Motor effects of cueing in Parkinson’s rehabilitation

Uni e sidade do Minho
Escola de Engenha ia
Ma ia Miguel Sá Ma ques Pimen a
Mo o e ec s o Cueing in Pa kinson’s
ehabili a ion
Oc obe 2024
Mo o e ec s o Cueing in Pa kinson’ s
ehabili a ion
Ma ia Miguel Sá Ma ques Pimen a
UMinho | 2024
Ma ia Miguel Sá Ma ques Pimen a
Mo o e ec s o Cueing in Pa kinson’s
Rehabili a ion
Oc obe 2024
Mas e ’s Disse a ion
Mas e Deg ee in Biomedical Enginee ing
Medical Elec onics B anch
Disse a ion supe ised by
P o esso Doc o C is ina Peixo o dos San os
i
DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO
POR TERCEIROS
Es e é um abalho académico que pode se u ilizado po e cei os desde que espei adas as eg as e
boas p á icas in e nacionalmen e acei es, no que conce ne aos di ei os de au o e di ei os conexos.
Assim, o p esen e abalho pode se u ilizado nos e mos p e is os na licença abaixo indicada.
Caso o u ilizado necessi e de pe missão pa a pode aze um uso do abalho em condições não
p e is as no licenciamen o indicado, de e á con ac a o au o , a a és do Reposi ó iUM da Uni e sidade
do Minho.
Licença concedida aos u ilizado es des e abalho
A ibuição
CC BY
h ps://c ea i ecommons.o g/licenses/by/4.0/
ii
STATEMENT OF INTEGRITY
I he eby decla e ha ing conduc ed his academic wo k wi h in eg i y. I con i m ha I ha e no
used plagia ism o any o m o undue use o in o ma ion o alsi ica ion o esul s along he p ocess
leading o i s elabo a ion. I u he decla e ha I ha e ully acknowledged he Code o E hical Conduc o
he Uni e si y o Minho.

iii
AGRADECIMENTOS
An es de mais, peço desculpa po des oa es a secção das es an es colocando-a em po uguês,
mas ealmen e só a ia sen ido se assim o osse. Es e ano oi um ano de mui o c escimen o pessoal.
Esc e e uma disse ação sem dú ida oi uma a e a que posso desc e e como complicada, não pelo
p oje o em si po que como se cos uma dize “quem co e po gos o não cansa”, mas po exigi
momen os mais soli á ios que não combinam an o comigo. No en an o, oi bas an e menos complicada
po e ido a so e de se acompanhada pelas pessoas ce as. E como es e abalho não e ia sido
possí el ealiza sem a ajuda di e a ou indi e a dessas pessoas, me ecem ambém es a aqui
mencionadas.
P imei amen e, gos a ia de ag adece à minha o ien ado a, a p o esso a C is ina, po odas as
opo unidades, po e ac edi ado em mim mesmo quando eu p óp ia pode ia acha que não se ia capaz,
e po e me pe mi ido abalha di e amen e com pessoas excecionais, de onde salien o a C is iana. A
C is iana oi sem dú ida essencial pa a odo o deco e des e p oje o, e um pa ág a o ou uma página
nunca se ão ag adecimen o su icien e. Ob igada po odos os ensinamen os, po odo o empo in es ido,
e p incipalmen e pela paciência pa a me a u a semp e.
De seguida, que o ag adece aos meus pais, po apoia em semp e odas as minhas decisões e
me pe mi i em ealiza as mesmas, e ainda à Ju, ao Luís e ao Sco , po se em o meu po o de ab igo
cons an e. Ag adeço ainda à es an e amília e amigos po es a em semp e p esen es e disponí eis,
inclusi e pa a se cobaias quando necessá io.
E como a ese não e ia sido ei a sem e passado pelos es an es anos do cu so, gos a ia de
ag adece ambém a odas as pessoas que o ma ca am. An es de mais, ag adece ao pes e po odas
as ba alhas conquis adas jun os, em especial ao Kim, o meu pa cei o ma a ilha na lu a que oi o 1.º ano
de mes ado. Às meninas, Demi, Pebbles, Red, Ma y e Api a, ag adeço po es a em semp e lá mesmo
não p ecisando de es a , po me aze em e o copo meio cheio quando es ou a olha pa a o azio, e po
me aze em en ende o e dadei o signi icado de amizade odos os dias. Res a-me ambém ag adece à
Bia, po se o sinónimo li e al de casa, po odas as boleias, e po celeb a semp e odas as i ó ias
comigo po mais pequeninas que sejam. Todos ocês são especiais, e le a - os-ei comigo semp e.
Pa a e mina , gos a ia de ag adece imensamen e a odos os pacien es de Pa kinson com os
quais i e a opo unidade de abalha , pela disponibilidade, in e esse e po o na em es e p oje o
ealmen e g a i ican e. A odos um eno me ob igada, e espe o que es e abalho os deixe ão elizes
quan o me deixou a mim.
i
RESUMO
A doença de
Pa kinson (PD),
é um dis ú bio neu odegene a i o que a e a p incipalmen e
neu ónios p odu o es de dopamina, aduzindo-se em sin omas mo o es incapaci an es como
pe u bações da ma cha e
eezing o gai (FoG)
[1], [2]. O
Pa kinson
impac a a au onomia dos pacien es
ao ealiza a e as mo o as diá ias, diminuindo assim a sua qualidade de ida [2], [3]. Es a égias de
aplicação de pis as senso iais em disposi i os es i éis, su gem como possibilidades p omisso as pa a
melho a a qualidade de ida (
QoL
) [3]–[5]. Pe sonalizadas pa a as necessidades dos pacien es de
PD,
pe mi em es imula di e en es caminhos neu ológicos pa a ge a espos a mo o a [6]. As pis as podem
assim ajuda os pacien es a es abelece con olo mo o e melho a a mobilidade dos mesmos, su gindo
como uma abo dagem holís ica pa a os e igo a e es au a a sua independência [5], [7].
Es e p oje o p e ende a alia o impac o das pis as senso iais, ocando-se no eino con ínuo com
a
+sensBand
em pacien es de
PD
, explo ando o seu po encial pa a melho a o con olo mo o e eduzi
episódios de
FoG
. A
+sensBand
é um disposi i o de
bio eedback
es í el, na o ma de uma cin a,
p oduzido pelo
BiRDLaB
, desenhado pa a po encializa eabili ação, moni o ização mo o a e ainda
adminis ação de pis as á eis ou isuais, pe sonalizada em pacien es de
PD
. O es udo em como obje i o
comp eende os e ei os mo o es do uso de di e en es es a égias da
+sensBand
, implemen ando no as
ocadas em sin omas mais a ançados como o
FoG
, e es ando-as a longo p azo sob e os pacien es,
a a és da es u u ação e condução de um p o ocolo pa a um es udo longi udinal clínico, e consequen e
análise dos dados adqui idos.
Os es udos de alidação ealizados demons a am que as pis as ib o á eis an o em
open
como
closed loop
, melho a am e icazmen e o desempenho da ma cha, ao abo da os aspe os hipociné icos e
b adiciné icos em pacien es de
PD
. A combinação de pis as isuais e soma ossenso iais, de ambos os
mé odos, numa no a es a égia ocada na p e enção de
FoG
, oi alidada com ele ada acei ação em
pa icipan es saudá eis, endo comp o ado o po encial p omisso pa a a eabili ação mo o a em
di e en es cená ios da ida diá ia. Es es esul ados sus en a am a sua aplicação a longo p azo com os
pacien es, endo a sua e icácia dependido de um eino consis en e. Adicionalmen e, o eino con ínuo
com es a es a égia eduziu as oco ências de
FoG
, melho ou o desempenho mo o e a
QoL
dos pacien es
de
PD
, des acando o impac o posi i o des a es a égia no quo idiano des es pacien es.
Pala as-cha e: Doença de Pa kinson; es udo longi udinal;
F eezing o Gai
; pis as senso iais;
Qualidade de ida
ABSTRACT
Pa kinson's Disease (PD) is a neu odegene a i e diso de ha a ec s dopamine gic neu ons,
leading o debili a ing mo o symp oms like gai dis u bances and F eezing o Gai (FoG) [1], [2]. PD
impac s pa ien 's au onomy pe o ming daily asks, diminishing hei o e all quali y o li e (QoL) [2], [3].
Cueing s a egies implemen ed on wea able de ices, eme ge as p omising possibili ies o enhancing
QoL [3]–[5]. Tailo ed o he needs o PD pa ien s, allow o explo e di e en neu ological pa hways o
gene a e mo o esponse [6]. Cueing s a egies can help PD pa ien s egain mo o con ol and imp o e
mobili y, o e ing a holis ic app oach ha empowe s pa ien s and es o es independence [5], [7].
This p ojec aims o assess he impac o cueing s a egies, ocusing on he use o +sensBand
by PD pa ien s du ing se e al sessions, explo ing i s po en ial in imp o ing mo o con ol and educing
FoG episodes. The +sensBand is a wea able cueing sys em, an ins umen ed wais band om BiRDLab
designed o enhance ehabili a ion, mo o assis ence, and pe sonalized clinical managemen on PD
pa ien s. The s udy aims o unde s and he mo o e ec s o di e en +sensBand s a egies, implemen
new s a egies ocused on a ge ing mo e ad anced symp oms such as FoG, and es ing he long- e m
ehabili a i e po en ial o he sys em in PD pa ien s by designing and conduc ing a p o ocol o a
longi udinal clinical s udy and consequen analysis o collec ed da a.
The alida ion s udies conduc ed demons a ed ha bo h open and closed loop ib o ac ile
app oaches e ec i ely imp o ed gai pe o mance by add essing hypokine ic and b adykine ic aspec s in
PD pa ien s. Fu he combina ion o bo h isual and soma osenso y cues a open and closed-loop, in a
no el s a egy ocused in add essing FoG, was alida ed wi h high accep abili y by heal hy pa icipan s
and p omising po en ial o mo o ehabili a ion in di e en scena ios o daily li ing. These indings
suppo ed i s long- e m applica ion wi h end-use s, wi h e ec i eness depending on consis en and
app op ia e use o cueing. Addi ionally, con inuous aining wi h he s a egy allowed imp o emen s in PD
pa ien s FoG ocu ences, mo o pe o mance and quali y o li e, highligh ing he posi i e impac o his
s a egy on daily li ing o PD pa ien s.
Keywo ds: Cueing; F eezing o Gai ; longi udinal s udy; Pa kinson’s Disease; Quali y o Li e
i
CONTENTS
1. INTRODUCTION .................................................................................................................... 1
1.1. MOTIVATION .......................................................................................................................................... 1
1.2. PROBLEM STATEMENT .......................................................................................................................... 4
1.3. DISSERTATION GOALS AND RESEARCH QUESTIONS ............................................................................. 4
1.4. CONTRIBUTION TO KNOWLEDGE .......................................................................................................... 7
1.5. PUBLICATIONS AND ORAL COMMUNICATIONS [24] .............................................................................. 7
1.6. MANUSCRIPT OUTLINE ......................................................................................................................... 8
2. LITERATURE REVIEW ON CUEING STRATEGIES FOR PD PATIENT’S GAIT REHABILITATION... 9
2.1. METHODOLOGY..................................................................................................................................... 9
2.1.1. RESEARCH METHODOLOGY .................................................................................................. 9
2.1.2. SELECTION STRATEGY ........................................................................................................ 10
2.1.3. DATA EXTRACTION .............................................................................................................. 10
2.2. RESULTS ............................................................................................................................................. 10
2.2.1. TECHNICAL SPECIFICATIONS .............................................................................................. 11
2.2.2. CLINICAL SPECIFICATIONS .................................................................................................. 27
2.3. DISCUSSION........................................................................................................................................ 40
2.3.1. RQ1: HOW HAVE CUEING STRATEGIES BEEN USED AND APPLIED FOR TARGETING MOTOR
SYMPTOMS OF PD? ............................................................................................................................ 40
2.3.2. RQ2: WHICH IS THE EFFECTIVENESS OF CUEING STRATEGIES CONCERNING MOTOR
EFFECTS AND QOL? ........................................................................................................................... 42
2.4. CONCLUSIONS .................................................................................................................................... 45
3. SOLUTION OVERVIEW ......................................................................................................... 47
3.1. HARDWARE ......................................................................................................................................... 47
3.2. CONTROL AND MONITORING STRATEGIES .......................................................................................... 48
3.2.1. MOTION MONITORING ......................................................................................................... 49
3.2.2. GAIT EVENT-DRIVEN CUEING ............................................................................................... 50
3.2.3. CONTINUOUS CUEING ........................................................................................................ 51
3.2.4. FOG PREVENTION CUEING .................................................................................................. 51
3.3. DESKTOP APP (MATLAB) ..................................................................................................................... 52
3.4. CONCLUSIONS .................................................................................................................................... 55
4. CROSS-SECTIONAL STUDY.................................................................................................. 57
4.1. METHODOLOGY................................................................................................................................... 57
4.1.1. PARTICIPANTS ..................................................................................................................... 57
xiii
JoLO – Judgemen o Line O ien a ion
L
LEDD – Le odopa Equi alen Daily Dosage
M
MBEST – Mini Balance E alua ion Sys ems
MDS-UPDRS – Mo emen Diso de Socie y-Uni ied Pa kinson’s Disease Ra ing Scale
MoCA - Mon eal Cogni i e Assessmen
MMSE - Mini-Men al S a e Examina ion
N
NFOG-Q – New F eezing o Gai Ques ionnai e
P
PD – Pa kinson’s Disease
PDQ-39 – Pa kinson’s Disease Ques ionnai e
PIGD – Pos u al Ins abili y and Gai Diso de
Q
QoL – Quali y o Li e
R
RAS – Rhy hmic Audi o y S imula ion
RCT – Randomized Con olled T ial
S
SD – S anda d De ia ion
SIP – S epping-in-place
SNC – Senio Neu ological Campus
STAI – S a e-T ai Anxie y In en o y
SUS – Sys em Usabili y Scale
T
TEDD – To al Elec ical Ene gy Deli e ed

xi
U
UMinho - Uni e si y o Minho
1
1. INTRODUCTION
The p esen epo aims o ou line he wo k es ablished, along wi h espec i e indings, o a
disse a ion in he scope o he second yea o he mas e ’s p og am in Biomedical Enginee ing – Medical
Ele onics ield, a he Uni e si y o Minho (UMinho), en i led “Mo o e ec s o Bio eedback in Pa kinson’s
ehabili a ion”. This disse a ion was designed o be accomplished du ing he academic yea o
2023/2024, a he Biomedical Robo ic De ices Labo a o y (BiRDLab) – Cen e o
Mic oElec oMechanical Sys ems (CMEMS), esea ch cen e o UMinho. Expe imen al ials we e also
execu ed in he scope o his disse a ion, aking place in bo h he Clinical Academic Cen e (2CA) a he
Hospi al o B aga, and he Senio Neu ological Campus (CNS), o B aga.
The wo k de eloped was supe ised by a obo ics p o esso a UMinho and p incipal in es iga o
o BiRD Lab, C is ina San os; and a Ph.D. biomedical enginee s uden also a BiRDLab, C is iana
Pinhei o. The p ojec also coun ed wi h he suppo o a neu ologis a B aga Hospi al, D ª Ana Ma ga ida
Rod igues, which pa ne ed wi h he p ojec ec ui ing PD pa ien s o he clinical ials in he scope o
his disse a ion.
1.1. MOTIVATION
Pa kinson’s Disease (PD) s ands as he second mos p e alen neu odegene a i e condi ion,
ollowing Alzheime ’s, impac ing pa icula ly he aging popula ion [8], [9]. The e o e, PD becomes an
exace ba ing conce n as he wo ld demog aphics con inue o shi owa ds an inc easingly aged
popula ion. This ansi ion no only esul s in an inc ease in he p e alence o he disease bu also ele a es
i s economic implica ions, ul ima ely posing as a subs an ial bu den on heal hca e sys ems wo ldwide
[1], [9].
PD is classi ied as a synucleinopa hy, a neu odegene a i e diso de cha ac e ized by he loss o
dopamine-p oducing neu ons wi hin he nig a egion o b ain [1], [2]. Al hough he causes o Pa kinson's
disease (PD) emain a challenge o asce ain, i is known ha abno mal alpha-synuclein p o ein buildup
in he subs an ia nig a causes dopamine-p oducing neu ons o degene a e, he eby lowe ing dopamine
le els [1]. The di ec and indi ec channels ha pass h ough he s ia um a e impac ed by his dec ease,
which in u n a ec s an indi idual's mo o esponse and, he e o e, his mo o ac i i y [1], mani es ing as
debili a ing mo o symp oms such as F eezing o Gai (FoG), b adykinesia (slowness o mo emen ),
akinesia (inabili y o olun a ily pe o m a mo emen ), emo s (in olun a y epea ed hy hmic
mo emen o a body pa ), igidi y (con inuous in olun a y muscle con ac ion), and impai ed mo o
2
e lexes [1], [8]–[10]. E en hough each pa ien a ec ed may expe ience di e en mo o symp oms, gai
dis u bances limi au onomy in ca ying ou e e yday cho es. As a esul , become mo e dependen on
help and Quali y o Li e (QoL) declines [7].
FoG is cha ac e ized by he sudden inabili y o gene a ing mo emen , he e o e eezing in place,
and is one o he main symp oms ha g ea ly a ec s QoL [3]. No all PD pa ien s’ expe ience FoG, since
i is a symp om ha mani es s in mo e ad anced s ages o he disease [11]. Howe e , i is a igh ening
eali y o pa ien s o conside has he disease ad ances. FoG mani es s spon aneously, being mainly
igge ed by ce ain asks like u ning o gai ini ia ion, bu also spa ial cons ain and s ess si ua ions,
like s ai cases and passing h ow na ow pa hs, as o example doo ames [3], [10]. The episodic na u e
o he symp om makes i one o he mos challenging ones o unde s and, bu also among he mos
dange ous o pa ien s since i can esul in alls [10].
Howe e , s udies ha e shown ha hese gai impai men s in PD pa ien s a e no solely mo o
phenomena. Ins ead, hey may a ise om an in ica e in e play among senso imo o , cogni i e, and
emo ional ac o s [11], [12]. This in e ac ion be ween hese domains leads o he exp ession o a b oade
spec um o symp oms in PD, including sleep dis u bances (insomnia), mood diso de s (anxie y and
dep ession), cogni i e impai men (a en ion de ici s), psychosis and con usion and e en au onomic
dys unc ion (balance impai men s and pos u al ins abili y) [1], [13].
A ho ough comp ehension and esea ch o hese symp oms in essen ial o disco e e ec i e
he apeu ic solu ions ocused on enhancing pa ien s’ QoL.
Despi e ongoing esea ch e o s and ad ances in unde s anding he disease, he e hology and
pa hophysiology o PD emains enigma ic, making i s managemen and ea men a o midable challenge.
The e o e, he need o comp ehensi e esea ch o unco e he unde lying mechanisms and explo ing
po en ial he apeu ic pa hways has ne e been mo e c ucial [3].
PD emains an incu able condi ion, unde sco ing he need o de elop he apeu ic in e en ions
ocused on enhancing pa ien s Quali y o Li e (QoL) [2], [3]. In he p esen days, pha maceu ical
in e en ions wi h dopamine gic medica ion, such as d ug deli e y using le odopa o ele a e dopamine
le els in he b ain, emains he p ima y ea men modali y [4], [8]. A c i ical challenge in PD lies in he
issues p esen ed by ce ain symp oms, such as wo sen FoG and inducing dyskinesias (abno mal
in olun a y mo emen s) associa ed wi h dopamine gic medica ion usage [8], [14]. These mo o
luc ua ions comp ehend an obs acle o op imal ea men , limi ing he e icacy o con en ional he apies,
hence, i is impe a i e o explo e he apeu ic al e na i es ha can mi iga e hese challenges.
3
Ne e heless, he e is a g owing in e es in al e na i e and complemen a y he apies, such as
Deep B ain S imula ion (DBS), occupa ional he apy, and physical he apy, which ha e gained
p ominence in ecen yea s [4], [8]. Howe e , hese he apies also en ail d awbacks, such as ela i ely
high cos o long- e m expenses [15]. Physical and occupa ional he apies, o be mo e e ec i e equi e
mo e expe ienced he apis s, he e o e being subjec i e which can also lead o slowe p og ess, ha can
a ec pa ien s’ mo i a ion [15]. Bo h me hods by depending on he apis don’ allow pa ien s o eel a
sense o au onomy, which is also essen ial o mo i a ing hem o pe o m he ea men s [16]. Some
me hods like DBS a e in asi e needing su ge y, which en ails associa ed isks. Addi ionally, he e a e
pa ien s ha a e no sui able o nei he al e na i e due o he he e ogeneous pa hophysiology challenge
o he disease, so i is necessa y o gi e pa ien s mo e op ions de eloped aking accoun abili y o hese
lapses [3].
Cueing s a egies a e he apeu ic app oaches ha eme ge as p omising al e na i es which ha e
g ea po en ial in ea ing PD [3], [4]. Cueing Sys ems (CSs) help pa ien s execu ing mo o asks, p o iding
ex e nal audi o y, isual, o soma osenso y cues deli e ed a a ime o dis ance pace, o con inuous
enhancemen o mo o ac i i y, dependen o independen o he end-use mo o beha iou [5]. Cueing
s a egies can be ca ego ized in wo g oups, depending on how hey a e deli e ed: open-loop o closed-
loop. Open-loop cueing p o ides cues a a ma ked pace, independen ly o he use s. On he o he hand,
closed-loop cueing is dependen o end-use , by adap ing cueing hy hm o he end-use s’ beha iou [17].
Bio eedback is a ype o closed-loop cueing s a egies since p o ides he oppo uni y o moni o and
con ol, o e ing eal- ime eedback, ha assis s pa ien s in becoming awa e and sel - egula ing hei
mo o unc ions [7]. Bo h me hods can be used sepa a ely and combined, allowing o moni o pa ien s’
eac ions o cueing du ing gai , which a ises g ea po en ial since p o ides a cus omized cou se o
ea men o he pa ien [7].
Cueing s imula es a di e en pa hway hen he one a ec ed by PD. Di e se cueing s a egies,
such as Rhy hmic Audi o y S imula ion (RAS), can ac i a e channels om di e en b ain egions no
di ec ly co ela ed wi h dopamine gic pa hway’s, including ce ebella and p e on al channels [6]. Thus,
since hese al e na i e pa hways a e spa ed du ing he diseases’ p og ession, e ain po en ial o e ec i e
compensa ion and es o ing o pace and hy hm du ing gai [6]
Wea able de ices a e he u u e o a -home o daily ea men op ions [7]. Mo o acking sys ems,
such as wea able ine ial senso s, allow he analysis o pa ien s’ gai , p o iding insigh s in o di e en gai
pa ame e s such as spa io empo al pa ame e s, kinema ic pa ame e s, An icipa o y Pos u al
Adjus men s (APAs) and du a ion and numbe o FoG episodes [4], [5], [7]. Closed loop cueing sys ems,
4
by in eg a ing mo o acking senso s wi h ac ua o s capable o deli e ing audi o y, isual o ib o ac ile
cues, a ise he po en ial o guiding and co ec ing gai impai men s o p oac i ely add essing mo o
symp oms p io o hei mani es a ion [4], [5], [7]. The e o e, closed loop cueing sys ems come o he
o e on as a pionee ing solu ion wi hin his amewo k [7].
1.2. PROBLEM STATEMENT
Despi e he a ailabili y al eady o exis ing he apeu ics and echnological solu ions o ea ing
Pa kinson, i is impe a i e o explo e no el op ions ha can enhance pa ien s QoL [3]. The e o e, long-
e m wea able de ices a e ising as a po en ial solu ion o a mo e con enien , low-cos , and po able a -
home he apy [7]. Howe e , i is essen ial o s udy and e alua e hei e icacy and hei impac on he
disease o de e mine hei iabili y as a po en ial ea men , since he aining o daily asks emains a
challenge [7].
The e a e s ill gaps ha need o be illed, ega ding some cueing s a egies, like ib o ac ile cues
[18]. Despi e di e en ypes o cueing s a egies al eady being es ed, he e is lack o in o ma ion abou :
he ypes o cueing, and con ol s a egies, ha ha e be e esul s in ea ing gai impai men s o PD
pa ien s; and he long- e m e ec s in enhancing PD pa ien s mo o e ec s and hus QoL [7], [18].
Since some cueing s a egies, like ib o ac ile cues, yield limi ed esul s also since benchma k
has no been accomplished, esul ing in lack o conclusions abou hei e iciency in ea ing PD
symp oms, i is impo an o explo e he po en ial o his ype o cueing [18], [19]. The e o e, in his
p ojec i will be analysed he po en ial o he +sensBand, a wea able de ice, de eloped in BiRDLab, ha
uses ib o ac ile cueing o enhance PD pa ien s’ ehabili a ion [19].
To ully comp ehend he po en ial o cueing s a egies, i is c ucial ha clinical ials a e
conduc ed [7]. Howe e , elying solely on single-use ials p o es insu icien , gi en he abundance o
exis ing esul s. The nex s ep is o conduc long- e m s udies, designed o assess he iabili y o
in eg a ing hese s a egies o ongoing u iliza ion and in o daily asks [7]. Only h ough such means will
i be possible o inno a e and s a impac ing pa ien s QoL.
1.3. DISSERTATION GOALS AND RESEARCH QUESTIONS
The main goal o his disse a ion is he de elopmen and consequen analysis o a longi udinal
clinical s udy o Cueing s a egies in PD pa ien s, e alua ing he long- e m use mo o and QoL impac and
po en ial o wea able cueing sys ems, using +sensBand. The +sensBand is a plug-and-play wea able

5
cueing sys em c ea ed by BiRDLab. I in eg a es an ine ial senso and a minia u ized came a, o ack
and moni o use s’ mo emen s, and ea u es ib a ion senso s ha p o ide soma osenso y cueing,
aiming o coun e ac any symp oms ha may a ise du ing gai asks, such as FoG.
To achie e he main goal, se e al sub-objec i es ha e been de ined, wi h some espec i e key
pe o mance indica o s (KPI’s), o accomplish g adually along his disse a ion:
• Objec i e 1: Li e a u e e iew and analysis ocused on Pa kinson clinical
s udy, using cueing s a egies. This e iew will in es iga e how cueing s a egies
ha e been used (audi o y, soma osenso y, and isual) and applied (single use sessions
o longi udinal s udies) o comba ing mo o symp oms o PD and hei e ec i eness
(e.g., e ec s o e clinical scales, APAs, s ep and s ide leng h, cadence, eloci y, numbe
o FoG episodes). A clea summa y o he main cueing s a egies implemen ed,
speci ica ions and espec i e ou comes e iewed should be accomplished (KPI 1).
• Objec i e 2: Familia iza ion wi h he +sensBand sys em and examina ion o
con ol s a egies sui able o long- e m clinical ials wi h Pa kinson.
Th oughou he p ojec , he in en ion is o le e age he p e-exis ing echnology,
+sensBand, and subsequen ly analyse i s po en ial, alongside no el cueing s a egies
de eloped based on he conduc ed li e a u e e iew. The e o e, i is impe a i e o
conside an acquain ing phase dedica ed o unde s anding he de ice and i s ope a ion
mode. Mo eo e , amilia iza ion will include he pa icipa ion on he conduc ion and da a
analysis o a cu en ly in cou se c oss-sec ion clinical s udy wi h +sensBand in Pa kinson.
As o KPI 2, i included he summa y o he di e en con ol s a egies o +sensBand
ha should be conside ed o he u he p o ocol de elopmen and included also da a
collec ion o 15 mo e pa ien s gai me ics o comple e he on-going c oss-sec ional s udy
and i s analysis.
• Objec i e 3: De elopmen and alida ion o a no el cueing s a egy adequa e
o a ge FoG. This s a egy will be designed o PD pa ien s in mo e ad anced s ages
o he disease, based on he s a egies al eady de eloped and li e a u e conclusions,
allowing a mo e comple e and in ense s a egy a ge ing FoG. I is expec ed o pe o m
a c oss-sec ional s udy, add essing he usabili y and wo kload o he s a egy, o g an
ha should be sui able o use in scena ios o daily li ing. This objec i e will be measu ed
by KPI 3: (i) implemen a ion o a no el cueing s a egy o assis PD pa ien s wi h FoG;
6
(ii) implemen a ion o bo h isual and soma osenso y cues simul aneously; and (iii)
g an ing ha he s a egy does no o e s imula e pa ien s wi h SUS sco es abo e 65
poin s and acco ding o he espec i e indi idual i ems e e ence benchma ks [20].
• Objec i e 4: De elopmen , alida ion and conduc ion o he p oposed
p o ocol wi h end-use s. This p o ocol will be designed o PD pa ien s, based on he
li e a u e conclusions, unc ionali ies o +sensBand, and by se ing he necessa y ade-
o s wi h clinicians. I is expec ed o pe o m a longi udinal s udy, add essing he
limi a ions bu also po en ial o mo o and QoL e ec s o long- e m cueing use. This
objec i e will be measu ed by he de elopmen o a con inuous s udy ci cui ha will
in eg a e di e en mo o asks suscep ible o inducing FoG. To e alua e he di e en
aspec s o he de ice, i is impe a i e o pe o m he mul iple aining sessions
(longi udinal s udy) using he +sensBand de ice and include e alua ion ime poin s o
acqui ing senso based and clinical da a, assessing he mo o (FoG episodes and
spa io empo al pa ame e s) and QoL e ec s o he aining. Du ing he p o ocol
conduc ion i should be accomplished a educ ion o pa ien ’s FoG episodes (KPI 4).
• Objec i e 5: Da a analysis and discussion ega ding he mo o and QoL
e ec s o long- e m cueing use in Pa kinson. The da a collec ed om he p o ocol
conduc ion should be examined and discussed, o d aw conclusions abou he cueing
long- e m use e ec s o e PD pa ien s’ symp oms. This goal will be measu ed
conside ing KPI 5, indica ing ha i should be accomplished spa io empo al di e ences
o a leas 30% be ween s udy g oups [21], PDQ-39 sco e minimal clinical imp o emen s
o a leas 4.72 poin s [22], and minimal NFOG-Q sco e changes o a leas 3 poin s [23].
• Objec i e 6: Dissemina e esea ch by w i ing and p esen ing disse a ion
documen , and a icles (KPI6).
Fu he mo e, o encou age he esea ch o he li e a u e e iew, his disse a ion aimed o
answe he ollowing que ies:
• Resea ch Ques ion 1: How ha e cueing s a egies been used and applied o a ge ing
mo o symp oms o PD?
7
• Resea ch Ques ion 2: Which is he e ec i eness o cueing s a egies conce ning
mo o e ec s and QoL?
• Resea ch Ques ion 3: How does +sensBand cueing sys em long- e m use a ec
Pa kinson’s mo o symp oms and QoL?
1.4. CONTRIBUTION TO KNOWLEDGE
The wo k done du ing his disse a ion con ibu es wi h no el cueing solu ions based on wea able
de ices o imp o e mo o symp oms o PD pa ien s. The main con ibu ions ou lined a e:
❖ A li e a u e e iew abou clinical s udies using senso y cueing solu ions o PD mo o
ehabili a ion, compiling in o ma ion abou : di e en cueing de ices used, cueing modali ies used
and he way hey we e deli e ed, senso s used o da a collec ion, sample sizes and s udy g oups
conside ed, p o ocol du a ion, aining dosage and ou comes measu ed (Chap e 2, Objec i e
1).
❖ Finding how cueing s a egies ha e been applied o a ge PD mo o symp oms and hei
e ec i eness ega ding hose e ec s bu also QoL (Chap e 2, Objec i e 1).
❖ Implemen a ion o a no el senso y cueing s a egy (FoG p e en ion cueing) in eg a ed on he
+sensBand de ice designed o a ge FoG in PD pa ien s (Chap e 3, Objec i e 3).
❖ Imp o emen o an ongoing s udy, esul ing in a jou nal pape o be published, abou he
e iciency o using di e en soma osenso y s a egies in PD pa ien s, compa ing he e iciency o
con inuous hy hmic ib a ion wi h ib a ions deli e ed a a pe sonalized a e acco ding o
pa ien s' inal con ac s o he igh oo (Chap e 4, Objec i e 2).
❖ Valida ion o he usabili y and wo kload o he no el senso y cueing s a egy, esul ing in he
publica ion o a con e ence pape (Chap e 5, Objec i e 3) and u he alida ion o he
s a egy es ing in PD pa ien s (Chap e 6, Objec i e 3).
❖ A longi udinal s udy abou he use o he no el cueing s a egy in PD pa ien s o mo o
ehabili a ion and QoL imp o emen s, ega ding he long- e m e ec s as a po en ial solu ion o
mi iga ing FoG (Chap e 6, Objec i es 4 and 5).
1.5. PUBLICATIONS AND ORAL COMMUNICATIONS [24]
This disse a ion esul ed in wo a icles: one con e ence pape as i s au ho en i led “Robo ic
senso y cueing du ing gai : a usabili y s udy wi h scena ios o daily li ing”, IEEE In e na ional Con e ence
8
on Au onomous Robo Sys ems and Compe i ions, Pa edes de Cou a, Po ugal, 2024, Ma ia Pimen a,
C is iana Pinhei o, C is ina P. San os [24]; and one jou nal pape , as i s au ho , cu en ly unde e ision
en i led “Gai e en -d i en bio eedback o imp o e gai impai men s in Pa kinson’s disease”,
Mo emen
Diso de s, O icial Jou nal o he In e na ional Pa kinson and Mo emen Diso de Socie y (impac ac o
o 7.4), 2024, Ma ia Pimen a, C is iana Pinhei o, Helena R. Gonçal es, Ana Ma ga ida Rod igues,
C is ina P. San os.
Following he “Robo ic senso y cueing du ing gai : a usabili y s udy wi h scena ios o daily
li ing”[24], besides he published con e ence a icle, his wo k also esul ed in an o al p esen a ion du ing
he ICARSC 2024 con e ence.
1.6. MANUSCRIPT OUTLINE
This documen is o ganized in o se en chap e s. Chap e 1 p esen s he mo i a ion o his
disse a ion, as well as he p oblem s a emen , objec i es o accomplish, he con ibu ion o his wo k o
u he esea ch and he esul ing publica ions and p esen a ions. In he ollowing chap e , Chap e 2,
he s a e o a is ou lined p esen ing he li e a u e e iew ound un il he p esen momen . Chap e 3
p esen s a concep ual o e iew o he main componen s, ha dwa e and so wa e, used du ing his
disse a ion.
Chap e s 4, 5 and 6 con ain he desc ip ion and esul s o he h ee di e en s udies done,
espec i ely: a c oss-sec ional s udy ha compa ed di e en cueing s a egies in PD pa ien s; a usabili y
and men al wo kload s udy o a no el cueing s a egy pe o med wi h heal hy pa icipan s; and a
longi udinal s udy using his no el s a egy in PD pa ien s wi h FoG.
Las ly, Chap e 7 esembles he conclusions o all he wo k done o his p ojec .
15
S udies
(yea )
Senso y cues
Gai assessmen senso s
De ice
ype
De ice desc ip ion
Loca ion
Mode
De ice
Loca ion
Pa ame e
Ginis e
al.
(2016)[51
]
CuPiD: audiobio eedback (ABF-gai app) and FOG
aining (FOG-cue app) using a sma phone (Galaxy
S3-mini, Samsung, Sou h Ko ea),
Ea phones o sma phone
speake
2 IMUs (EXLs3, I aly)
Shoes using ABF app, and abo e ankles
using FOG app
Spa io empo al ( eloci y, cadence, s ide-leng h)
Clinical (2-min Walk Tes , MBEST, FSST, FES-I,
PASE)
Tosse am
s e al.
(2022)[52
]
Me onome
NM
3-dimensional mo ion cap u e analysis
(VICON om Ox o d Me ics, UK)
16 ma ke s placed ollowing he Plug-in
Gai Lowe Body Model
Spa io empo al (gai a iabili y, s ide ime)
Capa o e
al.
(2020)[53
]
Speake (JBL Go Po able Wi eless Speake ):
me onome sound
NM
Video came as
NM
Clinical (MBEST, TUG dual ask pe o mance,
UPDRS)
Chang e
al.
(2019)[54
]
Me onome
NM
EMG h ough su ace ele odes
Tibialis an e io muscle
Physiological (MEP, CSP, AMEP, SICI, ICF)
Kis le o ce pla e (Kis le AG, Swi ze land)
On he loo
Spa io empo al (SIP pe o mance)
GAITRi e Analysis Sys em
Pa hway
Spa io empo al (gai speed, a iabili y, cadence,
and s ide leng h)
Li ani-
Sil a e
al.
(2017)[55
]
Non-
ele onic
Tex u ed insoles: wi h semiesphe ed ele a ions
Dis al phalanx o he
hallux, heads o
me a a sophalangeal
join s and heel
op oelec onic idimensional sys em
(No he n Digi al Inc.)
4 ma ke s: 5 h igh and 1s le me a a sal
join s and la e al ace o he igh
calcaneus and medial ace o he le
calcaneus
Spa io empo al (s ide leng h, du a ion and
eloci y, cadence, s ep wid h, s ance phase
du a ion and % o double-suppo phase du a ion)
Schlens e
d
e al.
(2020)[56
]
Ele onic
Vib oGai
W is s
3 Opal IMUs (APDM Inc., USA)
1 a pos e io unk (L5) and 1 a each
shin
Kinema ic (APAs)
Su ace ele omyog aphy (EMG, Wa e by
Come a)
Tibilias an e io and gas ocnemius
mediali bila e ally
Physiological (TFL muscle ac i i y)
Rosen hal
e al.
(2018)[57
]
CueS im (cus om-buil elec ical s imula o ): wo n
on he wais wi h skin su ace ele ods
Hams ings o quad iceps
o body side mos
a ec ed
NM ( eco dings)
NM
Spa io empo al ( ime o comple ing walking ask)
Clinical (VAS, FPR scale, numbe o FoG
episodes)
Peppe e
al.
(2019)[58
]
Equis asi®: wea able ocal mechanical ib a o y
de ice
C7 e eb a and
gas ocnemius
6-came as s e eopho og amme ic
sys em synch onized o 8-channel sEMG
sys em
4 muscles bila e ally: Rec us Femo is,
Tibialis An e io , Biceps Femo is and
Gas ocnemius La e alis
Physiological (Peak o he en elope and
occu ence)
El-
Tamawy
e al.
Vib a o y de ices (VDs) (OPTEC Co. L d., Japan)
In o-shoes
Qualysis P oRe lex mo ion analysis sys em
(Qualysis Medical AB, Sweden)
NM
Spa io empo al (cadence, s ide leng h)
Kinema ic (hip, knee, and ankle join s’ angula
excu sion)

16
S udies
(yea )
Senso y cues
Gai assessmen senso s
De ice
ype
De ice desc ip ion
Loca ion
Mode
De ice
Loca ion
Pa ame e
(2012)[59
]
Zhao e
al.
(2016)[60
]
Google Glass (Explo e e sion 2, XE 22.0, And oid
4.0)
Headse
+
7 IMU’s (MVN mo ion cap u e sui (Xsens,
Ne he lands))
Pel is, uppe legs, lowe legs and ee
Spa io empo al ( eloci y, cadence, s ide leng h
and a iabili y)
Vídeo came as
S a and midpoin o walkway
Clinical (p esence and se e i y o FoG)
Ca pinell
a e al.
(2017)[61
]
Gamepad: 6 wea able ine ial senso s (TMA) and
PC wi h LCD sc een (16'')
NM
6 IMUs (TMA)
Uppe and lowe unk and lowe limbs
h ough elas ic bel s
Spa io empo al ( eloci y, balance (BBS), 10-min
walk es )
Yoge -
Seligmann
e al.
(2023)[62
]
Sel P o sys em: compu e , p ojec o and spa ial
came a (MS Kinec Azu e came a & So wa e
De elope Ki ; SDK)
P ojec ed ligh s ipes and me onome bea s
Placed a 2m om he
walking a ea (s ipes
p ojec ed a 110% o he
use ’s s ep leng h)
Spa ial came a
NM
Spa io empo al (s ep leng h and ime)
Lee e al.
(2023)[63
]
Google glass: cus om-made audi o y- isual cue
applica ions "Walk Wi h Me" and "Un eeze Me"
Headse
NM ( ideo aped)
NM
Spa io empo al ( ime o comple e ask)
Janssen
e al.
(2017)[64
]
Sma glasses
Headse
Vídeo came as
NM
Clinical (numbe and du a ion o FoG episodes)
Me onome
NM
17 IMU’s (XSens, Ne he lands))
Fee , lowe legs, uppe legs, pél is,
hands, o ea ms, uppe a ms, s e num,
shoulde s and head
Spa io empo al (s ide leng h, ime, cadence and
eloci y)
Yang e
al.
(2022)[65
]
Wi eless-con olled comme cial ea phones: music
wi h di e en hy hms
Ea s
Sma insoles: wi h 6 PSUs
1 PSU se a he heel and 5 a he
o e oo (3 a me a a sal heads and 2 o
de ec oo -in e sion and oo -e e sion)
Spa io empo al ( eloci y, a iabili y and s ide
leng h)
Cus om-made lase -ligh gene a o
Wais o op o shoes,
p ojec s line in on o
pa ien s
Jiang e
al.
(2006)[66
]
Ele onic
Piezo buzze connec ed o a s imula o
NM
Op o ak 3020 sys em (No he n Digi al
Inc., Canada): wi h 3 IREDs:
1 o each heel and 1 o e he zygoma ic
bone.
Spa io empo al (s ep ime and leng h, eloci y)
Non-
ele onic
S ipe ape lines
Placed a pa ien s s ep
leng h o 40% o hei
heigh
Fo cepla e (Ad anced Mechanical
Technology Inc., USA)
Whe e pa ien s s ood
Kine ic ( iming and magni ude o weigh shi and
push-o o ce)
Lee e al.
(2012)[67
]
Ele onic
Me onome
NM
3-dimensional compu e ized mo ion
analyze (Mo ion Analysis; CA)
19 ma ke s: lowe limbs and unk,
enabling 3D analysis o he oo , shank,
high, pel is, and unk body segmen s
Spa io empo al ( eloci y, cadence, s ide leng h,
s ep leng h, and double- and single-limb s anding
ime);
Kinema ic (deg ees o maximum pel ic il , hip
lexion, knee lexion, and ankle do si lexion in he
sagi al plane)
Clinical ( equency o FoG episodes, o al s ep
numbe s, and o al ime pe walking cycle)
Non-
ele onic
S ipe ape lines
Placed a no malized s ep
leng hs dis ance
17
S udies
(yea )
Senso y cues
Gai assessmen senso s
De ice
ype
De ice desc ip ion
Loca ion
Mode
De ice
Loca ion
Pa ame e
Sijobe e
al.
(2017)[68
]
Ele onic
Ele os imula o
Fee : anode (do sum)
ca hode (a ch)
Video came as
NM
Clinical (numbe and du a ion o FoG episodes)
Lase cane
P ojec ed ho izon ally
om cane ip on he loo
Mancini
e
al.
(2018)[69
]
Speake s
NM
8 Opal IMUs (APDM Inc., USA)
Bo h shins, ee , w is s, on he s e num
and pos e io unk (a ound L5)
Clinical (du a ion o FoG and FoG- a io,
smoo hness)
Vib oGai ib a ion senso s
W is
Kla e e
al.
(2023)[70
]
Me onome (Na u al me onome App)
NM
Vídeo came a
NM
Clinical (numbe o FoG episodes and du a ion)
Vib a ing socks: la mini ib a ion mo o (Ada ui
Mini Mo o Disc 1201)
Below a ch o bo h ee
17 IMU’s (Xsens, Ne he lands)
NM
Spa io empo al ( eloci y and a iabili y)
Sil a-
Ba is a e
al.
(2022)[71
]
Speake : deli e ed me onome bea
Cen e o labo a o y
8 IMU’s (APDM Inc., USA)
Bo h shins, ee , w is s, on he s e num
and pos e io unk (a ound L5)
Clinical (FoG index)
Vib ogai
Bo h w is s
Lu e al.
(2017)[72
]
Ve ical a ay o ligh emi ing diodes (CREE)
3m in on o pa icipan s
2 o ce pla es (9260AA; Kis le )
On he loo , whe e pa ien s a s ask
Kine ic (g ound eac ion o ces and CoP)
Kinema ic (APA’s iming and ampli ude)
Vib o ac ile ansduce (C-3 Tac o , Enginee ing)
Ankle (la e al malleolus)
o ini ial s ance leg
Speake s (HPG-100N, A las Sound)
NM
McCandle
ss e al.
(2016)[73
]
Me onome (Pe e son BodyBea Pulsing
Me onome)
Bel , placed o e igh
side o he pel is
+
10 came a Qualisys mo ion analysis sys em
13 ma ke s: ee , shanks, highs, pel is,
unk, head, a ms and o ea ms
Clinical (numbe o FoG episodes)
Spa io empo al ( i s and 2nd s ep leng h)
Lase cane (Lase Cane, U-S ep
P ojec ed line on he loo
4 o ce pla es (Ad anced Mechanical
Technology, Inc.)
NM
Kine ic ( o wa d and sideways CoM eloci y;
o wa d and backwa d sways and numbe o
sideways sways)
Nieuwboe
e al.
Ligh emi ing diode: lash
On glasses
Vi apo Ac i i y Moni o (TEMEC
Ins umen s Inc, Ne he lands), wi h 5
accele ome e s
1 on each leg (la e al high), 3 placed on
he lowe hi d o he s e num o ien a ed
Spa io empo al ( u ning speed and ime du ing
ask)
18
IMUs: Ine ial measu emen uni s; EEG: Ele oencephalog a y; NM: no men ioned; TUG: Timed Up an Go; EMG: ele omiog aphy; ML APAs: Medio la e al An e o pos e ial adjus men s; LED: Ligh emmi ing
diode; VDs: Vib a o y de ices; PSU: P essu e sensi i e uni s; CoM: Cen e o Mass; CoP: Cen e o P essu e; IREDs: In a eds;
: Visual cues; : Audi o y cues; : Soma osenso y cues.
S udies
(yea )
Senso y cues
Gai assessmen senso s
De ice
ype
De ice desc ip ion
Loca ion
Mode
De ice
Loca ion
Pa ame e
(2009)[74
]
Audio one
Via ea phone
in he sagi al, longi udinal, and
ans e se planes.
Minia u e cillinde : pulsed ib a ion
W is band
Supu i ad
a e al.
(2022)[75
]
Wais bel cueing de ice
LED lase
P ojec ed line in on o
pa ien
2m RS oo scan
Embedded in he cen e o walkway
Spa io empo al ( eloci y, s ide leng h, and
cadence)
Ele onic
buzze
NM
Vib a ion
mo o
Wais
19
SENSORY CUES
This li e a u e e iew in ends o explo e cueing s a egies in PD pa ien s. The e o e, one main
opic o conside is he di e en de ices used o cueing. By examining he p esen s udies, i was possible
o spli he cueing de ices in 2 ca ego ies: ele onic, conside ed when men ioned he use o a leas an
ele onic de ice p o iding he cueing mechanism (43 s udies) [32]–[54], [56]–[75]; and non-ele onic,
when i is no men ioned he de ice in i sel used o he cue does no need a obo ic de ice o be p o ided
(14 s udies) [26]–[32], [43]–[45], [52], [55], [66], [67]. I is wo h no ing ha 7 s udies, [32], [43]–[45],
[52], [66], [67], used bo h ypes o cueing de ices, one ele onic and ano he non-ele onic, and
compa ed he mo o e ec s o each cueing de ice in PD pa ien s.
Cueing s a egies can also be o ganized by he di e en cueing mode used. Essen ially, he e a e
3 ypes o cueing modes: isual cues, audi o y cues, and soma osenso y cues; depending on he sense
hey ac on, and i is possible o ac ua e using one single cueing mode o di e en modes combined.
Visual cueing was he mos used mode (33 s udies) [26]–[45], [60]–[68], [72]–[75]; ollowed by audi o y
cueing (24 s udies) [46]–[54], [60]–[67], [69]–[75]; and las ly soma osenso y cueing (13 s udies) [55]–
[59], [68]–[75]. Howe e , ou di e en combina ions we e ound wi hin his s udies, being he mos
common es ing bo h Audi o y and Visual cueing (8 s udies) [60]–[67]; ollowed by es ing all h ee
Audi o y, Visual and Soma osenso y cueing (4 s udies) [72]–[75], hen es ing Audi o y and
Soma osenso y Cueing (3 s udies) [69]–[71] and las ly Soma osenso y and Visual cueing (1 s udy) [68].
O he non-ele onic de ices used o cueing, all s udies used isual cues [26]–[32], [43]–[45],
[66], [67] excep one s udy ha used soma osenso y cueing [55] and ano he ha used audi o y cueing
[52]. The non-ele onic isual cueing de ices consis ed o ape lines (8 s udies) [27], [28], [30], [31],
[43], [44], [66], [67] o ape s a s a angemen s (1 s udy) [26] o isually help pa ien s pe o m he gai
ask by spa ially ma king he s eps o he u ning posi ions; checke ed loo pa e ns o assis pa ien s
loo isualiza ion, in ending o he squa es o assis oo placemen du ing gai (1 s udy) [32], objec s
such as wooden ods o gi e a 3D spa ial sensa ion o s ep posi ioning (1 s udy) [29], and an in e ed
wooden walking s ick, wi h a sla o wood a ached a he bo om o gi e a s ep line cue o spa ial guidance
and suppo , o pa ien s when ha ing FoG episodes (1 s udy) [45]. As o he non-ele onic
soma osenso y cueing de ice, Li ani-sil a
e al.
[55], used ex u ed insoles wi h 9mm o diame e sphe e
ele a ions, loca ed unde he dis al phalanx o he hallux, heads o me a a sophalangeal join s and heel,
o s udy he p op iocep i e plan a e ec s in gai . Las ly, Tosse ams
e al.
[52] used an in e nal cueing
20
echnique ha consis ed o he pa ien sel -coun ing and ma ching each s ep o he coun ing hy hm as a
non-ele onic cueing me hod o gai ehabili a ion.
As o he ele onic de ices, 22 s udies used audi o y cues [46]–[51], [53], [54], [60]–[65], [67],
[69]–[75], 21 s udies used isual cues [33]–[42], [60]–[65], [68], [72]–[75] and 12 s udies used
soma osenso y cues [56]–[59], [68]–[75]. O ha , 6 s udies used bo h isual and audi o y cueing
elec onic de ices [60]–[65], 4 s udies used all 3 modes o cueing [72]–[75], 3 s udies used bo h
soma osenso y and audi o y cue modes [69]–[71], and 1 used bo h isual and soma osenso y cueing
[68].
AUDITORY CUEING
The e iewed audi o y cueing elec onic de ices consis ed mainly o me onomes (11 s udies)
[47], [49], [52], [54], [62], [67], [70], [73], [53], [69], [71] o gi e empo al hy hmic inpu in PD pa ien s
by making hem ma ch each s ep o he me onome bea s. Simila ly, and wi h he same pu pose, 3
s udies used sma glasses wi h audi o y hy hmic cues, 1 using Cinop ics (Cinop ics, Maas ich , he
Ne he lands) [64], and 2 using Google Glass (Explo e e sion 2, XE 22.0, And oid 4.0) [60], [63]; 2
s udies used buzze s ha p oduced hy hmic beeps, 1 using a Piezo buzze connec ed o a s imula o
[66] and ano he using an elec onic buzze [75]; and 1 s udy used a hy hmic audi o y one deli e ed
ia ea phones [74]. All audi o y cueing de ices e iewed deli e ed open-loop cueing.
As o o he audi o y cueing s a egies used, 1 s udy used a piano a angemen o child en’s song
“Row, Row, Row, You boa ”, p o iding hy hm o help se he pace o he s ide [46]; 1 s udy used
speake s o deli e ed di e se imed ones [72]; o he s udy used an iPod Touch playing music o podcas s
as a ewa d when s epping co ec ly [50]; o he used a Gamepad sys em ha p oduced a sound ala m
o guide pa ien s when doing a w ong mo emen [61]; and las ly one s udy used a sma phone based
sys em, called CuPiD, wi h an audio eedback App (ABF-gai App) and an ins umen ed cueing App o
FoG aining (FOG-cue App) o gi e di e en audi o y s imuli [51]. Al hough Ho in
e al.
[46] being a he
ocused on he song choice and cha ac e is ics o he song o be used as a cueing s a egy, i was
conside ed ha i used an elec onic de ice, such as speake s o a compu e , o p o ide he music
a angemen .
Lu
e al.
[72] used speake s (HPG-100N, A las Sound) o deli e a wa ning and a go one. The
wa ning one was se a 80dB and 1000Hz, and he go one se a 90dB and 2000Hz, and pa ien s we e
ins uc ed o s a walking only a e hea ing he go one. The ones we e deli e ed using 3 di e en iming
p o ocols: a ixed delay; a coun down and a andom delay. The coun down consis ed o gi ing 3 wa ning

21
ones ollowed by a go one, each 1s apa , gi ing pa ien s he sensa ion o coun ing 3-2-1-go. The ixed
delay was based on a simila p inciple bu ins ead ga e a wa ning cue ollowed 3s la e wi h a go one.
As o he andom delay cue, a wa ning cue was played and 4 o 12s la e a go cue ollowed. Di e en
imings we e s udied o analyse i had any implica ions in gai ini ia ion o PD pa ien s.
No e ha in Chomiak
e al.
[50] an iPod Touch was s apped in pa ien s’ knee and pai ed o an
ea phones’ headse . Pa ien s we e di ided in wo g oups, one ha ecei ed music s imuli and o he ha
ecei ed podcas s imuli. In o de o main ain he music o podcas playing, pa ien s had o pe o m a
s epping-in-place (SIP) ask co ec ly, which means essen ially, hey had o s and on one leg while li ing
he o he . In his s udy he music and podcas cues we e gi en as a ewa d o co ec ask pe o mance,
howe e since music p o ides hy hmic inpu , and podcas s p o ide in o ma ion ha need mo e o
pa ien s’ a en ion, he wo g oups wi h di e en audi o y s imuli we e analysed and compa ed o
unde s and he di e en impac in gai aining.
VISUAL CUEING
The e iewed isual cueing ele onic de ices included: lase beam p ojec ion de ices (15 s udies),
such as wais bel s [43], [75], shoes [33], [40], lase canes [36], [45], [68], [73], wheeled lase walke s
[35], [37], [38], lase ankle b acele s [39] o jus lase gene a o s [34], [44], [65] a ached a di e en
posi ions such as wais , ches o ee o p ojec guide lines on he walking pa h, o help pa ien s isualize
each s ep posi ion; sma glasses (4 s udies) such as Cinop ics [64], Google Glass [60], [63] o HoloLens
[41], p o iding augmen ed and mixed eali y cues like s ipe lines, wa ning ligh lashes, oo s eps, 3D
ba s and s ai cases o analyse he e ec s o FoG inducing ambien s and spa ial awa eness cues in
mi iga ing FoG, and also compa ing he e ec s o audi o y and isual s imuli combined o be e spa ial
and empo al o ien a ion [60], [63], [64]; p ojec o s (2 s udies) such as Sel P o sys em [62], used o
p ojec s ipe lines o s epping, o 3 Digi al Ligh P ocessing p ojec o s [32], used o p ojec di e en
checke ed pa e ns on he loo o he walking pa h, and compa e he mo o e ec s in PD pa ien s o his
ype o isual cue wi h he same non-ele onic cueing s a egy; and ligh emi ing diodes (2 s udies)
p o iding a e ical a ay o ligh s [72] o a ached o pa ien s’ glasses [74], o gi e wa ning cues o gai
ini ia ion o hy hmic lashes o ligh o s ep hy hmici y du ing walking. O he me hods such as a
Gamepad [61] o an iPad [42] we e also used.
I is impo an o no e ha only 8 s udies used closed-loop isual cueing s a egies [33], [34],
[37]–[39], [42], [45], [61], 3 o which compa e open and closed-loop isual cueing modali ies [34], [38],
[45].
22
Ba hel
e al.
[33], uses lase shoes as he cueing mechanism, o p o ide lase lines in on o
pa ien s du ing gai . The lase beam is ac i a ed by a swi ch loca ed unde he sole o he con ala e al
oo , he e o e ac i a ing he lase du ing heel s ike o he opposi e oo , guiding he li ed oo whe e o
s ep. When he oo is li ed lase is deac i a ed while simul aneously he opposi e oo , now on he g ound,
ac i a es he opposi e lase beam. This cycle epea s i sel du ing he gai cycle, p o iding cues only
du ing ce ain gai -e en s (heel-s ike), he e o e using closed-loop mode. Ra anasu i anon e al. [39],
simila ly p o ided lase lines au oma ically a closed loop, bu ins ead o using lase shoes i used an
ankle b acele wi h an ine ial senso o de ec pa ien s walking pa e n and p o ide he lase line ollowing
each s ep.
Ca pinella
e al.
[61] used a Gamepad, ha combined bo h audi o y and isual cueing, o gi e
bio eedback du ing di e en mo o asks o pa ien s o co ec hei mo emen s. The Gamepad in eg a ed
an LCD sc een ha allowed pa ien s o see ba s illing o ca oons mi o ing hei mo emen s, and when
doing an inco ec mo emen (which means a mo emen bellow o abo e a ce ain h eshold) he sc een
ga e a ed wa ning, a wa ning one, o bo h, o indica e pa ien s o epea he mo emen co ec ly. Fig.
2 shows examples o he exe cises included in he Gamepad.
Fig. 2. Examples o Gamepad isual cueing exe cises [61].
Simila ly, Byl
e al.
[42] uses an iPad o display oo p essing indica o s in g een, yellow o ed,
du ing di e en mobili y asks, o le pa ien s awa e o hei gai pe o mance and co ec hei
mo emen s.
Las ly, as o he 3 s udies ha use bo h open and closed loop cueing, Tang
e al.
[34] compa es
lase lines p o ided con inuously, i.e. ixed p ojec ion on he g ound du ing walking (open-loop), wi h
p o ided hy hmically, i.e. pe sonalizing he equency o lase lines p ojec ion acco ding o pa ien s s ide
cadence (closed-loop). Cubo
e al.
[38] and Kompoli i
e al.
[45] bo h used simila me hods by compa ing
wheeled walke and lase cane, espec i ely, p ojec ion o lase lines con inuously du ing gai (open-loop)
wi h p o ided only when pa ien s choose o, ia swi ches on he handles ha allowed hem o ac i a e
and deac i a e he cues (closed-loop). Bun ing-Pe y
e al.
[37], al hough no compa ing wi h open-loop
cue deli e y, i used he same handle swi ch mechanism o p o ide closed-loop lase lines du ing gai .
23
SOMATOSENSORY CUEING
As o he soma osenso y cueing de ices, he e can be essen ially wo ypes o de ices: ib a o y
senso s o elec os imula ion senso s. The e o e, 10 o he e iewed s udies implemen ed ib a o y
sys ems, mo e speci ically op ed o p o ide closed-loop cueing using sys ems such as Vib oGai [56],
[69], [71], and ib a o y socks [70]; o op ed o open-loop cue deli e y using de ices such as a
me onome in a ib a o y mode [73], Equis asi [58], o o he open-loop ib a o y mo o s and ansduce s
[59], [72], [74], [75]. Two s udies op ed o elec os imula ion, using a wi elessly p og ammable elec o-
s imula o (Phenix Neo Usb) [68] o deli e closed-loop cueing, o a cus om-buil elec ical s imula o
named cueS im [57] p o iding open-loop s imula ion. As o he loca ions o posi ion he de ices o p o ide
soma osenso y cues we e mainly applied a he w is s (4 s udies) [56], [69], [71], [74]; wais (2 s udies)
[73], [75]; legs (2 s udies) [57], [58]; ee (2 s udies) [68], [70]; ankles (2 s udies) [59], [72]; and neck
(1 s udy) [58].
Vib oGai is a ib o ac ile cueing sys em, ha in eg a es an ine ial mo emen senso (Opal IMUs),
a con olle uni , and a ac o uni , as shown in Fig. 3 a) [76]. Those componen s allow Vib oGai o
gene a e ib a ion on he use when de ec ing s ance phase, which is a phase ha usually can igge
FoG episodes. The ac ua o pa s o he de ice consis only in he ac o uni since i is he pa esponsible
o es ablishing he ib a o y cue. This uni is placed on he w is s, whe e he ib a ion is applied [56],
[69], ollowing he se up shown in Fig. 3 b). The ac o uni y, o op imal senso y cueing ib a es a 200-
300 Hz ange, simila ly o he ib a ion o a cell phone in ib a ion mode [69].
Fig. 3. Vib oGai Sys em cha ac e is ics [26].
Peppe
e al.
[58] used a di e en ib a o y app oach, wi h a wea able p op iocep i e de ice based
on ocal mechanical ib a ion named Equis asi. Equis asi is made o nano echnological ibe s sensi i e o
empe a u e, able o ans o m body he mal ene gy in o mechanical ene gy as a o m o high equency
ib a ion. The Equis asi pa ch was placed in he C7, as shown in Fig. 4, and in he gas ocnemius,
bila e ally, and as soon as i con ac ed wi h pa ien s skin i s a ed ib a ing. The e o e, Equis asi p o ided
open-loop cueing.
a) Componen s
b) Expe imen al se up
24
Fig. 4. Ve eb al placemen o Equis asi aken om [77].
Only wo s udies used elec ical s imula ion as soma osenso y cueing s a egy. Sijobe
e al.
[68]
used a wi elessly p og ammable ele o-s imula o (Phenix Neo Usb, Mon pellie , F ance). This s imula o
was placed a ound he shank, and wo ele odes, an anode and a ca hode, we e placed on he do sum
and a ch o he oo , espec i ely, whe e he s imula ion was applied. The s imula ion consis ed o a
closed-loop cueing pa e n o i e 500 ms/phase pulses a 200 Hz, epea ed ou imes a 10 Hz. An IMU
(HikoB) placed unde he la e al malleolus on he sagi al plane, was used o de ec spa io empo al
ou comes and moni o gai e en s o p o ide he s imula ion. When de ec ing heel-o , he s imula ion was
igge ed and a swing phase s opped [68].
As o cueS im, i is a wea able wo-channel elec ical s imula o de ice, wi eless con olled, ha
allows o moni o gai and FoG episodes in PD pa ien s by using closed-loop elec o s imula o y cues [57],
[78]. The de ice is used as a wais bel , and in eg a es skin su ace elec odes, placed in he hams ings
o quad iceps muscles o he body side mos a ec ed by he disease, as shown in Fig. 5 a) [57], [78].
The elec odes a e esponsible o inducing he elec os imula ion. By con olling he ol age, a con inuous
se ies o elec ic bu s s ac on he muscles, deli e ing soma osenso y cues o mo o ehabili a ion
adjus ed o each pa ien gai [57]. Each bu s was deli e ed wi h 100ms o amp-up ime, hen 500ms
ON s imula ion, ollowed by ano he 100ms o amp-down ime [57]. The e was no in e al ollowing
each bu s , he e o e OFF ime was o 0ms [57]. Fig. 5 b) ep esen s he cueing deli e y s a egies du ing
gai [57], [78].
Fig. 5. cueS im Sys em cha a he is ics [22], [27].
a) Expe imen al se up
b) Cueing deli e y s a egy du ing gai
31
S udies
(yea )
Sample Size
(PD, HC)
T aining Dosing
Mo o ask
E alua ion Time
poin s
Medica ion
S a e
Ou comes
On- a ge cueing e ec s
O - a ge cueing e ec s
Mancini
e al.
(2018)[69]
43(43*,0)
*25 PD+FoG;
18 PD-FoG
3 hou s assessmen ( epea ed 3
blocks o he same ask unde 3
di e en condi ions)
Tu ning
While pe o ming he
di e en asks, wi h a
10-15min b eak h ough
each ask
O
Be e u ning smoo hness;
Numbe o FoG episodes.
u ning eloci y and
numbe o u ns.
Spildoo en e al.
(2017)[47]
29 (15*,14)
*15 PD+FoG
6 ials pe condi ion (3 u ning
le and 3 igh )
Du ing mo o ask
O
Numbe o FoG episodes;
Imp o ed head o a ion
and head-pel is
sepa a ion.
Axial mo emen de ici s
did no imp o e.
Das
e al.
(2022)[26]
63 (43*, 20)
*22 PD+FoG;
21 PD-FoG
2h session
Du ing mo o ask
On
u ning eloci y (excep
PD+FOG).
-
Tang e al.
(2017)[34]
23 (23*,0)
*23 PD+FoG
Few ials wi h 2-day in e al
be ween each condi ion
Du ing mo o ask
On
Rhy mic lase cue:
s ep ime, u ning ime
and gai a hy hmici y;
Numbe o FoG episodes.
-
Zhao e al.
(2016)[60]
12 (12,0)
Each condi ion es ed wice o
each u ning ask
Du ing mo o ask
O
Mo e s able gai pa e n;
Me onome was mo e
e ec i e and mos
p e e ed;
Posi i e usabili y eedback
o Google Glass.
-
Willems e al.
(2007)[49]
28 (19*,9)
*9 PD+FoG;
10 PD-FoG
3 ials pe condi ion
Du ing mo o ask
On
-
Non-signi ican e ec s
(simila esul s o
noncued)
Nieuwboe e al.
(2009)[74]
133 (133*,0)
*68 PD+FoG;
65 PD-FoG
8 ials pe condi ion
Du ing mo o ask
On
Audi o y cueing was mo e
e ec i e;
Imp o ed u n imes.
-
Sil a-Ba is a e
al. (2022)[71]
13 (13,0)
2 s MRI scans consis ing o 10
minu es
Du ing mo o ask
O
FoG Index imp o emen s;
Imp o ed u ning wi h bo h
open and closed-loop
cueing.
-
Holmes e al.
(2015)[35]
6 (6*,0)
*6 PD+FoG
45-min session wi h 3 ials pe
condi ion
TUG
Du ing mo o ask
On
Imp o ed walke
posi ioning.
-
S ua
e al.
(2021)[27]
63 (43*,20)
*22 PD+FoG;
21 PD-FoG
2h session
Gai
ini ia ion
and Walking
Du ing mo o ask
On
gai eloci y and s ide
leng h;
powe spec al densi ies
o e senso y egions.
-
Sijobe e al.
(2017)[68]
13 (13*,0)
*9 PD+FoG;
4 PD-FoG
5 ials pe condi ion (3 las ials
we e analysed)
Du ing mo o asks
NM
ime o comple e ask;
Numbe o FoG episodes.
-
B yan e al.
(2010)[36]
7 (7*,0)
*7 PD+FoG
Res pe iods be ween asks
Walking and
Tu ning
Du ing mo o ask
On and O
G een ligh :
Numbe o FoG episodes;
-

32
S udies
(yea )
Sample Size
(PD, HC)
T aining Dosing
Mo o ask
E alua ion Time
poin s
Medica ion
S a e
Ou comes
On- a ge cueing e ec s
O - a ge cueing e ec s
Tu n ime and numbe o
s eps.
Ra anasu i anon
e al.
(2023)[39]
4 (4*,0)
*4 PD+FoG
2 sessions o TUG and walking;
10 min washou pe iod be ween
walking wi h lase on and o
Walking and
TUG
Baseline;
Tes ed again 1 week
a e
On
s ide leng h and eloci y;
TUG es ime;
% FoG ime.
-
Kla e e al.
(2023)[70]
31 (31*,0)
*31 PD+FoG
2 sessions on 2 sepa a e days (1
On and o he O medica ion)
Walking
(passing
doo way)
and Tu ning
Du ing mo o ask
On and O
>10% imp o emen in
FOG;
Use sa is ac ion.
Closed-loop and open-loop
ac ile, and audi o y cueing
did no signi ican ly impac
he % ime and numbe o
FoG episodes, o
spa io empo al pa ame e s
Lee e al.
(2023)[63]
10 (10*,0)
*10 PD+FoG
3 ials pe ask
Du ing mo o ask
On
“Walk wi h me” p og am:
Be e s aigh walk, dual
ask, and walk h ough
doo way.
“Un eeze me” p og am:
Only dual ask imp o ed.
UPDRS III did no had
signi ican imp o emen s;
“Walk wi h me” p og am:
Wo se 180 deg ee u ning.
An e al.
(2023)[40]
10 (10*,0)
*10 PD+FoG
10 min amilia iza ion; 3 ials
pe condi ion; 10 min b eak in
be ween asks
Gai
ini ia ion;
Walking and
Tu ning
Du ing mo o ask
On
Numbe , du a ion and
p opo ion o FoG;
G ea e up ake o FP-CIT.
-
Longi udinal
Li ani-Sil a e al.
(2017)[55][50]
19 (19**,0)
**10 ex u ed
insole;
9 con en ional
insole
5 ials wi hou insoles
Walking
P e- es ;
A e 1 week wea ing
he g oupspeci ic
insoles (Pos - es ); A e
1 week wea ing
con en ional insoles
( ollow-up).
On
Pos - es ex u ed insoles:
s ide leng h and plan a
sensa ion;
Follow-up:
Only plan a sensa ion
imp o emen s main ained.
-
El-Tamawy e al.
(2012)[59]
30 (30,0)
45-min session o low in ensi y
physio he apy p og am, 3 imes
a week, o 8 weeks; 5 min
eadmill aining a baseline and
25 min a he end o ea men
Be o e and a e
ea men
On
Imp o ed lowe limb join s’
angles;
Cadence and s ide
leng h.
-
Ginis e al.
(2016)[51]
40 (40**,0)
**22 CuPiD;
18 CG
walk a leas 3 imes pe week o
30 min, du ing 6 weeks
FoG: addi ional 30 min, 3 imes a
week using he FOG-cue app
baseline;
pos - es (a e 6
weeks);
ollow-up ( ollowing 4
weeks)
On
Pos - es :
MBEST sco e;
Follow-up:
Main ained SF-36 (while
con ols dec eased).
MBEST no main ained a
ollow-up.
33
PD: Pa kinson’s Disease; HC: heal hy con ols; PD+FoG: PD pa ien s wi h FoG; PD-FoG: PD pa ien s wi hou FoG; EG: expe imen al g oup; CG: Con ol g oup; ML APAs: Medio La e al An e o Pos e ial Adjus men s;
MDS-UPDRS: Mo emen Diso de Socie y – Uni ied Pa kinsons Disease; MoCA: Mon eal Cogni i e Assessmen ; % ime FoG: pe cen ime spen ozen; TUG: Timed Up an Go; EMG: ele omiog aphy; CoM:
Cen e o Mass; MBEST: Mini Balance E alua ion Sys ems es ; CoP: Cen e o P essu e; FOG-Q: F eezing o Gai ques ionnai e
S udies
(yea )
Sample Size
(PD, HC)
T aining Dosing
Mo o ask
E alua ion Time
poin s
Medica ion
S a e
Ou comes
On- a ge cueing e ec s
O - a ge cueing e ec s
Peppe e al.
(2019)[58]
20 (20,0)
8 week aining
Mo o
Balance
be o e and a e i s 8
weeks; be o e and a e
ollowing 8 weeks
On
S ance and double
suppo ime;
Veloci y and s ide leng h;
Se e e pa ien s ( H&Y)
had be e imp o emen s
in MDS-UPDRS III;
Placebo showed some
imp o emen s in MDS-
UPDRS III.
-
Ca pinella e al.
(2017)[61]
42 (42**,0)
**22 EG;
20 CG
20 sessions o 45 min each, 3
pe week
baseline; pos - aining;
1 mon h ollow-up
On
Pos - aining:
Be e balance
pe o mance;
CoP ML.
A 1-mon h imp o emen s
we e no main ained.
Capa o e al.
(2020)[53]
154 (154*,0)
56 RAS;
50 wi hou RAS;
48 educa ional
p og am
10 sessions o 45 min; 2
sessions pe week, o e 5 weeks
baseline (14 days p io
o
aining);
one day a e he las
5 h week aining;
1-mon h ollow-up;
6-mon hs ollow-up
On
MBEST sco e;
Re ained imp o emen s a
1 and 6-mon h ollow-ups;
Imp o ed UPDRS.
-
Chang e al.
(2019)[54]
21 (21*,0)
*10 PD+FoG;
11 PD-FoG
10 ounds o 50-s ep SIP
mo emen s
SIP
baseline and cued;
(washou in e al o a
leas 1 week be ween
he 2 condi ions)
O
Cadence;
S ep a iabili y;
Co ical Silen Pe iod;
Imp o ed Sho In e al
In aco ical Inhibi ion
-
Chomiak e al.
(2017)[50]
11 (11**,0)
**6 music;
5 podcas
3 imes pe week; 10-20 min
aining sessions
Be o e and a e 4-week
SIP a -home aining
On
Imp o emen on dual- ask
s ep au oma ici y wi h
music bu no podcas
No signi ican e ec o
aining on FES-I, MoCA, o
FOG-Q
Byl e al.
(2015)[42]
24 (12**,0)
**7 EG;
5 CG
6–8 weeks, 12 isi s (90 min
each) o al o 18h o aining
Mobili y,
ange o
mo ion,
balance,
and s eng h
exe cises
baseline;
6-weeks pos - aining
NM
Numbe o alls;
S ep leng h;
Imp o ed s eng h, Be g
balance and ange o
mo ion.
Non signi ican
imp o emen s in eloci y
and TUG ime.
34
STUDY TYPES
Pape s we e classi ied as 2 di e en ypes: c oss-sec ional s udies, when e alua ing only single
session sho - e m e ec s o cueing (41 s udies) [26]–[41], [43]–[49], [52], [56], [57], [60], [62]–[75];
o longi udinal, when pa ien s ained wi h cueing s a egies o a con inuous pe iod, and mul isession
da a analysis we e pe o med in o de o e alua e he long- e m usage e ec s o cueing (9 s udies) [42],
[50], [51], [53]–[55], [58], [59], [61].
Ra anasu i anon
e al.
[39], al hough being longi udinal, i does no e alua e he long- e m use
o cueing s a egies in mo o ehabili a ion, i only di ides da a collec ion du ing 2 sessions 1-week apa .
The e o e, his pape ou comes we e analysed as a c oss-sec ional s udy.
PD PATIENTS AND CONTROL GROUPS
The o al numbe o PD pa ien s summa ized in his e iew, he e o e he sum o all a icles PD
pa ien s’ samples, is 1543, being 43 he median sample size. O he 50 a icles analyzed, 18 s udies
[26]–[28], [30], [31], [43]–[46], [48], [49], [54], [56], [66]–[69], [72], [74] sepa a ed PD pa ien s in
bo h eeze s (wi h FoG), and non- eeze s (wi hou FoG). O he 16 s udies conside ed only eeze s o
he clinical p o ocol [32]–[35], [37]–[41], [45], [47], [62]–[64], [70], [73], and 1 s udy used compa ed
one g oup o eeze s wi h heal hy con ols [47]. The e o e, his esul ed in a o al o summing a o al o
642 eeze s and 378 non- eeze s, wi h co esponding median sample sizes 20 and 20, espec i ely,
and lea ing 523 PD pa ien s indisc imina e ega ding su e ing o no o FoG. The maximum sample size
o PD pa ien s was 154, ound in Capa o
e al.
[46]
and he minimum was 4, ound in Ra anasu i anon
e al.
[57].
The main ac o s e iewed o cha ac e ize PD pa ien s we e age (48 s udies) [26]–[49], [52]–
[75]; Mo emen Diso de Socie y-Uni ied Pa kinson’s Disease Ra ing Scale (MDS-UPDRS) (43 s udies)
[26]–[49], [52]–[75]; disease du a ion [26]–[49], [52]–[75]; and Hoehn and Yah scale (H&Y) [26]–[49],
[52]–[75] (bo h in 40 s udies); gende (34 s udies) [26]–[49], [52]–[75]; and New F eezing o Gai
Ques ionnai e (NFOG-Q) o o he FoG ques ionnai es (35 s udies) [26]–[29], [31]–[37], [39]–[43], [46],
[49]–[54], [56], [57], [60], [62]–[64], [67], [68], [70], [71], [73], [74].
Less han 50% o he s udies cha ac e ized pa ien s e alua ing o he clinical scales such as and
Mon eal Cogni i e Assessmen scale (MoCA) (15 s udies) [26]–[28], [30]–[32], [41], [43], [50], [52],
[56], [63], [68], [69], [73]; Mini-Men al S a e Examina ion (MMSE) (16 s udies) [37], [38], [40], [43],
[46]–[49], [54], [55], [57]–[59], [62], [64], [70]; Falls E icacy Scale (FES–I) (9 s udies) [26], [27], [29],
[32], [46], [50], [53], [68], [74]; Pa kinson’s Disease Ques ionnai e (PDQ-39) (2 s udies)[33], [43]; and
Mini Balance E alua ion Sys ems es (MBEST) (2 s udies) [52], [53]. Fewe s udies also e alua ed less
35
common scales such as Ge ia ic Dep ession Scale – 15 (GDS-15), Pos u al Ins abili y and Gai Diso de
(PIGD) Sco e, F on al Assessmen Ba e y (FAB) To al Sco e, Beck’s Dep ession In en o y (BDI) Sco e,
Royal’s Clock D awing ask (CLOX 1&2), and S a e-T ai Anxie y In en o y (STAI) Sco es.
Besides clinical scales, PD pa ien s we e cha ac e ized acco ding o Le odopa Equi alen Daily
Dosage (LEDD) (16 s udies) [26], [27], [29], [32], [35], [36], [44], [46], [52], [60], [62]–[64], [66], [67],
[73]; heigh (14 s udies) [26]–[28], [34], [36], [43], [44], [49], [55], [59], [64]–[66], [75]; and weigh
(10 s udies) [26]–[28], [34], [36], [49], [55], [59], [65], [75]. In some pape s (4 s udies) [63] usabili y
ques ionnai es we e also pe o med o analyze pa ien s eedback abou he cueing de ice.
I is impo an o no e ha hese cha ac e is ics e alua ions we e no conside ed as clinical
pa ame e s analysed, since only p o ide in o ma ion o cha ac e ize he popula ion being s udied.
As o medica ion s a e, he majo i y o he s udies we e assessed on medica ion (31 s udies)
[26]–[28], [30]–[32], [34], [35], [37]–[41], [43]–[46], [49]–[51], [53], [55], [57]–[59], [61]–[63], [65],
[66], [74]. Al e na i ely, 12 s udies [54], [71], [75], op ed o assessing p o ocols du ing o medica ion,
and 5 pape s [33], [36], [48], [52], [70], es ed bo h on and o s a es, o exclude bias o dopamine gic
medica ion in ake. Byl
e al.
[42] and Sijobe
e al.
[68] did no men ion in which medica ion s a e
pa ien s we e assessed.
Only 11 s udies used heal hy con ol g oups o compa e esul s wi h he PD popula ions being
sampled [26]–[29], [31], [44], [46]–[49], [67]. This esul ed in a o al o 203 heal hy con ols, being he
median sample size o heal hy con ol g oups 18. The maximum sample size o con ols was ound in
Del al
e al.
[48],
and he minimum, besides no ha ing con ol g oup, was ound in Willems
e al.
[49],
wi h 30 and 9 heal hy pa icipan s, espec i ely.
I is impo an o no ice ha 6 s udies used only one g oup o PD pa ien s o pe o m di e en
cued and uncued asks, in andom o de , o compa e he cueing s. non-cueing e ec s o e mobili y [52],
[53], [57], [60], [65], [71], [75]; 2 s udies dis ibu ed PD pa ien s in wo al e na i e g oups being one
cued and o he Placebo, which mean pa ien s used a simila de ice bu wi hou p o iding any ype o
cue [55], [58]; 4 o he pape s di ided pa ien s in expe imen al g oups, ecei ing cueing, and con ol
g oups, ha did no ain wi h he cueing de ices [42], [59], [61]; and 2 s udies op ed o o he al e na i e
sample g oups o ganiza ions [50], [53].
Chomiak
e al.
[50] and Capa o
e al.
[53] di e en g oup assignmen s a e wo h explaining
sepa a ely. Chomiak e al. [50] di ided pa ien s in 2 g oups, bo h ecei ing audi o y s imuli in di e en
manne s: one g oup ecei ed music (5 pa ien s) and o he ecei ed s imuli in a o m o a podcas (6
pa ien s). These 2 g oups we e analysed o unde s anding he mo o and FoG e ec s o a mo e hy hmic
36
induced audi o y cue (music) when compa ed o an audi o y cue ha equi es almos dual- ask ocus
(podcas ). Al e na i elly, Capa o
e al.
[68] di ided PD pa ien s in 3 g oups: one g oup ha ecei ed
Rhy mic Audi o y S imuli (RAS)-suppo ed mul imodal balance aining (56 pa ien s), one g oup ha
pe o med he same aining bu wi hou cueing (50 pa ien s) and a hi d con ol g oup ha ins ead o
pe o ming he aining, ecei ed an educa ional p og am explaining hem he RAS aining e ec s in
add essing PD mo o symp oms. These hi d g oup was assigned o compa e he aining and cueing
e ec s wi h a g oup ha al hough no being cue ained, is ale ed o he in en ions behind cue aining.
I is [53]also impo an o no e ha Byl
e al.
[42], besides e alua ing PD pa ien s also [42]e alua ed
pa ien s ha su e ed om s okes, bu o he pu pose o his esea ch ha g oup o pa ien s was
excluded.
To summa ize he di e en combina ions and o ganiza ions o sample g oups e iewed he e
we e iden i ied 9 s udies ha only di ided eeze s s. non- eeze s [30], [43], [54], [56], [66], [68], [69],
[72], [74]; 9 s udies ha di ided eeze s, non- eeze s and heal hy con ols [26]–[28], [31], [44], [46],
[48], [49], [67]; a single s udy ha compa ed a eeze g oup wi h heal hy con ols [47]; a single s udy
ha compa ed one PD g oup and one heal hy con ol g oup [29]; 16 s udies ha only used one g oup o
eeze s [32]–[35], [37]–[41], [45], [47], [62]–[64], [70], [73]; 6 s udies ha only used one g oup o PD
pa ien s [52], [57], [60], [65], [71], [75]; 2 s udies ha analyzed one Placebo g oup s. one expe imen al
g oup [55], [58]; 4 s udies designing one expe imen al s. one con ol g oup o PD pa ien s [42], [51],
[59], [61]; and 2 s udies op ing o wo di e en cued condi ions [50], [53]. Ul ima ely, he mean numbe
o expe imen al g oup pa icipan s was 19, 10 o Placebo g oup pa icipan s and 13 o con ol g oup
wi h PD pa ien s. The espec i e minimum and maximum numbe s we e 7 and 22 o expe imen al g oup
sizes, 10 and 10 o Placebo g oup sizes; and 5 and 20 o con ol wi h PD pa ien s g oup sizes.
INCLUSION AND EXCLUSION CRITERIA
Al hough no lis ed in he able, i was also impo an o analyse he mos p e alen inclusion and
exclusion c i e ia conside ed, o u he conside a ion du ing he p o ocol design. I is impo an o no e
ha no all s udies p esen ed hese c i e ia. Fou s udies did no men ion any exclusion c i e ia, bu
p esen ed inclusion c i e ia [57], [62], [73], [75]; o he ou s udies did no men ion he inclusion c i e ia,
bu p esen ed exclusion c i e ia [30], [44], [67], [72]; and i e s udies did no men ion bo h [36], [38],
[58], [63], [68].
The main and inhe en ly p esen inclusion c i e ia conside ed was pa icipan s being diagnosed
wi h PD, we e 10 s udies inclusi ely men ioned pe o ming he diagnosis acco ding o he UK b ain bank
c i e ia [26], [27], [31], [32], [40], [41], [53], [60], [64], [65]. The g ea majo i y o he s udies only

37
included pa ien s ha we e able o walk o s and independen ly (24 s udies) [26], [27], [31], [34], [35],
[37], [39], [42], [43], [46], [47], [49]–[51], [53]–[55], [59]–[62], [65], [66], [73], o g an ha pa ien s
we e able o pe o m he mo o asks; in eg a ed a ce ain ange o s ages o H&Y scale (21 s udies) [26]–
[28], [31], [32], [34], [42], [47], [49], [51], [53]–[57], [61], [65], [69], [71], [74], [75], being he mos
p e alen including pa ien s ha anged om s age 2 o 3 o pa ien s anging om s age 1 o 4; we e
e alua ed wi h his o y o p esence o FoG (19 s udies) [32]–[35], [37], [39]–[41], [45], [47], [48], [57],
[59], [60], [62], [64], [65], [70], [73], using o ha pu pose NFOG-Q o UPDRS-III; we e able o pe cei e
isual, audi o y o ib a o y s imuli (12 s udies) [26], [27], [31], [35], [37], [46], [53], [54], [62], [70],
[73], [75], mainly using Snellen cha o e alua e isual acui y; ha ing a ce ain age (12 s udies) [26]–
[28], [31], [32], [35], [41], [46], [54], [64], [74], [75], being mos ly common o conside pa ien s ha
ha e a leas 50 yea s; and absence o cogni i e impai men o no being demen ed (11 s udies) [29],
[32], [37], [46], [47], [51], [53], [54], [57], [62], [74], equen ly e alua ed conside ing MoCA sco e o
MMSE sco e la ge han 24.
As o exclusion c i e ia, mos o hem comple e he inclusion c i e ia by ein o cing de e mined
aspec s. Pa icipan s mos p edominan ly we e excluded i p esen ed o he neu odegene a i e o heal h
diso de s (34 s udies) [26]–[35], [39]–[44], [46], [48]–[52], [54], [56], [59], [60], [64]–[67], [69]–[72],
such as Alzheime ’s disease, s oke o gai diso de s; i p esen ed demen ia o se e e cogni i e
impai men making hem unable o ollow ins uc ions (25 s udies) [26]–[31], [37], [39], [41], [43], [44],
[48]–[50], [54], [55], [59]–[61], [64], [65], [67], [69], [70], [72]; isual, hea ing o soma osenso y
pe cep ion impai men s (22 s udies) [30], [32]–[35], [37], [39], [43], [44], [46], [50]–[52], [55], [59],
[60], [64]–[66], [69], [70], [72]; and i pa ien s we e unable o walk o walk independen ly (10 s udies)
[26]–[28], [41], [45], [52], [56], [64], [69], [71]. I is also wo h men ioning ha 5 s udies also conside ed
o exclude pa ien s ha had DBS o o he unc ional b ain su ge y, o p e en isk o bias [46], [47], [49],
[64], [74].
MOTOR TASKS
To e alua e pa ien s’ mo o pe o mance using di e en cueing me hodologies, i was essen ial
o analyse di e en mo o asks. Essen ially, mo o asks can be di ided in 3 ca ego ies: walking asks
(34 s udies) [27]–[33], [36]–[46], [51]–[53], [55], [57]–[59], [61]–[65], [67], [68], [70], [75]; u ning
asks (13 s udies) [26], [34], [36], [38], [40], [47], [49], [60], [63], [69]–[71], [74], such as 8-shaped
u ning ask, 180º le U- u n, u ning in place and 360º u ns; [27]–[33], [36]–[46], [51]–[53], [55],
[57]–[59], [61]–[65], [67], [68], [70], [75] gai ini ia ion asks (10 s udies) [27], [28], [37], [40], [48],
38
[56], [66], [68], [72], [73], such as i s s ep ini ia ion, si ing- o-s anding o s epping-in-place (SIP), and
hey can be pe o med indi idually o combined.
Di e en walking asks we e pe o med such as passing h ough a doo way (7 s udies) [30], [37],
[38], [44], [45], [63], [70], wi h he pu pose o exposing pa ien s o po en ial FoG igge ing scena ios;
ange o mo ion asks (1 s udy) [42] and balance aining asks (4 s udies) [42], [53], [58], [61], ha
consis in walking o s anding in di e en su aces, o du ing condi ions such as eyes closed, o analyse
pos u al and mo o con ol aspec s. Timed Up and Go (TUG) ask is an example o combining bo h gai
ini ia ion, walking and u ning asks. TUG ask consis s o e alua ing gai pe o mance when s anding up
om chai , walking, u n a ound, walk back and si -down. In his esea ch 7 s udies pe o med TUG asks
[35], [39], [44], [52], [53], [61], [68], being Capa o
e al
. [53] he only s udy ha pe o med TUG dual-
ask.
The e o e, i is also impo an o highligh dual- ask aining (6 s udies) [31], [35], [39], [44],
[51]–[53], [56], [61], [63], [68]. Dual- ask aining consis ed o simul aneously pe o ming gai and
a en ional asks. Th ee s udies pe o med dual- ask aining by coun ing backwa ds by 7 [63] o by 3
[56], while walking (2 s udies) [56], [63] o pe o ming TUG dual- ask es [53]. Simila ly, G aham e al.
[31] pe o med a o wa d digi span es while walking. As o Sijobe
e al.
[68] pe o med SIP dual- ask
es , by naming backwa ds he mon hs o he yea while s epping in place. Las ly, Ginis
e al.
[51]
pe o med walking while eci ing as many wo ds om a le e as possible.
OUTCOMES
All 5 di e en pa ame e s (Spa io empo al, Kine ic, Kinema ic, Physiological o Clinical) we e
analysed o e alua e cueing e ec s o e PD pa ien s gai pe o mance and in mi iga ing mo o symp oms,
such as FoG. Di e en esul ou comes we e ob ained, meaning posi i e (45 s udies) [26]–[36], [39]–
[44], [46]–[48], [50]–[55], [57]–[75], and nega i e (24 s udies) [28], [30], [37], [38], [41], [42], [44],
[45], [47]–[51], [56], [61]–[64], [66], [67], [69], [70], [72], [73], cueing e ec s o e mo o esponse
we e analysed.
Spa io empo al pa ame e s we e he pa ame e s ha showed g ea e imp o emen s wi h cueing
(31 s udies) [26]–[29], [31], [32], [34], [36], [39], [42]–[44], [46], [48], [52], [54], [55], [57]–[60], [62],
[64]–[68], [72]–[75], mainly inc easing pa ien s’ eloci y while pe o ming gai asks [26]–[28], [32],
[39], [46], [52], [58], [65], [66], [75], and cadence [44], [46], [54], [59], [67], he e o e educing ime
o comple e asks [34], [36], [39], [57], [68], [74]; inc eased s ep leng h [32], [42], [43], [62], [66], [73],
and s ide leng h [27], [28], [31], [39], [44], [46], [55], [58], [59], [64], [65], [67], [75]; dec eased s ep
39
o s ide ime [31], [34], [46], [48], [62] , and a iabili y [31], [52], [54]; allowing be e ask pe o mance
and mo e s able/smoo he gai pa e ns [60], [69]. I is wo h no ing ha in Mancini
e al.
[69] educing
he numbe and eloci y o u ns was no conside ed a eal nega i e aspec , since i was hypo hesized
ha his dec ease was due o pa ien s being mo e awa e and ocused du ing he u ning asks wi h cueing.
Pa ien s’ smoo hness du ing u ning inc eased which indica ed ha hey we e mo e cau ious du ing
u ning pe o mance, making hem slowe bu in al e na i e p o iding hem an imp o ed gai pa e n and
educing FoG episodes.
As o Clinical pa ame e s, he main imp o ed ou come was he numbe and du a ion o FoG
episodes (14 s udies) [30], [33], [34], [36], [39]–[41], [47], [57], [65], [67]–[69], [73]. Gee se
e al
.
[41] ini ially showed nega i e e ec s as he % ime FoG and numbe o episodes inc eased a i s .
Howe e , i was conside ed a habi ua ion pe iod, since as pa ien s p oceeded wi h aining hose nega i e
e ec s we e o e come.
O he clinical posi i e e ec s o cueing we e educing numbe o alls [42]; imp o ed MBEST
sco e [53], al hough a 4-week ollow-up no main ained [51]; imp o ed UPDRS [53], and be e balance
pe o mance [42], [61].
Tosse ams
e al.
[52] e i ied also a ela ionship be ween some clinical and spa io empo al
esul s, such as lowe MDS-UPDRS III sco es being linked o imp o ed gai a iabili y, meaning pa ien s
wi h less se e e disease symp oms had be e imp o emen s in gai pa e n; and highe MoCA sco es
linked o g ea e e icacy o ex e nal cueing, meaning less cogni i e impai ed pa ien s showed be e
imp o emen s when using cues. Ano he clinical ou comes wo h men ioning a e Peppe
e al.
[58]
conclusions ha showed ha pa ien s wi h highe H&Y sco e, he e o e mo e se e e disease s age,
showed be e imp o emen s in MDS-UPDRS III, meaning hei mo o symp oms and disease pe cep ion
may ha e imp o ed. The Placebo g oup also showed some imp o emen s in MDS-UPDRS III, meaning
pa ien s pe cep ion o he disease may ha e been also imp o ed by he sense o ha ing a ’ alse’ cueing
de ice e en when i was no deli e ing cues.
As o Kinema ic posi i e e ec s, he main imp o emen s we e imp o ed APA’s du a ion [28],
[48], [72] and ampli ude [72]; imp o ed head o a ion and head-pel is sepa a ion du ing u ning [47];
imp o ed o wa d Cen e o Mass (CoM) eloci y and sideways sways [73]; imp o ed lowe limbs join s
and ee s ike heads [31], [59]; and inc eased knee lexion and ankle do si lexion [67]. Howe e , some
s udies showed ha APA’s ampli ude did no inc ease ampli ude in all pa ien s, indica ing ha pa ien s
wi h highe UPDRS III sco es dec eased APA’s ampli ude [28], meaning APA’s imp o emen s may no be
p onounced o all pa ien s.
40
In o he hand, kine ic ou comes ga e in o ma ion abou p essu e and o ces e ec ing du ing
di e en mobili y asks. The main imp o emen s obse ed we e imp o ed posi ioning [35] and imp o ed
Cen e o P essu e (CoP) sways in he medio la e al plane [61]. Las ly, physiological ou comes mainly
se ed as in o ma ion o ela e wi h o he pa ame e s. The main physiological esul s ob ained we e
imp o ed saccade, he e o e imp o ed apid mo emen o he eyes be ween ixa ion poin s, using isual
cues [31], meaning pa ien s we e mo e ocused on p epa ing hei mo emen when p o ided isual
eedback; and g ea e up ake o FF-CT in he be e cueing esponse g oup [40], which may indica e a
po en ial connec ion be ween he compensa o y e ec s o isual cueing and he dopamine gic unc ion
in speci ic b ain egions in PD pa ien s.
One s udy indica ed imp o emen s o Co ical Silen Pe iod (CSP) and Sho In e al In aco ical
Inhibi ion (SICI), meaning ha co ical inhibi ion may ha e been educed by audi o y cues, leading o
be e gai pe o mance [54]. As o EMG esul s, Schlens ed
e al.
[56] men ioned ha , al hough i s
s ep dec easing when in dual- ask pe o mance, con ac ion o he enso asciae la ae inc eased,
he e o e i was no iced an imp o emen o muscle ac i i y using cueing.
2.3. DISCUSSION
2.3.1. RQ1: HOW HAVE CUEING STRATEGIES BEEN USED AND APPLIED FOR TARGETING
MOTOR SYMPTOMS OF PD?
Cueing s a egies ha e been employed o a ge PD mo o symp oms by p o iding ex e nal s imuli
o imp o e mo emen and gai . They can be di ided essen ially in 3 ypes acco ding o he senso y aspec
a ge ed: isual, audi o y, o soma osenso y cueing; and i was possible o unde s and hei
implemen a ions sepa a ely o combined. Cues we e deli e ed also ia elec onic de ices, non-ele onic
s a egies, o bo h me hods combined. The g ea majo i y o he s udies implemen ed Visual o Audi o y
cueing, essen ially and espec i ely ia lase de ices o p ojec lase lines o pa ien s o s ep o
me onomes o p o ide hy hm o pa ien s o ollow, while Soma osenso y cueing was p esen in less
han a hi d o he e iewed pape s. O his hi d, mo e han hal combined Soma osenso y cues wi h
Visual and/o Audi o y s imuli, which means ha he e a e a lack o esea ch e alua ing Soma osenso y
cues alone in add essing mo o symp oms such as FoG. The main soma osenso y de ices used we e
ib a o y sys ems, such as Vib oGai , o p o ide ib a o y cues a a ce ain hy hm o when de ec ing a
de e mined gai -e en o phase.
47
3. SOLUTION OVERVIEW
This disse a ion is he p og ess o se e al clinical s udies ha aim o analyse and compa e
di e en cueing s a egies, o conclude which may cause be e mo o pe o mance imp o emen s in a
c oss-sec ional s udy; de elop a no el s a egy; and p ac ice a long- e m p o ocol, in PD pa ien s, using
+sensBand and e alua ing he mo o e ec s o hose s a egies in gai ehabili a ion, ocusing on FoG and
QoL. To achie e his, i is necessa y o i s conside he ma e ials necessa y o de elop he no el solu ion.
In his chap e , i is ou lined: (i) an o e iew o he sys em ocusing on he main ha dwa e componen s,
(ii) an in oduc ion and desc ip ion abou he di e en s a egies used and (iii) a desc ip ion o he MATLAB
App implemen ed o u he da a p ocessing.
3.1. HARDWARE
In which conce ns he echnologies used in his p ojec , i is impo an o ake in o accoun wo
di e en sys ems: he +sensBand sys em, esponsible o p o iding bo h soma osenso y cueing, isual
cueing and o da a collec ion (Chap e 4, 5 and 6), and he Xsens mo ion acking sys em wi h Awinda,
used o gai assessmen alida ion in Chap e 6. Fig. 7 and Table 3 summa ize he echnologies used.
Table 3. Technologies used du ing his disse a ion
Technologies
Pu pose
Da a collec ed
Pa ame e s
e alua ed
+sensBand
P o ide di e en con ol
s a egies o
soma osenso y and isual
cueing, du ing pa ien s gai
pe o mance, and
Collec s accele a ion and
angula eloci y.
Allow o de e mine
gai eloci y,
cadence, s ep
leng h, s ep ime,
s ide leng h, s ide
b) Xsens
a) +sensBand
Fig. 7. Technology sys ems used.

48
simul aneously collec ing
gai me ics.
ime, asymme y,
and a iabili y
Xsens
Valida e +sensBand da a
collec ion and p o ide
addi ional biomechanical
ou comes
Collec s mo ion da a o
each body segmen
posi ion
Fi s , and since i is he main componen o add ess in his p ojec , i is impo an o s a by
unde s anding he +sensBand sys em and he di e en componen s in eg a ed.
The +sensBand de ice i sel includes 6 sys ems: a p ocessing sys em; an ac ua ion
sys em, ha includes ib a ion mo o s esponsible o soma osenso y cueing and 2 lase p ojec o s o
isual cueing; a senso y sys em, including 4 IMU’s esponsible o da a collec ion and a came a o
doo de ec ion; a memo y sys em; a communica ion sys em, esponsible o communica ing ia
Blue oo h wi h a sma phone App, designed o de ice con igu a ion, and ia SSH wi h he clinical desk op
App, designed o da a p ocessing; and a powe supply, ha consis s o a Cha mas powe bank 10000.
All 5 sys ems a e loca ed inside a 3D p in ed box, excep o he senso y sys em, whose IMU’s a e loca ed
h oughou he wais band ex ile i sel , a 4 di e en loca ions a ound he wais : back; le side; igh side
and na el poin . +sensBand allows o moni o axel accele a ions and angula eloci ies on all hese
loca ions, simul aneously o no , o assess gai me ics, while he ac ua ion sys em p o ides isual o
ib o ac ile cues acco ding o he con ol s a egy selec ed.
The Xsens sys em con ains 6 wi eless Mo ion T acke s (MTw), ha consis o minia u ized IMU’s
ha inco po a e 3D accele ome e s, gy oscopes, magne ome e s and a ba ome e , he e o e p o iding
3D o ien a ions as well as linea accele a ion, angula eloci y, magne ic ield and p essu e da a. This
en ails mo e de ailed me ics han he +sensBand senso y sys em, allowing o alida e he +sensBand
collec ed da a (ac ing as a g ound u h) bu also p o iding mo e in o ma ion o assess pa ien s mo ion
ou comes. The IMU’s a e posi ioned on he pa ien s’ body (e.g. lowe and uppe legs, uppe a ms, w is s,
and pel is), using Velc o body s aps. To cha ge and calib a e he MTW’s, an Awinda S a ion is connec ed
o an ou le , ia DC powe connec o , and o a compu e , o using an Awinda USB Dongle, ia USB,
allowing synch oniza ion o da a assessmen when using he di e en MTW’s simul aneously.
3.2. CONTROL AND MONITORING STRATEGIES
The +sensBand o iginally o e ed i e possible con ol s a egies, ha can be selec ed using he
sma phone App, including (i) gai -e en d i en cueing; (ii) con ex awa eness cueing; (iii) u ning
49
phase-dependen cueing; (i ) pos u al s abili y phase-dependen cueing and ( ) con inuous cueing.
Based on he li e a u e e iew conclusions, i was u he de eloped a no el senso y cueing s a egy, he
( i) Fog p e en ion cueing. Fo his p ojec , only s a egies (i), (ii) and ( i) we e explo ed, he e o e
only hose will be u he discussed.
Besides p o iding di e en cueing s a egies, he +sensband also p o ides a mo ion
moni o ing op ion.
All he di e en s a egies can be selec ed o con igu e he +sensband de ice using a sma phone
app, as shown in Fig. 8. In case he use p e ends o no use any ype o cueing, bu a he only moni o
he pa ien ’s gai , i should selec he Mo ion moni o ing op ion di ec ly. When he Senso y cueing op ion
is selec ed ins ead, a page wi h he di e en cueing s a egies a ailable will appea and he s a egy
p e ended should be selec ed and hen con igu ed. Each o he s a egies will be u he explained in his
sec ion.
Fig. 8. +sensBand sma phone App.
3.2.1. MOTION MONITORING
PD mo o symp oms a e commonly p esen du ing pa ien s' e e yday asks. Changes in pa ien s
gai pa e n such as educed ampli ude o mo emen , slowness, and sho e s eps, can be an indica o
o mo e se e e symp oms such as FoG. The e o e i is o eno mous impo ance o analyze pa ien s gai
me ics.
By hese means, he +sensBand was de eloped conside ing also he de elopmen o a mo ion
moni o ing s a egy. This s a egy ac i a es only he IMU’s o he +sensBand, o collec accele a ion
and angula eloci y da a, o u he analysis o gai me ics.
Gai e en de ec ion equi es he Back IMU da a as a e e ence, so i is impo an o gua an ee
ha he IMU is placed co ec ly a he lowe back (app oxima ely abo e L5). The +sensBand in eg a es a
50
gai e en de ec ion algo i hm ha ecei es he IMU da a in eal- ime, hen calib a es i o elimina e
measu emen e o s, applies mo ion compensa ion o adjus o unk o a ions, and il e s da a o de ec
signi ican changes indica i e o gai e en s.Then, he senso y da a uns ough a eal- ime adap i i e
algo y hm, ha allows o iden i y i e key gai e en s: Heel s ike (HS), he ini ial con ac (IC) momen ;
Foo la (FF); Mid s ance (MS ); Heel o (HO); and Toe o o he inal con ac (FC) o he oo . This
algo i hm was de eloped speci ically o PD pa ien s, con aining adap a i e h esholds o each gai e en ,
ha a e upda ed along he mo ion moni o ing assessmen . The adap a i e h esholds a e ecalcula ed
and upda ed pe iodically in se s o 10 gai cycles, conside ing he da a o he p e ious gai cycle se .
The mo ion moni o ing s a egy is also au oma ically ac i a ed when using all he di e en cueing
s a egies, since i equi es o ac i a e mo ion moni o ing o ob ain he pa ien gai da a ile collec ed.
3.2.2. GAIT EVENT-DRIVEN CUEING
Se e al cueing s a egies o +sensBand we e explo ed, o comp ehend each unc ionali y bu also
o selec which should be adap ed and implemen ed in he clinical p o ocols.
Gai e en -d i en cueing consis s in p o iding closed-loop ib a o y s imula ion, igge ed
e e y ime a igh - oo oe-o is de ec ed. Toe-o gai e en was selec ed since li e a u e epo s
obse a ion o inadequa e swing phase be o e he occu ence o FoG episodes [79].
The ib a o y equency and ime on which he cueing s a egies a e p o ided, which mo o s will
apply he ib a ion, and he equency on which he da a will be collec ed a e all selec ed/con igu ed
using he sma phone App o +sensBand. Du ing he use o his s a egy all mo o s we e selec ed o
p o ide ib a ions a ound he wais , and we e se a 200Hz o ib a e du ing 200ms, acco ding o [19].
Fig. 9. Gai e en -d i en cueing deli e y schema ic.
The gai e en de ec ion algo i hm plays an impo an ole also in his cueing s a egy. By
segmen ing he gai e en s, i is possible o de ec he TO’s o bo h ee , ha a e hen dis inguished
conside ing ha he igh oo HS’s p esen s angula eloci y in he An e o Pos e io plane (
gy _AP
) below
0, and le HS’s p esen
gy _AP
abo e 0 [80].
Closed-loop ib a ion (200ms a igh TO)
51
When he TO’s a e de ec ed and hey co espond o he igh oo TO’s, he mo o s a e ac i a ed,
and he pa ien ecei es he ib a ions se a he espec i e du a ion and equency.
3.2.3. CONTINUOUS CUEING
Li e a u e e iew as shown ha one o he main s a egies o empo al o ien a ion o PD pa ien s
was he use o audi o y cues se as me onomes, since i p o ides a ixed hy hm o pa ien s o ollow
and y o ma ch hei pace. Howe e , audi o y cueing comp ises se e al disad an ages conce ning
p ac ical aspec s, conside ing ha audio can be in usi e. Soma osenso y cueing, on he o he hand,
when applied a a ixed hy hm can be also used wi h he same pu pose bu in a mo e speci ic and
disc e e manne .
Con inuous cueing, like a me onome, consis s in p o iding con inuous ib a o y open loop
cueing o a de ined pe iod o ime and a a ce ain equency. This s a egy was con igu a ed in o de o
p o ide ib a ions du ing 2s pulses, se a 200 Hz, acco ding o [19]. Fig. 10 shows he schema ic o his
cueing mode deli e y.
Fig. 10. Con inuous eedback deli e y schema ic.
3.2.4. FOG PREVENTION CUEING
By analysing he li e a u e e iew i was no ed ha , o e all, he cueing s a egies ha
p edominan ly esul ed in highe gai imp o emen s and in highe FoG episodes educ ion we e he isual
cues. When applied a open loop p o iding pa ien s spa ial o ien a ion on whe e o place each s ep, such
as ape lines, ac ed as ac i e s a egies o pa ien s o o e come FoG.
Soma osenso y cues also had g ea po en ial o PD mo o ehabili a ion and in p e en ing FoG
episodes, especially when applied a a closed loop hy hm, allowing o be pe sonalized o each pa ien 's
gai pe o mance o help hem be mo e awa e on when o s ep.
Combining bo h open loop isual cues wi h closed loop soma osenso y cues, he e o e, may ha e
g ea e po en ial in a ge ing FoG, since i can ac as bo h an ac i e and p e en i e s a egy, allowing a
mo e comple e senso y s imula ion o he pa ien s. The e o e, i was de eloped a no el senso y cueing
Open-loop ib a ion (2s pulses)
52
s a egy using he +sensBand sys em ha simul aneously p o ided bo h wo cueing modali ies, he FoG
p e en ion cueing.
FoG p e en ion cueing ini ially p o ides ib a ion a a ixed con inuous a e (open-loop) o help
gai ini ia ion, along wi h he p ojec ion o wo ixed lase lines in on o he pa ien (also open-loop). A e
s a ing walking, he lase lines emain ixed and open-loop. Howe e , he ib o ac ile cues change o
closed-loop a e he i s TO o he igh oo being de ec ed. Fig. 11 ep esen s he cueing s a egy
deli e y.
Fig. 11. FoG p e en ion cueing deli e y schema ic.
Closed-loop ib a ion is enabled a he igh oo ’s oe-o gai e en du ing 200 ms and is se a
200 Hz, using he same algo i hm o gai e en de ec ion as he gai e en -d i en s a egy. The ib a ion
is se o 2s pulses. Since he s a o walking usually elici s FoG, hy hmic open-loop ib a ion is enabled
be o e gai ini ia ion, ein o cing he i s s ep execu ion [56], [73]. The e o e, a e de ec ing he i s oe-
o , he con ol s a egy swi ches o closed-loop ib a ion o pe sonalize walking acco ding o he use s'
sel -selec ed pace.
Visual cues a e con inuously p o ided by ac i a ing he lase s o p ojec wo lines in on o he
pa icipan s. These lines mo e along wi h use s so hey a e always seeing he nex wo s eps whe e he
oo should be placed. The e o e, isual cues a e applied o assis du ing a FoG episode, while ib o ac ile
cues also assis du ing gai ini ia ion bu du ing walking o wa d, u ning, and passing h ough doo s a e
applied also o p e en he occu ence o FoG episodes.
3.3. DESKTOP APP (MATLAB)
In o de o u he p ocess and analyse he collec ed gai da a, and calcula e spa io empo al gai
me ics, i was de eloped a desk op clinical App using Ma LAB (Fig. 12).
Open-loop ib a ion (2s pulses)
Closed-loop ib a ion (200ms a igh TO)
Open-loop lase lines (adjus ed o pa icipan s s ep leng h)

53
Fig. 12. Desk op clinical App.
This App was designed o ead he IMU da a iles ob ained using he +sensBand, p ocess he
da a using se e al il e s, and hen unning he da a h ough a ini e s a e machine (FSM) o de ec IC’s
(HS) and FC’s (TO) o each oo , o u he calcula e gai me ics. The FSM algo i hm was de eloped o
emula e he FSM o he +sensband i sel . Howe e , o a oid he accumula ion o any e o s, such as leg
dis inc ion e o s, he Ma lab FSM includes wo addi ional condi ions ha ensu e ha a e de ec ing a
gai cycle o he igh leg, he nex gai cycle mus be de ec ed o he le leg, and ice e sa. Ins ead o
5 s a es, he imp o ed FSM con ained a o al o 10 s a es, co esponding o he di e en gai e en s in
o de bu o bo h igh and le ee (e.g. HS_ igh , FF_ igh , TO_le , MS _ igh , HO_ igh , HS_le ,
FF_le , TO_ igh , MS _le and HO_le ). I he FSM de ec s consecu i e gai cycles on he same leg, he
s a e ese s o co ec any po en ial e o s.
A TO i was inse ed he gy _AP condi ion o leg dis inc ion. A e de ec ing igh o le HS, he
subsequen e en s (FF, MS , and HO) a e associa ed o ha same leg, excep o he TO e en ha is
associa ed o he opposi e leg, in line wi h he desc ip ions o [81], [82]. Since i is impo an o ensu e
ha a e a igh leg de ec ion he ollowing me ics should be o he opposi e leg, and ice e sa, he
algo y hm was se by i s de ec ing he 5 e en s o he igh leg gai cycle ( HS_ igh , FF_ igh , TO_le ,
MS _ igh and HO_ igh ), and hen ollowing wi h he same sequence o gai e en s bu o he le leg. I
he e en s we e de ec ed wi h he gy _AP condi ion o leg dis inc ion de ec ing a igh o le leg, and he
ollowing and p e ious legs we e de ec ed as he opposi e legs, he gai e en s we e co ec ly de ec ed,
he e o e we e inse ed in he IC and FCs a ays o gai me ics calcula ion. Else, i mean ha wo gai
cycles o he same leg we e de ec ed consecu i ely, p esen ing an e o . In his case, he espec i e gai
e en s de ec ed in ha gai cycle we e emo ed, and Nans we e added on hose posi ions ins ead. This
app oach lags e o s, wi h Nans ma king cycles whe e he expec ed leg al e na ion ailed, hus p e en ing
inaccu a e gai me ics om being calcula ed.
54
Depending on which oo he pa ien ini ia es gai , wo FSM cycles we e c ea ed: one s a ing wi h
he le oo i s and o he s a ing wi h he igh . Fo be e unde s anding o he imp o ed FSM algo i hm,
Fig. 13 p esen s a lowcha wi h he o de o gai e en de ec ion highligh ed by he colo co esponding
o he i s oo de ec ed in he FSM algo i hm. When using he FSM_Righ _ i s cycle, e en s a e de ec ed
in he o de indica ed by he blue numbe s, whe eas in he FSM_Le _ i s cycle, e en s ollow he o de
ma ked by he ed numbe s. Ul ima ely, he inpu accele a ion and angula eloci ies componen s
(𝑎𝑐𝑐𝑛,𝑔𝑦𝑟𝑛), a e p ocessed h ough he FSM algo i hm o iden i y he espec i e IC’s and FC’s o each oo
(𝐼𝐶𝑅,𝐹𝐶𝑅,𝐼𝐶𝐿,𝐹𝐶𝐿).
Fig. 13. Flowcha o he imp o ed FSM o gai segmen a ion Ma Lab algo i hm.
55
A e p ocessing and de ec ing ICs and FCs o each oo , gai me ics we e calcula ed o bo h
igh and le legs, including s ep ime and leng h; s ide ime and leng h; s ance, swing and double
suppo phase; gai speed; cadence and numbe o s eps aken. Addi ionally, asymme ies and s anda d
de ia ions we e also compu ed by compa ing me ics o bo h ee and analysing a iabili y, espec i ely,
o a mo e comp ehensi e unde s anding o he pa ien 's gai cha ac e is ics. Table 4 p esen s he
equa ions used o ob ain each spa io empo al gai me ic calcula ed.
Table 4. Me hods o es ima ing gai spa io empo al me ics ep esen ed by domain, me ic, o mula and measu emen uni s
Domain
Me ic
Fo mula
Uni s
Pace
S ep leng h
𝑚𝑒𝑑𝑖𝑎𝑛((2√𝐿ℎ−ℎ2,ℎ=∬ 𝑎𝑐𝑐𝑉
𝐼𝐶2𝑖
𝐼𝐶1𝑖),(2√𝐿ℎ−ℎ2,ℎ=∬ 𝑎𝑐𝑐𝑉
𝐼𝐶1𝑖+1
𝐼𝐶2𝑖))
Me e s (m)
Veloci y
𝑚𝑒𝑑𝑖𝑎𝑛(𝑆𝑡𝑒𝑝 𝑙𝑒𝑛𝑔𝑡ℎ1𝑖
𝑆𝑡𝑒𝑝 𝑡𝑖𝑚𝑒1𝑖 ,𝑆𝑡𝑒𝑝 𝑙𝑒𝑛𝑔𝑡ℎ2𝑖
𝑆𝑡𝑒𝑝 𝑡𝑖𝑚𝑒2𝑖)
Me e s pe second (m/s)
Cadence
𝑚𝑒𝑑𝑖𝑎𝑛(𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑦1𝑖∙60
𝑆𝑡𝑒𝑝 𝑙𝑒𝑛𝑔𝑡ℎ1𝑖,𝑉𝑒𝑙𝑜𝑐𝑖𝑡𝑦2𝑖∙60
𝑆𝑡𝑒𝑝 𝑙𝑒𝑛𝑔𝑡ℎ2𝑖)
S eps pe minu e
(s ep/min)
Rhy hm
S ep ime
𝑚𝑒𝑑𝑖𝑎𝑛((𝐼𝐶2𝑖−𝐼𝐶1𝑖),(𝐼𝐶1𝑖+1−𝐼𝐶2𝑖))
Seconds (s)
S ance phase
𝑚𝑒𝑑𝑖𝑎𝑛((𝐹𝐶2𝑖−𝐼𝐶1𝑖),(𝐹𝐶1𝑖+1−𝐼𝐶2𝑖))
Pe cen age (%)
Swing phase
𝑚𝑒𝑑𝑖𝑎𝑛((100−𝑠𝑡𝑎𝑛𝑐𝑒𝑝ℎ𝑎𝑠𝑒1),(100−𝑠𝑡𝑎𝑛𝑐𝑒𝑝ℎ𝑎𝑠𝑒2))
Pe cen age (%)
Va iabili y
SD s ep leng h
𝑆𝐷(𝑠𝑡𝑒𝑝𝑙𝑒𝑛𝑔𝑡ℎ1,𝑠𝑡𝑒𝑝𝑙𝑒𝑛𝑔𝑡ℎ2)
Me e s (m)
SD eloci y
𝑆𝐷(𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦1,𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦2)
Me e s pe second (m/s)
SD s ep ime
𝑆𝐷(𝑠𝑡𝑒𝑝𝑡𝑖𝑚𝑒1,𝑠𝑡𝑒𝑝𝑡𝑖𝑚𝑒2)
Seconds (s)
Assyme y
AS s ep leng h
|𝑠𝑡𝑒𝑝𝑙𝑒𝑛𝑔𝑡ℎ1−𝑠𝑡𝑒𝑝𝑙𝑒𝑛𝑔𝑡ℎ2|
Me e s (m)
AS eloci y
|𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦1−𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦2|
Me e s pe second (m/s)
AS s ep ime
|𝑠𝑡𝑒𝑝𝑡𝑖𝑚𝑒1−𝑠𝑡𝑒𝑝𝑡𝑖𝑚𝑒2|
Seconds (s)
1 and 2: ep esen he i s and second leg acco ding o he o de de ec ed ( igh leg i s o le leg i s ); i: ins an ; L: dis ance om loo o
whe e he back IMU is placed in he use body; h: e e s o he double in eg a ion o e ical accele a ion; 𝒂𝒄𝒄𝑽: e e s o he e ical
accele a ion; SD: s anda d de ia ion; AS: asymme y
3.4. CONCLUSIONS
Following an in-dep h analysis o he cu en cueing s a egies and echnologies o gai
ehabili a ion in Pa kinson's disease (PD) pa ien s, i was ound ha he in eg a ion o ad anced sys ems
like he +sensBand and Xsens holds signi ican po en ial. These echnologies no only p o ide di e se
cueing modali ies bu also o e obus da a collec ion and alida ion capabili ies, which a e essen ial o
e alua ing mo o e ec s, pa icula ly in add essing FoG and imp o ing quali y o li e.
Thus, his disse a ion ocuses on explo ing and implemen ing hese echnologies wi hin he
amewo k o ongoing clinical s udies aimed a de eloping and assessing no el cueing s a egies o PD.
To his end, he disse a ion p o ides a de ailed o e iew o he sys em componen s, he con ol and
56
moni o ing s a egies employed, and he ools used o da a p ocessing and analysis. This wo k is pa o
he b oade +sense p ojec , con ibu ing speci ically o he ad ancemen o gai ehabili a ion echniques
using s a e-o - he-a equipmen such as he +sensBand sys em and he Xsens mo ion acking sys em.
63
I hink he a ious unc ions o he sys em a e e y well in eg a ed.
4.43±0.83
I hough he e was oo much inconsis ency in his sys em.
1.48±0.92
I would imagine ha mos people would lea n o use his sys em e y quickly.
4.38±1.09
I ound he sys em e y cumbe some o use.
1.29±0.73
I el e y con iden using he sys em.
4.34±0.76
I needed o lea n a lo o hings be o e I could ge going wi h his sys em.
1.45±0.90
Like scale: 1 – s ongly disag ee; 2 – disag ee; 3 – indi e en ; 4 – ag ee; 5 – s ongly ag ee.
4.2.1. GAIT METRICS
Table 9 shows ha pace was he domain ha p esen ed mo e signi ica i e di e ences. In his
gai domain, s ep leng h inc eased he mos in SG2 (≈18.95%). Veloci y only showed signi ican
di e ences be ween he CG and each s udy g oup (SG1 o SG2), inc easing wi h bo h cueing s a egies
in compa ison o he CG (≈10.70% o SG1 and ≈20.14% o SG2). No signi ican di e ences we e no iced
o eloci y in be ween he wo s udy g oups (SG1 and SG2). As o cadence, only we e no iced signi ican
di e ences be ween CG and SG1, whe e SG1 showed a g ea e inc ease o his me ic (≈8.05%). Bo h
s udy g oups we e able o coun e ac he p o o ypical pace beha iou o PD, imp o ing pace me ics wi h
cueing.
Simila ly, he hy hm domain p esen ed signi ican di e ences o he swing phase be ween he
CG and each s udy g oup (SG1 o SG2). Bo h g oups ha used he wo cueing modali ies showed g ea e
imp o emen s o swing phase (≈5.36% o SG1 and ≈8.96% o SG2), by p esen ing highe swing phases.
No signi ican di e ences we e obse ed o s ep ime o s ance phase. By conside ing he swing phase
imp o emen s, i is possible o unde s and ha bo h cueing s a egies we e able o coun e ac he
p o o ypical hy hm beha iou o PD.
As o he a iabili y and asymme y aspec s o gai , di e en esul s we e obse ed. Fo he
a iabili y domain, signi ican di e ences we e no iced o SD s ep leng h and SD eloci y. Fo bo h me ics
i was no iced ha cueing s a egies wo sened gai a iabili y by ollowing he p o o ypical beha iou o
PD. Howe e , in SD s ep leng h, SG1 p esen ed signi ican di e ences when compa ed wi h SG2
indica ing ha gai e en -d i en cueing (SG1) had be e esul s o SD s ep leng h han con inuous cueing
(SG2) (≈46.79% imp o emen be ween SG1 in compa ison o SG2). No signi ican di e ences we e
no iced be ween CG and SG1. SD eloci y, on he o he hand, only p esen ed signi ican di e ences
be ween CG and SG2. Those di e ences esul ed in wo sened SD eloci y o he SG2 (≈45.94%). Simila
esul s we e no iced o AS s ep leng h and AS s ep ime, whe e SG2 pe o med wo se han CG ( esul s
wo sened by almos wice he CG alue). The e o e, i is possible o no e ha con inuous cueing wo sened

64
bo h a iabili y and asymme y aspec s o gai pe o mance. Gai e en -d i en cueing, when compa ed o
he con inuous cueing, p o ides less a iabili y o s ep leng h du ing gai pe o mance.
O e all, he CG p esen ed less dispe sion o he SD alues, ollowed by SG1 and las ly SG2. SG2
was he g oup ha p esen ed highe dispe sion o he SDs.
4.2.2. SYSTEM USABILITY SCALE (SUS)
Conside ing each pa ien ’s o al SUS sco e, i was ob ained a mean±SD sco e o 86.32±6.90%.
This esul indica es ha he +sensBand was pe cei ed wi h an excellen usabili y o e all, being e alua ed
as an easy- o-use (4.77±0.46% o ques ion 3), cohesi e (4.43±0.83% o ques ion 5), in ui i e
(4.38±1.09 o ques ion 7), and com o able sys em (1.29±0.73 o ques ion 8). O e all, pa ien s also
el sa e and mo e con iden abou hei gai pe o mance when using he sys em (4.34±0.76% o
ques ion 9).
Howe e , some mo e neu al sco es we e no iced when answe ing he i s ques ion abou i
pa ien s would like o use he sys em equen ly (3.77±1.20%). A good pa o he pa ien s was s ill capable
o walking au onomously, wi hou any majo gai changes, he e o e when asked his ques ion, did no
eel he need o use a sys em daily o imp o e hei gai . Some commen s we e also no ed ega ding he
design o he de ice no being he mos appealing o day- o-day use, especially o use in public. The
weigh o he sys em was also commen ed abou o be u he imp o ed, since i may no be he mos
com o able o daily usage. On ano he hand, some pa ien s also no ed ha al hough a i s impac
eeling a bi in imida ed by he bulky design, hey quickly adap ed o i and accep ed i . Du ing he ials,
i was no epo ed any discom o while using he de ice.
4.3. DISCUSSION
Conside ing he p o o ypical mo o beha iou o PD, his sec ion in ends o analyze and discuss
he gai me ics esul s o unde s and i PD pa ien s imp o ed o no hei gai pe o mance. I is impo an
o compa e also bo h cued g oups' esul s o p o ide insigh s abou which o he wo cueing me hods
may be mo e bene icial o PD pa ien s' mo o ehabili a ion, and he e o e achie e he main goal o his
s udy.
The esul s ob ained showed ha o e all, bo h cueing s a egies we e able o imp o e pa ien s'
gai pe o mance by a ge ing he pace and hy hm aspec s o gai , he e o e mi iga ing he hypokine ic
and b adykine ic gai pa e ns o PD. When compa ing bo h cued g oups wi h he con ol g oup i was
possible o no e imp o emen s o s ep leng h (≈18.95% o he SG2), eloci y (≈10.7% o he SG1 and
65
≈20.14% o he SG2), cadence (≈8.05% o he SG1), and swing ime (≈5.36% o he SG1 and ≈8.96%
o he SG2). Al hough bo h gai e en -d i en (SG1) and con inuous (SG2) ib o ac ile cueing p esen ed
imp o emen s o bo h pace and hy hm, con inuous eedback p o ided g ea e imp o emen s o e all
(≈19.54% in mean o pace and ≈8.96% o hy hm) han gai e en -d i en cueing (≈9.38% in mean o
pace and ≈5.36% o hy hm), ha only showed g ea e imp o emen s o cadence.
Howe e , when compa ing di ec ly SG1 and SG2 esul s, mixed conclusions we e aken
depending on he me ics being analyzed. Only s ep leng h and SD s ep leng h me ics p esen ed
signi ican di e ences be ween he wo s udy g oups (SG1 and SG2), co esponding o wo di e en
aspec s o gai (pace and a iabili y, espec i ely). Fo s ep leng h, he g oup wi h con inuous eedback
(SG2) had g ea e imp o emen s han he one ecei ing gai e en -d i en cueing (SG1). On he o he
hand, when conside ing he SD s ep leng h, gai e en -d i en cueing showed imp o emen s when
compa ed o he g oup ecei ing con inuous eedback. Mo eo e , he SG2 g oup when compa ed o he
CG g oup did no p esen any imp o emen s o gai a iabili y and asymme y me ics, ins ead i was
no iced an inc ease, he e o e wo sen, o SD s ep leng h (≈60%), SD eloci y (≈45.94%), AS s ep leng h
(≈114.95%) and AS s ep ime (≈165.57%) wi h he use o open loop ib a ion.
Conside ing hese esul s, i was hypo hesized ha con inuous eedback, by deli e ing he
ib a ions a 2s pulses simila ly o a me onome, caused mo e ab up imp o emen s in pace and hy hm
han closed-loop ib a ion and also highe dispe sion o SDs. Al hough SG2 p esen ing highe eloci y and
s ep leng h han he o he g oups, he esul an cadence is lowe han he SG1. SG2 pa ien s ook
app oxima ely 3.41 s eps pe each 2s passed, meaning ha comple ed app oxima ely one gai cycle o
one oo ( om IC o he nex IC o he same oo ) du ing each 2s ecei ing he ib a ions. This sugges s
ha pa ien s may unin en ionally ollowed he hy hm o he ib a ions and ied o ill he pulse du a ion
by aking less s eps bu longe and sligh ly quicke . Howe e , since hese a e mo e ab up changes o
pace han he changes no iced in he SG1, i caused highe ins abili y o gai ansla ed by wo sened
a iabili y and asymme y. Gai e en -d i en cueing, on he o he hand, by p o iding he ib a ions a a
hy hm pe sonalized o he pa ien ’s own pace, i consis s o a mo e sub le s a egy. The e o e, mo e
sub le changes o he gai pa e n we e no iced, bu less gai a iabili y is no iced since pa ien s pe o m
an imp o ed e sion o hei own gai ins ead o o cing hei gai o ma ch he cues.
This highligh s he po en ial o gai e en -d i en cueing which should no be disca ded since i
causes g ea imp o emen s o pace and hy hm s ill and may su ge as a be e choice o s a egy in
p ese ing pa ien s' gai a iabili y. Also, since he s udy g oups consis ed in h ee independen
popula ions he esul s may be a ec ed by pa ien s di e en clinical cha ac e is ics such as H&Y s age
66
and UPDRS III ha was mo e ad anced in SG1 when compa ed o bo h CG and SG2 (SG1 showed mean
and SD o H&Y o 1.86±0.76 poin s, indica ing unila e al and axial owa ds bila e al in ol emen , and
UPDRS III sco e o 19.44±12.43 poin s, co esponding o a mild mobili y s age). Also, al hough he CG
g oup has mo e PD pa ien s wi h FoG (9 eeze s) han SG1 and SG2 (5 eeze s each), he SG1 eeze s
showed highe sco es o he NFOG-Q (21.40±7.47). These clinical cha ac e is ics may indica e ha
pa ien s ecei ing he gai e en -d i en cueing p esen ed mo e ad anced s ages o he disease, he e o e
p esen ing wo se mobili y and gai pa e ns. Ul ima ely, pa ien a iabili y may ha e a ec ed he esul s
o he s udy, and p o iding a c oss-o e analysis ollowing he p o ocol in he u u e should be conside ed
o imp o e he s udy conclusions.
Las ly, he SUS esul s o he usabili y aspec s o he de ice ha e shown ha o e all, he
+sensBand sys em was e alua ed wi h an excellen usabili y sco e, being accep able o u he
implemen a ion in longe ials. Howe e , in he u u e, he design should be imp o ed o become mo e
appealing o daily usage and less hea y o imp o e com o o mo e ex ended pe iods o use.
4.4. CONCLUSIONS
Bo h open and closed-loop cueing s a egies, as expec ed, imp o e gai pe o mance by
mi iga ing he hypokine ic and b adykine ic aspec s o gai . Al hough o e all con inuous eedback causing
g ea e imp o emen s o pace and hy hm han gai e en -d i en cueing, i inc eased gai a iabili y and
asymme y. Conside ing he di e en ad an ages o bo h cueing s a egies, hey bo h ha e g ea
accep abili y and po en ial o mo o ehabili a ion o PD pa ien s, mo eo e i explo ed in a combined
new s a egy.
67
5. USABILITY STUDY
This s udy se es o in oduce he FoG p e en ion cueing, wi h heal hy pa icipan s o a p ima y
alida ion o he usabili y and wo kload o his no el s a egy, while simul aneously using he +sensBand
de ice in di e en en i onmen s, o e alua e also he po en iali y o being u he used as a daily solu ion.
The e o e, he p ima y objec i e o his s udy is o alida e he usabili y and wo kload o his new
cueing s a egy in heal hy pa icipan s as a p elimina y s ep be o e es ing in he PD popula ion. By
e alua ing he sys em's impac on gai empo al me ics and usabili y h ough he Sys em Usabili y Scale,
his esea ch aims o ensu e he de ice's e ec i eness and com o in daily use. Posi i e indings in his
ini ial phase will pa e he way o u he ials ocused on PD pa ien s, ul ima ely aiming o imp o e hei
mobili y and o e all quali y o li e.
5.1. METHODOLOGY
5.1.1. PARTICIPANTS
A o al o i een heal hy pa icipan s (5 males and 10 emales; 30.47 ± 14.09 yea s; highe
educa ion le el; daily use o echnologies) we e included in his s udy. To be eligible o in eg a e his
s udy, pa icipan s had o mee he ollowing c i e ia: (i) no p esen any neu odegene a i e diso de s and
gai impai men s; (ii) be able o walk independen ly wi hou any assis ance; and (iii) sign in o med consen
o pa icipa e in he s udy. The s udy was app o ed by he e hical E hics Commi ee CEICVS 006/2020,
acco ding o he Decla a ion o Helsinki and O iedo Con en ion.
5.1.2. PROTOCOL
A e ins umen ing pa icipan s wi h he +sensBand wais band (Fig. 15 a)) and con igu a ing he
FoG p e en ion eedback s a egy, a i s calib a ion ial was execu ed, ins uc ing hem o ake one s ep
o wa d o adjus he +sensBand lase s o ma ch he use ’s sel -selec ed s ep leng h. A e calib a ing he
lase s, h ee walking ials we e pe o med, in one o h ee possible pa hs (in wo k o home-based
scena ios). Fig. 15 b) shows he di e en en i onmen s used.
a) Ins umen ed subjec
68
Fig. 15. FoG p e en ion eedback usabili y s udy es (a)) and cou ses (b)).
The wo k-based en i onmen was a longe walking pa h ( o al o 64m) pe o med on a hallway o
Uni e si y o Minho, a Azu ém Campus, and only 3 pa icipan s we e selec ed o his scena io due o
hei a ailabili y o a el o he si e and amilia iza ion wi h he cou se. As o he home-based scena ios
(A and B), hey ook place in wo di e en houses, common o he pa icipan s. Conside ing space
limi a ions o he di e en housings, sho e walking cou ses we e conside ed (A: o al o 13m; B: o al o
16m). The pu pose o a home ials was o assess pa icipan s in a secu e and amilia se ing, common
o hei daily ou ine, while using he de ice. By hese means, pa icipan s we e alloca ed o each cou se
conside ing hei day- o-day amilia iza ion wi h en i onmen s A o B.
A he s a o each ial, pa icipan s we e ins uc ed o s and in place o 20s, while ecei ing
he open-loop ib a ions and ixed lase lines. A e hose 20s, pa icipan s s a ed walking and he
ib o ac ile mode changed o closed loop a e de ec ing he i s igh oo TO. Du ing he walking pa h,
pa icipan s u ned 90 deg ees, passed h ough a leas one doo , u ned 180 deg ees and walked back
o he s a posi ion. Du ing each ial he esea che s accompanied he pa icipan s o con i m ha he
+sensBand was p o iding he cues co ec ly.
5.1.3. DATA ACQUISITION
Each pa icipan 's demog aphic in o ma ion was collec ed conside ing age, gende , educa ion
le el, and echnology usage.
To assess sys em's usabili y and wo kload, he Sys em Usabili y Scale (SUS) and NASA Task
Load Index (TLX) ques ionnai es we e answe ed anonymously o educe isk o bias. SUS ques ionnai e
allowed o e alua e use expe ience aspec s such as willingness o use he sys em o en, complexi y, ease
o use, consis ency, com o and con enience. As o he NASA TLX ques ionnai e, h ee ypes o demands
we e e alua ed (men al, physical, and empo al), bu also pe o mance, e o , and us a ion el . The
A
B
3m
3.50m
5m
3m
20m
7m
5m
b) S udy walking cou ses

69
s a emen s om he SUS and NASA TLX ques ionnai es we e a ed using a Like scale, anging om 1
(S ongly disag ee) o 5 (S ongly ag ee) and 1 (Low) o 21 (High), espec i ely. Resul ing SUS sco es
anged om 0 o 100, allowing o be ca ego ized as aw ul, poo , easonable, good, o excellen o sco es
lowe hen 51, om 51 o 68, equal o 68, om 68 o 80.3, o g ea e hen 80.3, espec i ely. Simila ly,
NASA TLX esul ing sco es anged also om 0 o 100, ca ego izing wo kload as e y high, high, somewha
high, medium, o ligh o sco es highe hen 80, om 50 o 80, om 30 o 50, om 10 o 30, o lowe
hen 10, espec i ely.
5.1.4. DATA PROCESSING
Gai me ics we e also assessed by he +sensBand du ing each ial, by ac i a ing he mo ion
moni o ing op ion while pe o ming he FoG p e en ion eedback. The da a iles ob ained we e hen
p ocessed using he desk op clinical App, o ob ain he ollowing empo al gai me ics: s ide, s ep,
s ance, and swing imes, and cadence.
T ansi ions om HS o FF, and om FF o TO we e iden i ied and s o ed in lis s o de ec Ini ial
Con ac s (IC) and Final Con ac s (FC) du ing gai . Those lis s we e hen used o calcula e each empo al
pa ame e as ollows: s ep ime and s ide ime we e calcula ed as he empo al dis ance om IC o one
oo o he ollowing, and one IC o he nex o he same oo , espec i ely; s ance ime was calcula ed as
he empo al dis ance om IC o FC o he same oo ; swing ime was calcula ed as he di e ence be ween
s ide and s ance ime ( espec i e phases we e ob ained by con e ing hose imes in pe cen ages); and
cadence was calcula ed by coun ing he numbe o s eps, co esponding o he leng h o he lis o IC,
and di iding i by he gai du a ion, he e o e ime dis ance be ween i s and las IC. The mean empo al
gai me ics we e calcula ed o each pa icipan conside ing he gai cycles pe o med du ing he p o ocol.
Finally, mean and s anda d de ia ion o he empo al gai me ics ob ained om all pa icipan s we e
p esen ed o he esul s sec ion.
5.2. RESULTS
Table 11 p esen s bo h demog aphics and esul s ob ained o each pa icipan based on he
ques ionnai e answe s e alua ing he usabili y and wo kload o he FoG p e en ion eedback using he
+sensBand. Fu he mo e, mean and s anda d de ia ions (mean±SD) o bo h SUS and NASA TLX scales
a e p esen ed. Table 12 hen p esen s he mean esul s o pa icipan s' empo al gai me ics.
70
Table 11. Heal hy pa icipan s’ age, gende (F – emale; M – male), Educa ion le el, Technology usage, Walking cou se
pa h (A – Home-based pa h A; B – Home-based pa h B; W – Wo k-based), SUS and NASA TLX sco es, and co esponding
quali a i e le els (colo ed h ough g een o ed om be e h ough wo s usabili y, espec i ely)
ID
Age
(yea s)
Gende
Walking
pa h
SUS
(sco e, le el)
NASA TLX
(sco e, le el)
P1
19
F
B
82.50
Excellen
18
Medium
P2
M
B
80
Good
11
Medium
P3
22
F
W
90
Excellen
9
Ligh
P4
A
75
Good
30
Somewha
high
P5
A
100
Excellen
18
Medium
P6
A
92.50
Excellen
11
Medium
P7
24
F
A
90
Excellen
15
Medium
P8
25
F
A
82.50
Excellen
15
Medium
P9
27
F
W
92.50
Excellen
21
Medium
P10
F
W
52.50
Poo
19
Medium
P11
M
B
77.50
Good
37
Somewha
high
P12
30
M
B
72.50
Good
21
Medium
P13
55
M
B
82.50
Excellen
8
Ligh
P14
57
F
B
97.50
Excellen
10
Medium
P15
59
M
B
77.50
Good
29
Medium
(mean±SD)
30.47±14.09
F
B
83±14.08
Excellen
18.1±9.36
Medium
Table 12. Heal hy pa icipan s’ mean s ep, s ide, s ance and swing imes, s ance and swing phases, and cadence, and
espec i e o al means and s anda d de ia ions
ID
S ep ime
(s)
S ide ime
(s)
S ance ime
(s)
S ance
phase (%)
Swing
ime (s)
Swing
phase (%)
Cadence
(s eps/min)
P1
0.69
1.37
0.91
66.20
0.46
33.80
87.20
P2
0.68
1.34
0.91
68.20
0.43
31.80
88.24
P3
0.80
1.41
1.02
72.24
0.39
27.75
74.69
P4
0.63
1.25
0.77
61.66
0.48
38.33
95.15
P5
4.11
6.29
4.68
74.40
1.61
25.60
14.58
71
P6
1.10
1.95
1.44
73.46
0.52
26.54
56.31
P7
2.58
3.92
2.78
71.59
1.15
28.41
29.35
P8
1.14
1.94
1.29
66.52
0.65
33.48
54.27
P9
0.59
1.11
0.82
73.78
0.29
26.22
102.41
P10
1.05
2.06
1.33
64.40
0.74
35.60
56.85
P11
0.89
1.63
1.11
67.98
0.52
32.02
70.34
P12
1.78
3.16
2.24
70.60
0.92
29.40
33.61
P13
1.23
2.13
1.47
69.83
0.66
30.17
49.99
P14
2.48
4.63
2.83
61.72
1.80
38.28
25.19
P15
2.75
4.62
2.95
64.72
1.67
35.28
25.41
(mean±SD)
1.50±0.99
2.59±1.52
1.77±1.07
68.49±4.09
0.82±0.49
31.51±4.09
57.57±27.17
5.2.1. SYSTEM USABILITY SCALE (SUS)
SUS sco es we e p edominan ly excellen , wi h nine pa icipan s (8 emales and 1 male) sco ing
be ween 82.50 and 100, aged 19-57 yea s, and p edominan ly aining in home-based scena ios. Fi e
pa icipan s (1 emale and 4 males) sco ed good usabili y (72.50-80), wi h ages anging om 19 o 59
yea s, and aining mainly in home-based walking pa h B. Ul ima ely, one pa icipan ( emale) sco ed poo
usabili y (52.50), wi h 27 yea s and walking in wo k-based walking pa h. O e all, he esul s ob ained
we e posi i e, conside ing an excellen a e age sco e (83 ± 14.08).
5.2.2. TASK LOAD INDEX (NASA TLX)
Only wo pa icipan s (1 emale and 1 male) sco ed he lowes wo kload le el (ligh ), anging om
8 o 9, aged 22 and 55, and execu ing wo k and home-based B walking pa hs, espec i ely. Medium
wo kload was he p edominan sco e le el, wi h ele en pa icipan s (8 emales and 3 males) sco ing
be ween 10 and 29, aged 19-59 yea s, and p edominan ly aining in bo h home-based cou ses. Two
pa icipan s (1 emale and 1 male) p esen ed ewe posi i e esul s, sco ing somewha high (30 and 37),
wi h ages 22 and 27 yea s, and aining in home-based pa hs A and B, espec i ely. Al hough no eaching
he bes wo kload sco e le el, he men al demand was s ill posi i e, wi h an o e all medium a e age le el
(18.10 ± 9.36).
5.2.3. TEMPORAL GAIT METRICS
By p ocessing he da a me ics ob ained, i was possible o calcula e he mean and s anda d
de ia ions conside ing all pa icipan s o s ep ime (1.50s±0.99s), s ide ime (2.59s ± 1.52s), s ance
ime (1.77s ± 1.07s), swing ime (0.82s ± 0.49s) and cadence (57.57s eps/min ± 27.17s eps/min).
72
Al hough no eaching he ideal s ance and swing phases (60% and 40%, espec i ely) he means ob ained
we e app oxima e o he expec ed esul ing in 68.49% ± 4.09% and 31.51% ± 4.09%, espec i ely [23].
5.3. DISCUSSION
Rega ding usabili y, pa icipan s gene ally e alua ed he cueing s a egy and +sensBand wi h an
excellen SUS sco e, he e o e conside ing i as highly usable [20]. Ul ima ely, he sys em was well-
ecei ed, conside ing i in ui i e o use and use - iendly among he pa icipan s. Howe e , i was epo ed
one poo le el o SUS sco e.
Al hough answe ing anonymously, mos o he pa icipan s answe ed he ques ionnai es in he
p esence o he esea che s so ha hey could be able o cla i y any doub s ha may a ise. Howe e ,
some pa icipan s couldn’ answe igh a e he ials he ques ionnai es, answe ing u he an online
e sion wi hou he esea che ’s p esence. Ul ima ely, hose cases epo ed di icul y in in e p e ing he
ques ionnai e. Ques ions e alua ing he necessi y/likabili y o including his de ice in hei daily ou ines
we e answe ed nega i ely in his case due o he pa icipan ocusing on hei lack o need o he de ice
a he han i s usabili y. Also, ques ions abou he pe o mance and echnological in eg a ion aspec s o
he de ice we e no answe ed co ec ly, since some pa ien s op ed o gi e a neu al sco e since hey
didn’ in end o commen on hose aspec s. All hese misin e p e a ion e o s ul ima ely may ha e a ec ed
he inal sco e. In he u u e, clea e ins uc ions should be conside ed o cla i y he scales, as long as
p io i izing answe ing he ques ionnai e immedia ely a e he mo o pe o mance, wi h he esea che s
s ill p esen .
As o wo kload pe cep ion including bo h physical and men al demands, analysis o NASA TLX
sco es e ealed a ying le els o wo kload expe ienced by pa icipan s du ing he ask. While mos
epo ed a medium wo kload le el, some expe ienced ligh e o somewha highe demands. I is impo an
o no e ha one pa icipan , du ing he walking cou se epo ed ha he lase p ojec ed lines, since we e
ad ancing as he pa icipan also p og essed, made he pa icipan us a ed du ing he i s ial ying
o always s ep o e hem. This may ha e ansla ed in o a high sco e o wo kload pe cei ed by his
pa icipan . Howe e , he es o he pa icipan s did no epo any ype o us a ion and conside ed he
lase lines an in ui i e cueing modali y, e en mo e han he ib a ions. When compa ing home wi h wo k-
based walking pa hs, he esul s o pe cei ed wo kload end o be posi i e in bo h en i onmen s. Home-
based en i onmen s may p esen low men al demand due o amilia i y and com o . Howe e , he wo
79
episodes occu ing o he ials bu du ing he sessions, i.e. du ing es in e als o o he momen s, we e
no coun ed as FoG obse a ions since we e no p esen when using he s a egy bu we e no ed abou
he asks done when occu ing. Du ing he e alua ion ime poin s es ablished in he p o ocol, NFOG-Q
answe s we e also collec ed o p o ide a clinical e alua ion o FoG e olu ion be o e and a e in e en ion,
and i he esul s we e main ained a e a week ( ollow-up).
Clinical da a was collec ed, e alua ing H&Y, MDS-UPDRS III and PDQ-39 clinical scales a speci ic
ime poin s o he p o ocol (p e- aining, pos - aining and ollow-up), be o e he mo ion moni o ing ials
du ing each p o ocol ime-poin , o e alua e PD clinical e olu ion be o e and a e ecei ing cues.
Simila ly, some ques ionnai es we e answe ed o e alua e di e en pa ame e s o he cueing
s a egy used along he sessions. NASA TLX answe s we e collec ed a he end o each aining session,
o e alua e cogni i e and physical demand o he s a egy, as well as answe ing a pe sonalized
ques ionnai e abou pa ien s’ pe cep ion o he di e en cues ecei ed du ing he ials, o e alua e how
he cueing s a egies we e being used along he aining sessions. The pe sonalized ques ionnai e
con ained ou ques ions, answe ed on a ange om 1 o 5, as shown in Appendix B. A he end o he
aining p o ocol (a e comple ing he session 12 cued aining ials) pa ien s answe s o he SUS we e
also collec ed o p o ide end-use s e alua ion o he usabili y o he cueing s a egy a e long- e m usage.
Du ing p e- aining, bo h Xsens and +sensBand allowed o acqui e, a 100Hz, and collec
accele a ion, angula eloci y and gai e en s da a. The +sensBand also allowed o collec hese me ics
in he middle o aining (session 6 o 7), a pos - aining and ollow-up. These wo sys ems we e used a
p e- aining o compa e bo h me ics ob ained a baseline and he e o e mo e ho oughly alida e he
+sensBand sys em pe o mance o gai e en de ec ion. The +sensBand sys em was also used du ing
he di e en ime poin s o unde s and pa ien s' gai pe o mance e olu ion be o e and a e using he
cueing s a egy.
6.1.4. DATA PROCESSING
Due o he educed numbe o pa ien s, o a be e unde s anding o he esul s da a p ocessing
consis ed in pe o ming desc ip i e analysis wi h he acqui ed da a wi h he esul s a ying along he
espec i e p o ocol imeline. A i s , i was impo an o analyze pa ien s calenda ized sessions and
ans o m ha calenda in a imeline.
Conside ing pa ien s' a ailabili y and di e en s a s o he p o ocol, he 12 sessions o aining
we e spaced di e en ly o each pa ien . Since he p e- aining and he i s session, which occu ed on
he same day, we e he e e ence sessions s a ing a day 0, each session ha ollowed was ma ked on
a imeline by conside ing he numbe o days ha passed since his e e ence. Ul ima ely, wo p o ocol

80
imelines we e ob ained, a imeline anging om 0 o 54 days o pa ien P1, and ano he anging om
0 o 46 days o pa ien P2.
O e all, he p o ocol o pa ien P1 ook 54 days o comple e, whe e aining ook 47 days, and
o pa ien P2 ook 46 days o comple e, whe e aining ook 39 days.
Acco ding o each ype o acqui ed da a and espec i e ime poin s o da a acquisi ion, he
imeline o he espec i e g aphic a ied acco ding o he sessions in be ween which he da a we e
collec ed. This means ha , da a ha was collec ed con inuously along each session, such as numbe o
FoG episodes pe cei ed in-be ween sessions and numbe o FoG episodes obse ed in each session
du ing ials, was g aphically p esen ed showing he esul s along each session o he ull p o ocol ( om
p e- aining o ollow-up). Da a ha ins ead o being collec ed a each session was collec ed only a he
p o ocol e alua ion ime poin s (p e- aining, mid- aining, pos - aining and ollow-up), such as +sensBand
mo ion moni o ing esul ing gai me ics, Xsens gai me ics, NFOG-Q esul s, MDS-UPDRS III esul s and
PDQ-39 esul s, was g aphically p esen ed showing each esul along he espec i e e alua ion imepoin
session. Las ly, da a e alua ed du ing he aining sessions (+sensBand gai me ics acqui ed while
ecei ing cues, NASA TLX esul s, and cue pe cep ion ques ionnai e esul s) was displayed g aphically
along a imeline anging om he i s aining session o he las (session a e p e- aining o session
be o e pos - aining).
I is impo an o no e ha H&Y, since i is a mo e speci ic scale designed o assess he o e all
s age o PD p og ession, was only e alua ed be o e and a e ecei ing he cued aining (p e- aining,
pos - aining, and ollow-up). This decision was made as he scale is no as sensi i e o sho - e m changes
h oughou he in e en ion, unlike he o he me ics. The e o e, gi en i s ela i ely s able na u e o e
ime, he H&Y esul s we e be e sui ed o be displayed in a able o ma .
SUS esul s we e also op ed o be displayed in able o ma since we e only assessed once, a e
comple ing he aining sessions and he e o e we e no e alua ed along a imeline.
Las ly, he +sensBand gai da a was p ocessed, simila ly o Chap e s 4 and 5, o ob ain he
ollowing gai me ics: s ep leng h, eloci y and cadence, o he pace domain; s ep ime, s ance phase
and swing phase, o he hy hm domain; SD o s ep ime, eloci y and s ep ime, o he a iabili y
domain; and AS o s ep ime, eloci y and s ep ime, o he asymme y domain o gai .
The Xsens da a had o be impo ed om MVN, selec ing accele a ion, angula eloci y bu also
oo con ac o se e as g ound u h o he gai e en de ec ion. Then he iles had o be con e ed o
ma
ype iles, o be able o be ead in Ma lab. Las ly, using he Clinical Desk op App he Xsens oo
con ac da a was p ocessed ob aining he same gai me ics as he +sensBand gai da a. Since i was
81
assessed h ee ials in each imepoin , i was calcula ed a each e alua ion ime poin s he mean and
SD o he ials o each pa ien +sensBand esul ing gai me ics. Simila ly, a p e- aining he mean and
SD o he ials o each pa ien we e calcula ed o bo h +sensBand and Xsens gai me ics. The
+sensBand esul s, since a e impo an o be analyzed along he di e en assessmen momen s, we e
hen displayed g aphically along he e alua ion imepoin s imeline.
6.2. RESULTS
Fo be e comp ehension o he esul s, his sec ion is segmen ed in ou subsec ions acco ding
o he di e en ypes o da a acqui ed: (i) FoG esul s, whe e obse ed episodes du ing ials, pe cei ed
episodes in be ween sessions, episodes du a ion e olu ion along he p o ocol imeline and NFOG-Q esul s
we e conside ed; (ii) clinical scales esul s, including H&Y, UPDRS III and PDQ-39, e alua ed only a
p o ocol ime-poin s; (iii) aining ques ionnai es esul s, including ques ionnai es done con inuously
in each aining session, such as NASA TLX and a cus om-made pe cei ed cues ques ionnai e, bu also
a he end o aining such as SUS ques ionnai e; and (i ) gai me ics esul s, analyzing bo h Xsens
and +sensBand gai collec ed da a du ing he e alua ion ime-poin s bu also con inuously along he
aining sessions.
6.2.1. FOG
Fig. 18, Fig. 19 and Fig. 20 p esen he g aphics ob ained o he FoG esul s.
Fig. 18. G aphics o FoG pe cep ion in be ween sessions s. FoG occu ences du ing ials, o pa ien s P1 and P2.
When analysing he esul s o pa ien s' pe cei ed FoG in be ween sessions s. FoG occu ences
du ing ials (Fig. 18), i is no iceable ha , o bo h pa ien s, he pe cei ed cu e (black line poin s) ends
o ha e unp edic able beha iou , bu he obse a ional cu e ( ed line poin s) ends o dec ease.
82
Pa ien P1 did no eeze du ing any ial along he p o ocol imeline. As o FoG episodes epo ed
in be ween sessions, pa ien P1 pe cei ed a maximum o 5 episodes in be ween sessions 5 and 6, ha
ook place 4 days apa . When analysing only he e alua ion ime poin s, om p e- aining o pos - aining
he numbe o FoG episodes pe cei ed inc eased (≈50%), and om pos - aining o ollow-up assessmen
he numbe o FoG episodes pe cei ed imp o ed g ea ly by dec easing o 0 episodes occu ing in be ween
pos - aining and ollow-up assessmen s.
Pa ien P2 p esen ed mo e cohesi e obse a ional esul s being no iceable a dec ease o FoG
du ing ials along he h ee main e alua ion ime poin s (≈100% in be ween p e- aining and pos - aining,
and ≈66.67% in be ween p e- aining and ollow-up). F om pos - aining o ollow-up, howe e , i was
obse ed an inc ease (≈33.33%) o he numbe o FoG episodes occu ing a e 7 days wi hou ecei ing
cues.
Fig. 19. G aphics o gai ask esponsible o FoG (Gai ini ia ion, Tu ning, o Passing h ough doo ways) along he sessions
imeline and espec i e o al pe cen age, bo h o pa ien P2.
I is impo an , o analyse also which asks lead o eezing (Fig. 19). In p e- aining he main
FoG-inducing ask was u ning du ing he walking cou se (≈100% o FoG a session 0). O e all, coun ing
all sessions, u ning was he main FoG-inducing ask (≈53% o all FoG episodes), ollowed by passing
h ough doo ways (≈30% o all FoG episodes), and las ly gai ini ia ion (≈17% o all FoG episodes).
Howe e , a ollow-up, he only FoG episode no iced occu ed du ing gai ini ia ion a he s a o he i s
moni o ing ial. The maximum numbe o FoG episodes obse ed du ing ials occu ed in session 3,
consis ing o a o al o 19 FoG episodes (≈5.26% du ing gai ini ia ion, ≈63.16% du ing u ning, and
≈31.58% while passing h ough doo s). As o P2 pe cep ion o FoG in-be ween sessions, he maximum
numbe o FoG episodes epo ed we e 17 episodes a ollow-up, a e 7 days wi hou ecei ing cueing.
Howe e , as a as conside ing he a io be ween numbe o episodes epo ed and numbe o days
passed, he maximum numbe o episodes be ween sessions was epo ed in be ween session 5 and 6
(≈4 a io), epo ing 16 FoG episodes along 4 days wi hou cueing.
83
Fig. 20. G aphics o NFOG-Q sco es along each ime-poin e alua ion, o pa ien s P1 and P2.
When analysing he NFOG-Q answe s along he p o ocol ime-poin s (Fig. 20), o e all pa ien P2
epo ed highe sco es han pa ien P1. Pa ien P1 along he p o ocol ime poin s p esen ed inc easingly
wo se esul s o NFOG-Q, p esen ing a maximum sco e o 10 poin s a bo h pos - aining and ollow-up.
As o pa ien P2, he esul s a ied along he ime poin s, being no iced wo s sco e esul s a mid-
aining, wi h 19 poin s. A bo h p e- aining and pos - aining P2 ob ained he same minimum NFOG-Q
sco e o 15 poin s, howe e a ollow-up he sco e inc eased by one poin .
6.2.2. CLINICAL SCALES
Table 14, Fig. 21 and Fig. 22, show he able and g aphics ob ained o clinical scales esul s.
Table 14. H&Y sco e e olu ion be o e and a e using he FoG p e en ion cueing, o pa ien s P1 and P2
S age
Desc ip ion
Impai men
P e- aining
Pos -
aining
Follow-
up
0
Wi hou any signs o he disease.
1
Unila e al impa imen .
Mild o
mode a e
1.5
Axial and unila e al impai men .
P1
P1
2
Bila e al disease wi hou balance de ici .
P1
P2
P2
2.5
Mild bila e al disease, wi h eco e y in he
'push es '.
P2
3
Mild o mode a e bila e al disease; some
pos u al ins abili y; abili y o li e
independen ly.
4
Se e e disabili y, s ill able o walk o s and
wi hou assis ance.
Se e e
84
5
Con ined o bed o a wheelchai unless
assis ed.
When conside ing he H&Y esul s (Table 14) i is possible o obse e ha o bo h pa ien s, when
compa ing he momen s be o e (p e- aining) and a e (pos - aining) ecei ing he cued aining wi h he
FoG p e en ion s a egy, bo h pa ien s imp o ed hei mo o impai men s. These sco e imp o emen s
we e main ained a ollow-up. Bo h pa ien s showed an imp o emen o 0.5 sco e poin s, howe e , P2
p esen ed s ill highe H&Y s age and, he e o e wo se PD symp om se e i y when compa ed o P1.
Bo h MDS-UPDRS III and PDQ-39 g aphics (Fig. 21 and Fig. 22, espec i ely) showed an o e all
simila p o ile, wi h a endency o ake a U-shape o ma .
Fig. 21. UPDRS III sco es e olu ion along each e alua ion imepoin , o pa ien s P1 and P2.
Fig. 22. PDQ-39 sco es e olu ion along each e alua ion imepoin , o pa ien s P1 and P2.
When obse ing he MDS-UPDRS III esul s (Fig. 21) o e all bo h pa ien s dec eased hei sco es,
meaning imp o emen s o mo o unc ion a each e alua ion ime poin when compa ed o p e- aining
(≈50%, ≈35.71% and ≈21.43% o P1, and ≈52.17%, ≈82.61% and ≈78.26% o P2, a mid- aining, pos -
aining and ollow-up, espec i ely). Howe e , bo h pa ien s expe ienced a sligh eg ession owa ds he
end o he ollow-up pe iod (≈22.22% o P1 and ≈25% o P2, eg ession om pos - aining). Pa ien P1
showed maximum imp o emen s o UPDRS III sco es a mid- aining, while pa ien P2 showed maximum

85
imp o emen s a pos - aining. Pa ien P2, al hough no ing a eg ession a ollow-up, p esen ed a be e
MDS-UPDRS III sco e a ollow-up han a mid- aining, explaining he less p ominen U-shape pa e n o
his g aphic. Pa ien P1, al hough showing esul s a ollow-up wo se han mid- aining, hose esul s we e
s ill be e han p e- aining.
As o he PDQ-39 sco e esul s (Fig. 22) o e all bo h pa ien s also expe ienced an imp o emen
in QoL o e he e alua ion ime poin s. When compa ing bo h pa ien s, P1 showed g ea e a ia ions
along each imepoin ’s sco e, while P2 showed a less p ominen imp o emen , wi h a mo e s able pa e n
o e all. P1 achie ed a minimum sco e, meaning maximum imp o emen el o QoL, a mid- aining
(≈16.10% imp o emen om p e- aining). A pos - aining, al hough ha ing a small eg ession om he
mid- aining esul s (≈5.74% eg ession), s ill showed some deg ee o imp o emen when compa ed wi h
he p e- aining PDQ-39 sco e (≈11.29% imp o emen ). A small eg ession owa ds he end o he ollow-
up pe iod ocu ed, bu wi hou e u ning o p e- aining le els. Pa ien P2 howe e showed maximum
imp o emen a pos - aining (≈19.45% imp o emen om p e- aining), imp o ing g adually om p e-
aining o mid- aining (≈15.28% imp o emen ) and con inuing o imp o e om mid- aining o pos -
aining (≈4.92% imp o emen ) eaching i s peak a his imepoin . A ollow-up, o bo h pa ien s al hough
being no iced a eg ession o he esul s since pos - aining (≈7.30% eg ession o P1 and ≈8.64%
eg ession o P2), QoL imp o ed since p e- aining (≈4.81% imp o emen o P1 and ≈12.49%
imp o emen o P2). O e all, pa ien P2 al hough showing a mo e s able pa e n, p esen ed highe
imp o emen s o QoL om p e- aining o bo h pos - aining and ollow-up.
6.2.3. TRAINING QUESTIONNAIRES
Fig. 23, Fig. 24 and Table 15 p esen he g aphics and able o he aining ques ionnai es
subsec ion.
Fig. 23. NASA TLX sco es e olu ion along each aining session, o pa ien s P1 and P2.
86
When analysing he men al and physical wo kload (Fig. 23) a he las aining session bo h
pa ien s P1 and P2 epo ed lowe NASA TLX sco es, he e o e imp o ing along he sessions he men al
e o el du ing he cued aining (imp o emen s o ≈37.5% o pa ien P1 and ≈67.74% o P2). Pa ien
P2 showed o e all be e esul s o NASA TLX sco es along he aining imeline, howe e bo h pa ien s
s a ed he p o ocol conside ing he wo kload a a somewha high le el (39 sco e o P1 and 31 sco e o
P2) bu dec eased i o a medium wo kload a bo h mid and pos - aining (28 and 25 sco es, espec i ely,
o pa ien P1, and 10 sco es a bo h momen s o P2).
Fig. 24. Cue pe cep ion Ques ionnai e answe s o each ques ion (Q1: “How pe cep ible was he lase ?”, Q2: “Wi h wha
equency did you use he lase cues?”, Q3: “How pe cep ible was he ib a ion du ing he ial?”, and Q4: “How pe cep ible
was he ib a ion in he i s 20s o he ial?”) e olu ion along each aining session, o pa ien s P1 and P2.
Cue pe cep ion esul s (Fig. 24) showed some a iabili y ac oss he wo pa ien s. Bo h pa ien s
epo ed high pe cep ion o ib a ion in he i s 20 seconds o ials (Q4), howe e du ing he walking
ials, he ib a ions pe cei ed a igh oe-o s we e ligh ly pe cei ed (Q3). O e all, bo h pa ien s could
pe cei e he lase e y well (Q1), howe e , only used i sligh ly (Q2). Pa ien P1 did no el he need o
use he lase and el sligh ly con used by i . Pa ien P2 on he o he hand epo ed using he lase lines
igh be o e FoG inducing asks, o p e en om eezing. When he numbe o FoG episodes du ing he
sessions s a ed o diminish, his pa ien also s a ed o epo he need o using he lase less.
Table 15. SUS answe s o each indi idual ques ion, o pa ien s P1 and P2
SUS Ques ions
P1
P2
I hink ha I would like o use his sys em equen ly.
4
4
I ound he sys em unnecessa ily complex.
1
2
I hough he sys em was easy o use.
5
5
I hink ha I would need he suppo o a echnical pe son o be able o use his sys em.
1
1
I hink he a ious unc ions o he sys em a e e y well in eg a ed.
4
3
87
I hough he e was oo much inconsis ency in his sys em.
1
3
I would imagine ha mos people would lea n o use his sys em e y quickly.
5
5
I ound he sys em e y cumbe some o use.
1
1
I el e y con iden using he sys em.
4
5
I needed o lea n a lo o hings be o e I could ge going wi h his sys em.
1
1
Like scale: 1 – s ongly disag ee; 2 – disag ee; 3 – indi e en ; 4 – ag ee; 5 – s ongly ag ee.
As o he SUS esul s (Table 15) bo h pa ien s showed o e all exellen usabili y eaching o al
sco es o 92.5 and 85, o pa ien s P1 and P2, espec i ely.
6.2.4. GAIT METRICS
Fi s , he me ics ob ained a p e- aining (baseline) wi h he +sensBand we e compa ed wi h he
ones cap u ed by he Xsens ha se ed as ou g ound u h. I he +sensBand esul ing me ics ell wi hin
he ange o he espec i e Xsens me ics we e deemed accep able, he e o e alida ing hose +sensBand
esul s. Table 16 summa izes he esul ing me ics ob ained using bo h de ices (Xsens and +sensBand)
o each pa ien P1 and P2.
Table 16. Gai me ics ob ained wi h he Xsens s. +sensBand, o each pa ien P1 and P2 a p e- aining
Xsens
+sensBand
Gai me ics
P1
P2
P1
P2
(p e- aining)
Mean±SD
Mean±SD
Mean±SD
Mean±SD
S ep leng h [m]
0.565±0.021
0.484±0.035
0.515±0.015
0.604±0.047
Veloci y [m/s]
0.974±0.009
0.988±0.114
0.950±0.075
1.062±0.089
Cadence [s eps/min]
102.572±0.877
115.035±10.295
112.549±10.931
89.864±8.606
S ep ime [s]
0.585±0.005
0.513±0.032
0.550±0.050
0.597±0.030
S ance phase [%]
75.420±16.937
57.974±8.206
78.909±0.701
61.693±4.425
Swing phase [%]
24.580±16.937
47.050±8.206
21.091±0.701
38.307±4.425
SD S ep leng h [m]
0.140±0.027
0.058±0.036
0.283±0.037
0.426±0.113
SD Veloci y [m/s]
0.234±0.022
0.227±0.026
0.479±0.073
0.489±0.104
SD S ep ime [s]
0.901±0.147
0.113±0.062
0.228±0.155
0.137±0.022
AS S ep leng h [m]
0.059±0.032
0.087±0.061
0.242±0.207
0.576±0.130
AS Veloci y [m/s]
0.128±0.122
0.347±0.010
0.216±0.177
0.133±0.083
AS S ep ime [s]
0.038±0.012
0.148±0.107
0.167±0.110
0.177±0.051
Bold highligh s he +sensBand me ics alida ed when compa ed wi h he Xsens me ics.
88
O e all, eloci y, s ance phase and swing phase we e he only me ics alida ed o bo h pa ien s.
Pa ien P1, howe e , p esen ed bo h cadence, AS s ep leng h and AS eloci y wi hin he espec i e XSens
me ics ange, he e o e hose +sensBand me ics we e also alida ed o his pa ien . As o pa ien P2,
only SD and AS o s ep ime we e u he alida ed.
Las ly, Fig. 25, Fig. 26, Fig. 27, Fig. 28 and Fig. 29, p esen he g aphics o he gai me ics
esul s. When analysing he +sensBand gai me ics g aphics along he p o ocol e alua ion imepoin s, i
is possible o no e di e en pa e ns along some me ics bu also along each indi idual pa ien .
Fig. 25. Mean s ep leng h, espec i e SD and AS e olu ion along p o ocol e alua ion imepoin s, o each pa ien .
Fig. 26. Mean eloci y, espec i e SD and AS e olu ion along p o ocol e alua ion imepoin s, o each pa ien .
S ep leng h (Fig. 25) and eloci y (Fig. 26) we e he me ics ha showed simila p o iles o bo h
pa ien s, imp o ing g adually as expec ed along he aining sessions and eaching he max imp o emen
peak a pos - aining (≈17.13% and ≈10.85% o pa ien P1, and ≈20.39% and ≈22.89% o pa ien P2,
imp o emen s o s ep leng h and eloci y, espec i ely, om p e- aining o ollow-up). A ollow-up, he
esul s showed a sligh eg ession o bo h pa ien s and bo h me ics, howe e , we e s ill main ained
imp o emen s when compa ed o he p e- aining assessmen .
95
al hough being ex ensi e was e y well ecei ed, being pe cei ed wi h excellen usabili y and medium
wo kload.
Pa ien s’ mo o pe o mance was also imp o ed a pos - aining, especially when conside ing
imp o emen s o pace (s ep leng h and eloci y) bu also hy hm (s ep ime), ein o cing imp o emen s
o he hypokine ic and b adykine ic aspec s o gai .

96
7. CONCLUSIONS
This chap e p o ides inal e lec ions and esea ch indings, highligh ing he wo k con ibu ion
o knowledge. An o e iew o he esul s conside ed wi hin each chap e is p esen ed, and esea ch
ques ions a e answe ed. Fu u e s eps o be conside ed a e also discussed, enhancing he impac and
u he po en ial o his p ojec esea ch.
This disse a ion elied on he +sensBand sys em p ojec , aiming o imp o e i s s a egies,
de eloping and alida ing a new one o mi iga ing FoG, and es ing i s long- e m use po en ial o mo o
and QoL imp o emen s. The wo k de eloped inno a ed by o e coming some o he p e ious
challenges o he +sensBand esea ch p ojec , concluding esea ch on he e iciency o gai e en -
d i en cueing and con inuous eedback in PD pa ien s, p o iding a no el s a egy sui able o
mo e ad anced s ages o PD and pe o ming i s alida ion in a longi udinal p o ocol o allow insigh s on
a ehabili a i e pe spec i e o he +sensBand sys em. This app oach was based on ga he ing and
analyzing, bo h pa ien s’ eedback and mo o da a, wi h he aim o enhance pa ien s' mo o pe o mance
and educe mo o symp oms o PD du ing daily ac i i ies, he e o e imp o ing hei QoL. The esea ch
e o s, along wi h he encou aging esul s achie ed h ough expe imen al me hodologies wi h end-use s,
played a c ucial ole in achie ing he disse a ion goal, as ou lined below.
The ga he ed knowledge in Chap e 2, was he mo o o he p ojec de elopmen , allowing an
unde s anding o he di e en a ailable cueing solu ions being applied in PD pa ien s, as well as how
we e being applied and hei espec i e e ec s (KPI1). I was no iced ha cueing s a egies, pa icula ly
soma osenso y and isual s a egies, show p omising esul s in educing PD mo o symp oms such as
FoG. Closed-loop pa ien - ailo ed app oaches highligh ed hei po en ial o signi ican ly imp o e mo o ask
pe o mance. The li e a u e e iew also allowed o unde s and ha esea ch should p io i ize
spa io empo al ou comes, such as eloci y and FoG occu ences, while inco po a ing egula ,
consis en aining p o ocols and assessmen s o ensu e sus ained bene i s wi h long- e m cued
aining. Technological sys ems, along wi h he +sensBand solu ion componen s, conside ing he new
implemen ed and de eloped s a egy, we e in oduced and desc ibed in Chap e 3: (1) he ha dwa e,
conside ing bo h +sensBand and Xsens sys ems; (2) he cueing and moni o ing +sensBand s a egies
used; and (3) he clinical desk op App imp o ed o gai da a p ocessing.
The +sensBand allows wi h a single de ice o p o ide se e al cueing s a egies, h ough lase line
p ojec ion and/o ib o ac ile mo o s, while simul aneously collec ing pa ien s gai da a h ough an IMU-
97
based sys em. The Xsens sys em, by using high- ech IMU-based gai assessmen s, ac s as a g ound u h
o alida e he +sensBand gai assessmen s.
The +sensBand con igu a ion is es ablished ia a sma phone app, which allows o selec he
espec i e s a egy p e ended. I i ’s in ended o only moni o pa ien s gai wi hou p o iding cueing, he
mo ion moni o ing s a egy allows o ac i a e only he IMU senso s, collec ing pa ien s' angula eloci y
and accele a ion da a, and enabling eal- ime gai e en de ec ion. As o he cueing s a egies, h ee
modali ies we e conside ed. The gai e en -d i en cueing s a egy consis s on p o iding closed-loop
ib a o y cueing a igh - oo oe-o s. The con inuous eedback, al e na i ely, p o ides open-loop
ib a o y cueing, se o 2s on-and-o pulses. By conside ing ha combining soma osenso y closed-loop
cueing and isual open-loop cueing may a ge FoG and enhance gai pe o mance by ins uc ing when
and whe e o place he oo , espec i ely, he FoG p e en ion cueing was de eloped. Simila ly o he
gai e en -d i en cueing, i allows o p o ide ib a ions a igh - oo oe-o s, while simul aneously p o iding
open-loop lase line p ojec ion se o pa ien s' s ep leng h. Since gai ini ia ion was iden i ied as one o he
main FoG-inducing asks, he s a egy was adap ed o p o ide open-loop ib a ion as he con inuous
eedback s a egy un il he i s igh oo oe-o is de ec ed. A e de ec ing his e en i au oma ically
swi ches o he closed-loop mode.
I was also p esen ed he clinical desk op app, de eloped in Ma lab, ha allows o ans o m
he acqui ed da a in o he espec i e gai me ics being analyzed. The accele a ion and angula eloci y
da a un h ough calib a ion, il e ing and gai segmen a ion p ocesses (Fini e S a e Machine algo i hm)
o ob ain a lis o ICs and FCs o each oo . By using hese IC/FCs, gai me ics a e calcula ed o bo h
pace and hy hm domains, bu also a iabili y and asymme y me ics.
The ollowing h ee chap e s comp ise each o he h ee s udies pe o med along his disse a ion,
ega ding he espec i e p o ocol es ablished, esul s ob ained and conclusions.
Chap e 4 p esen ed he on-going c oss-sec ional s udy execu ed in which was accomplished
bo h he on- ield amilia iza ion wi h he +sensBand echnologies bu also he sample size goal o comple e
he s udy (KPI2). This s udy also allowed o unde s and he e icacy o bo h gai e en -d i en cueing and
con inuous eedback s a egies in imp o ing mo o pe o mance, being bo h capable o imp o ing bo h
pace and hy hm aspec s o gai . O e all, he e is po en ial o using bo h s a egies combined, ein o cing
he need o de elopmen o he FoG p e en ion cueing s a egy.
Chap e 5 ollowed wi h he execu ion o a usabili y and wo kload s udy, pe o med in heal hy
indi iduals, o alida ion o he FoG p e en ion cueing s a egy in di e en daily walking scena ios. F om
98
his s udy, i was concluded ha he s a egy was pe cei ed wi h excellen usabili y and low wo kload,
he e o e g an ing ha i should no o e s imula e pa ien s (KPI3).
A e he alida ion o he FoG p e en ion cueing s a egy in heal hy pa icipan s (Chap e 5),
Chap e 6 ollowed wi h he design and execu ion o he longi udinal s udy o add ess he long- e m
ehabili a i e po en ial o he +sensBand di ec ly wi h he end-use s. The esul s, al hough showing some
di e ences due o di e en pa ien cha ac e is ics, p o ed he po en ial o he FoG p e en ion cueing in
mi iga ing FoG du ing aining (KPI4). I s e ec i eness, howe e , depends on consis en use.
Imp o emen s in mo o pe o mance and quali y o li e we e also no ed, as long as con i ming he
excellen usabili y o he s a egy and mid-wo kload pe cei ed (KPI5). Fu u e wo k should conside
a ge ing a la ge numbe o FoG pa ien s o pa icipa e in he s udy, o allow g ea e impac o
spa io empo al esul s.
Along wi h his disse a ion documen , he wo k de eloped esul ed in 2 pape s, a jou nal pape
and a con e ence pape (KPI6).
In summa y, he esul s ob ained along his disse a ion expe imen al assessmen s con i med
he e ec i eness o he +sensBand sys em usage in imp o ing PD pa ien s mo o pe o mance, educing
mo o symp oms and ul ima ely eaching be e QoL, a bo h sho and long- e m uses. I was al eady
possible o pu he +sensBand o es in home-based scena ios and also a a longi udinal p o ocol se ing.
The indings ob ained should se e as mo i a ion o explo e new app oaches such as home-based long-
e m usage. Some adjus men s mainly o design should s ill be conside ed. Howe e , o e all, his
disse a ion esea ch p ojec made a signi ican and g a e ul con ibu ion o enhancing he daily li es o
PD pa ien s h ough he applica ion o biomedical enginee ing solu ions.
7.1. RESEARCH QUESTIONS
The esul ing disse a ion wo k p esen ed allowed o answe he RQs ou lined in Chap e 1. The
concluding answe s a e p esen ed in summa y as ollowed.
RQ1: How ha e cueing s a egies been used and applied o a ge ing mo o symp oms
o PD?
This RQ was answe ed in Chap e 2, by e ising he li e a u e a ailable abou di e en cueing
applica ions designed o a ge se e al mo o symp oms o PD. Cueing s a egies applied o PD pa ien s
p ima ily in ol e ex e nal s imuli such as isual, audi o y, and soma osenso y cues, deli e ed h ough
elec onic and non-elec onic me hods. Visual cues, like lase lines, a e mainly used o imp o e spa ial
99
awa eness, a ge ing mainly he pace aspec s o gai (imp o e hypokine ic impai men s), while audi o y
cues, like me onomes, enhance hy hm du ing gai (imp o e b adykine ic impai men s). Soma osenso y
cues, o en ib a o y, a e less esea ched and mos ly used in combina ion wi h o he s imuli. Cueing can
be open-loop, p o iding he cue a a ixed hy hm, o closed-loop, synch onized wi h pa ien s’ gai
pe o mance o a mo e ailo ed app oach. S udies ha e ocused on using cueing du ing se e al walking
asks, o en inco po a ing FoG igge s, o e alua e cueing e icacy using es ablished clinical scales and
collec ing gai me ics.
RQ2: Which is he e ec i eness o cueing s a egies conce ning mo o e ec s and QoL?
Chap e s 2 and 4, h ough he li e a u e e iew execu ed and he c oss-sec ional s udy
de eloped wi h PD pa ien s es ing di e en cueing s a egies, allowed o unde s and hei e icacy in
imp o ing bo h mo o symp oms and QoL. Conside ing he di e se ange o cueing op ions, cueing
s a egies ha e shown a ied e ec i eness in add essing mo o symp oms and imp o ing QoL in PD.
Sho - e m s udies (c oss-sec ional s udies) indica ed ha isual and audi o y cues signi ican ly imp o e
gai pa ame e s, a ge ing mainly pace aspec s such as cadence, eloci y, and s ep leng h, and can also
educe he numbe FoG episodes in mo e ad anced s ages o he disease. Soma osenso y cues, on
ano he hand, also show po en ial, bu hei e ec i eness is less consis en . Combining di e en cue
ypes, o example, isual and soma osenso y cues, can enhance bo h mo o and QoL ou comes bu may
equi e a en ion in which conce ns a oiding o e s imula ion. Long- e m s udies (longi udinal s udies)
sugges ha while ini ial imp o emen s in mo o unc ion and QoL occu , hese e ec s may diminish o e
ime wi hou a con inuous use o cueing. Closed-loop cueing appea ed o p o ide be e long- e m bene i s
han open-loop me hods, mainly in a ge ing FoG, imp o ing pace (s ep leng h and eloci y) and some
aspec s o hy hm (s ance phase). Mo eo e , when compa ing he use o open and closed-loop
soma osenso y cueing s a egies in a c oss-sec ional s udy execu ed, i was p o en ha bo h cueing
deli e y me hods a e e icien in imp o ing gai pe o mance, howe e di e en aspec s may be imp o ed
sligh ly di e en ly wi h each me hod. Open loop cueing al hough showing g ea e o e all imp o emen s
in pace, con ibu ed o highe gai a iabili y and asymme y. In con as , closed-loop cueing led o mo e
sub le, s able imp o emen s by aligning wi h he pa ien ’s na u al hy hm, po en ially p ese ing gai
a iabili y.
RQ3: How does +sensBand cueing sys em long- e m use a ec Pa kinson’s mo o
symp oms and QoL?
The inal s udy chap e , Chap e 6, by add essing a longi udinal p o ocol in PD pa ien s, allowed
o answe his ques ion conside ing he esul s ob ained. The p omising indings suppo he po en ial o
100
+sensBand cueing s a egies, in he FoG p e en ion cueing o m, in a ge ing mo e ad anced symp oms
o PD, such as FoG (mi iga ing FoG ocu ences o ally du ing aining and educing up o 66.67% a e a
week wi hou ecei ing cued aining) , imp o ing mainly hypokine ic aspec s o gai (≈17.82%)., and
imp o ing o e all pe cei ed QoL (≈15.37%). Howe e , he e ec i eness o FoG mi iga ion depends also
on he egula i y o aining using he FoG p e en ion cueing s a egy, since he majo i y o he
imp o emen s we e no main ained a ollow-up, meaning ha he s a egy caused mainly immedia e
imp o emen s du ing aining ha we e no main ained a long- e m. The imp o emen s obse ed,
especially a pos - aining, demons a e ha senso y cueing can be a aluable ool o mi iga ing FoG, bu
i s applica ion mus be con inuous. These indings, howe e , a e no su icien o gua an ee he use o
his senso y solu ion as a ehabili a i e de ice ins ead o an assis i e de ice o mo o ehabili a ion,
he e o e needing mo e e idence o conclude abou hese aspec s.
7.2. FUTURE DIRECTIONS
Some limi a ions no iced along his disse a ion wo k no iced consis ed o : (i) high a iabili y o
PD pa ien s, ha may a ec g oup and o e all esul s; (ii) design limi a ions, ega ding he bulkiness o
he de ice; and (iii) small sample o PD pa ien s a ailable o pa icipa e in longi udinal s udies.
These aspec s should no disca d, howe e , he po en ial o u he in es iga ion wi h he de ice.
Fu he mo e, as pa ien s o en epo ed no pe cei ing he ib o ac ile cues du ing walking, u u e wo k
should explo e al e na i e +sensBand cueing s a egies, such as subliminal cues, o imp o e PD mo o
symp oms, as hese may be mo e seamlessly in eg a ed in o pa ien s' daily ou ine.
Las ly, he algo i hm o gai -e en de ec ion p o ed o also ha e some limi a ions ega ding
pa ien s gai a iabili y due o he unp edic able na u e o he disease. The e o e, o imp o e eal- ime
gai -e en de ec ion and gai me ics calcula ion pe o mance, he implemen a ion o a machine lea ning
algo i hm able o adap o di e en gai pa e ns should be conside ed ins ead o he de eloped FSM
algo i hm o IC/FC segmen a ion.

101
8. REFERENCES
[1] C. A. Da ie, “A e iew o Pa kinson’s disease”, doi: 10.1093/bmb/ldn013.
[2] B. Thomas and M. Flin Beal, “Pa kinson’s disease”, doi: 10.1093/hmg/ddm159.
[3] M. Delgado‐Al a ado, M. Ma ano, A. San u ún, A. U iaga‐Gallano, D. To desillas‐Gu ie ez, and
J. In an e, “Nonpha macological, Nonsu gical T ea men s o F eezing o Gai in Pa kinson’s
Disease: A Sys ema ic Re iew,”
Mo emen Diso de s
, ol. 35, no. 2, pp. 204–214, Feb. 2020,
doi: 10.1002/mds.27913.
[4] D. Bega, P. Gonzalez-La api, C. Zadiko , and T. Simuni, “A Re iew o he Clinical E idence o
Complemen a y and Al e na i e The apies in Pa kinson’s Disease,”
Cu T ea Op ions Neu ol
,
ol. 16, no. 10, p. 314, Oc . 2014, doi: 10.1007/s11940-014-0314-5.
[5] A. Nieuwboe , “Cueing o eezing o gai in pa ien s wi h Pa kinson’s disease: A ehabili a ion
pe spec i e,”
Mo emen Diso de s
, ol. 23, no. SUPPL. 2. 2008. doi: 10.1002/mds.21978.
[6] L. Wang
e al.
, “E ec s o Rhy hmic Audi o y S imula ion on Gai and Mo o Func ion in Pa kinson’s
Disease: A Sys ema ic Re iew and Me a-Analysis o Clinical Randomized Con olled S udies,”
F on
Neu ol
, ol. 13, 2022, doi: 10.3389/ neu .2022.818559.
[7] H. R. Gonçal es, A. M. Rod igues, and C. P. San os, “Vib o ac ile bio eedback de ices in
Pa kinson’s disease: a na a i e e iew”, doi: 10.1007/s11517-021-02365-3/Published.
[8] S. G. Reich and J. M. Sa i , “Pa kinson’s Disease,”
Medical Clinics o No h Ame ica
, ol. 103,
no. 2, pp. 337–350, Ma . 2019, doi: 10.1016/j.mcna.2018.10.014.
[9] S. Von Campenhausen
e al.
, “P e alence and incidence o Pa kinson’s disease in Eu ope,”
Eu opean Neu opsychopha macology
, ol. 15, no. 4, pp. 473–490, 2005, doi:
10.1016/j.eu oneu o.2005.04.007.
[10] N. Giladi and A. Nieuwboe , “Unde s anding and ea ing eezing o gai in pa kinsonism,
p oposed wo king de ini ion, and se ing he s age,”
Mo emen Diso de s
, ol. 23, no. S2, pp.
S423–S425, Jul. 2008, doi: 10.1002/mds.21927.
[11] A. Nieuwboe and N. Giladi, “Cha ac e izing eezing o gai in Pa kinson’s disease: Models o an
episodic phenomenon,”
Mo emen Diso de s
, ol. 28, no. 11. pp. 1509–1519, Sep. 15, 2013.
doi: 10.1002/mds.25683.
[12] J. Vandenbossche
e al.
, “F eezing o gai in Pa kinson’s disease: dis u bances in au oma ici y
and con ol,”
F on Hum Neu osci
, ol. 6, 2013, doi: 10.3389/ nhum.2012.00356.
102
[13] A. Sch ag, M. Jahanshahi, and N. Quinn, “How does Pa kinson’s disease a ec quali y o li e? A
compa ison wi h quali y o li e in he gene al popula ion,”
Mo emen Diso de s
, ol. 15, no. 6, pp.
1112–1118, No . 2000, doi: 10.1002/1531-8257(200011)15:6<1112::AID-
MDS1008>3.0.CO;2-A.
[14] L. M. Ambani and M. H. Van Woe , “S a Hesi a ion — A Side E ec o Long-Te m Le odopa
The apy,”
New England Jou nal o Medicine
, ol. 288, no. 21, pp. 1113–1115, May 1973, doi:
10.1056/NEJM197305242882108.
[15] H. Khalil and J. A. Bajwa, “Ba ie s and Facili a o s in Physical Rehabili a ion o Pa kinson’s
Disease in he A abian Wo ld,”
Mo Diso d Clin P ac
, ol. 2, no. 3, pp. 227–229, Sep. 2015, doi:
10.1002/mdc3.12200.
[16] T. C. Rubins ein, N. Giladi, and J. M. Hausdo , “The powe o cueing o ci cum en dopamine
de ici s: A e iew o physical he apy ea men o gai dis u bances in Pa kinson’s disease,”
Mo emen Diso de s
, ol. 17, no. 6, pp. 1148–1160, No . 2002, doi: 10.1002/mds.10259.
[17] P. Ginis, E. Nackae s, A. Nieuwboe , and E. He emans, “Cueing o people wi h Pa kinson’s
disease wi h eezing o gai : A na a i e e iew o he s a e-o - he-a and no el pe spec i es,”
Ann
Phys Rehabil Med
, ol. 61, no. 6, pp. 407–413, No . 2018, doi:
10.1016/J.REHAB.2017.08.002.
[18] C. Cosen ino
e al.
, “One cue does no i all: A sys ema ic e iew wi h me a-analysis o he
e ec i eness o cueing on eezing o gai in Pa kinson’s disease,”
Neu osci Biobeha Re
, ol.
150, p. 105189, Jul. 2023, doi: 10.1016/J.NEUBIOREV.2023.105189.
[19] H. Gonçal es, R. Mo ei a, A. Rod igues, and C. San os, “Finding Pa ame e s a ound he Abdomen
o a Vib o ac ile Sys em: Heal hy and Pa ien s wi h Pa kinson’s Disease,”
J Med Sys
, ol. 42, no.
11, p. 232, No . 2018, doi: 10.1007/s10916-018-1087-2.
[20] J. R. Lewis, “I em Benchma ks o he Sys em Usabili y Scale,” 2018. [Online]. A ailable:
h ps://www. esea chga e.ne /publica ion/330225055
[21] N. Mu huk ishnan, J. J. Abbas, H. A. Shill, and N. K ishnamu hi, “Cueing Pa adigms o Imp o e
Gai and Pos u e in Pa kinson’s Disease: A Na a i e Re iew,”
Senso s
, ol. 19, no. 24, p. 5468,
Dec. 2019, doi: 10.3390/s19245468.
[22] K. Ho á h
e al.
, “Changes in Quali y o Li e in Pa kinson’s Disease: How La ge Mus They Be o
Be Rele an ?,”
Neu oepidemiology
, ol. 48, no. 1–2, pp. 1–8, 2017, doi: 10.1159/000455863.
103
[23] U. M. Fie zek, S. J. Schulz, K. Ziegle , and A. O. Ceballos-Baumann, “The Minimal Clinically
Rele an Change o he FOG Sco e,”
J Pa kinsons Dis
, ol. 10, no. 1, pp. 325–332, Jan. 2020,
doi: 10.3233/JPD-191783.
[24] M. Pimen a, C. Pinhei o, and C. San os, “Robo ic senso y cueing du ing gai : a usabili y s udy
wi h scena ios o daily li ing,” in
2024 IEEE In e na ional Con e ence on Au onomous Robo
Sys ems and Compe i ions (ICARSC)
, May 2024, pp. 165–170. doi:
10.1109/ICARSC61747.2024.10535942.
[25] T. Zhang, D. Meng, D. Lyu, and B. Fang, “The E icacy o Wea able Cueing De ices on Gai and
Mo o Func ion in Pa kinson Disease: A Sys ema ic Re iew and Me a-analysis o Randomized
Con olled T ials,”
A ch Phys Med Rehabil
, Aug. 2023, doi: 10.1016/j.apm .2023.07.007.
[26] J. Das
e al.
, “Visual Cues o Tu ning in Pa kinson’s Disease,”
Senso s
, ol. 22, no. 18, 2022,
doi: 10.3390/s22186746.
[27] S. S ua , J. Wagne , S. Makeig, and M. Mancini, “B ain Ac i i y Response o Visual Cues o Gai
Impai men in Pa kinson’s Disease: An EEG S udy,”
Neu o ehabil Neu al Repai
, ol. 35, no. 11,
pp. 996–1009, No . 2021, doi: 10.1177/15459683211041317.
[28] Y. Russo, S. S ua , C. Sil a-Ba is a, B. B umbach, G. Vannozzi, and M. Mancini, “Does isual
cueing imp o e gai ini ia ion in people wi h Pa kinson’s disease?,”
Hum Mo Sci
, ol. 84, Aug.
2022, doi: 10.1016/j.humo .2022.102970.
[29] K. Poláko á
e al.
, “3D isual cueing sho ens he double suppo phase o he gai cycle in pa ien s
wi h ad anced Pa kinson’s disease ea ed wi h DBS o he STN,”
PLoS One
, ol. 15, no. 12 12,
Dec. 2020, doi: 10.1371/jou nal.pone.0244676.
[30] E. N. Beck, K. A. Ehgoe z Ma ens, and Q. J. Almeida, “F eezing o Gai in Pa kinson’s Disease:
An O e load P oblem?,”
PLoS One
, ol. 10, no. 12, p. e0144986, Dec. 2015, doi:
10.1371/jou nal.pone.0144986.
[31] L. G aham
e al.
, “Visual Explo a ion While Walking Wi h and Wi hou Visual Cues in Pa kinson’s
Disease: F eeze Ve sus Non-F eeze ,”
Neu o ehabil Neu al Repai
, ol. 37, no. 10, pp. 734–743,
Oc . 2023, doi: 10.1177/15459683231201149.
[32] O. Gál
e al.
, “Pa emen pa e ns can be designed o imp o e gai in Pa kinson’s disease pa ien s,”
Mo emen Diso de s
, ol. 34, no. 12, pp. 1831–1838, Dec. 2019, doi: 10.1002/mds.27831.
[33] C. Ba hel
e al.
, “The lase shoes,”
Neu ology
, ol. 90, no. 2, pp. e164–e171, Jan. 2018, doi:
10.1212/WNL.0000000000004795.
104
[34] C. G. D. W. A. L. S. C. D. G. Liang Tang, “Rhy hmic lase cue is bene icial o imp o ing gai
pe o mance and educing eezing o u ning in Pa kinson’s disease pa ien s wi h eezing o gai ,”
In J Clin Exp Med
, 2017, Accessed: Dec. 07, 2023. [Online]. A ailable:
h ps://www. esea chga e.ne /publica ion/330290372_O iginal_A icle_Rhy hmic_lase _cue_i
s_bene icial_ o _imp o ing_gai _pe o mance_and_ educing_ eezing_o _ u ning_in_Pa kinson
’s_disease_pa ien s_wi h_ eezing_o _gai
[35] J. D. Holmes
e al.
, “The E ec s o Manipula ing Spa ial Loca ion o Visual Cue Placemen on Gai
Among Indi iduals wi h Pa kinson’s Disease: A Pilo S udy,”
Phys Occup The Ge ia
, ol. 33, no.
3, pp. 263–278, Jul. 2015, doi: 10.3109/02703181.2015.1045109.
[36] M. S. B yan , D. H. Rin ala, E. C. Lai, and E. J. P o as, “A pilo s udy: in luence o isual cue colo
on eezing o gai in pe sons wi h Pa kinson’s disease,”
Disabil Rehabil Assis Technol
, ol. 5, no.
6, pp. 456–461, No . 2010, doi: 10.3109/17483107.2010.495815.
[37] L. Bun ing-Pe y, M. Spindle , K. M. Robinson, J. Noo igian, H. J. Cianci, and J. E. Duda, “Lase
ligh isual cueing o eezing o gai in Pa kinson disease:A pilo s udy wi h male pa icipan s,”
The Jou nal o Rehabili a ion Resea ch and De elopmen
, ol. 50, no. 2, p. 223, 2013, doi:
10.1682/JRRD.2011.12.0255.
[38] E. Cubo, C. G. Moo e, S. Leu gans, and C. G. Goe z, “Wheeled and s anda d walke s in
Pa kinson’s disease pa ien s wi h gai eezing,”
Pa kinsonism Rela Diso d
, ol. 10, no. 1, pp. 9–
14, Oc . 2003, doi: 10.1016/S1353-8020(03)00060-9.
[39] C. Ra anasu i anon , K. S isilpa, P. Te msa asab, and P. Ru hi aphong, “Sma ankle b acele -
lase de ice o imp o e gai and de ec eezing o gai in Pa kinsonism pa ien s: a case se ies,”
Assis i e Technology
, ol. 35, no. 5, pp. 417–424, Sep. 2023, doi:
10.1080/10400435.2022.2113179.
[40] E. J. An, W.-S. Sim, S. M. Kim, and J. Y. Kim, “Sui abili y o isual cues o eezing o gai in
pa ien s wi h idiopa hic Pa kinson’s disease: a case–con ol pilo s udy,”
J Neu oeng Rehabil
, ol.
20, no. 1, p. 91, Jul. 2023, doi: 10.1186/s12984-023-01214-8.
[41] D. J. Gee se, B. Coolen, J. J. an Hil en, and M. Roe dink, “Holocue: A Wea able Holog aphic
Cueing Applica ion o Alle ia ing F eezing o Gai in Pa kinson’s Disease,”
F on Neu ol
, ol. 12,
Jan. 2022, doi: 10.3389/ neu .2021.628388.
[42] N. Byl, W. Zhang, S. Coo, and M. Tomizuka, “Clinical impac o gai aining enhanced wi h isual
kinema ic bio eedback: Pa ien s wi h Pa kinson’s disease and pa ien s s able pos s oke.,”
111
ANOVA
S ep leng h [m]
Squa e sum
d
Mean squa e
Z
Sig.
Be ween g oups
.291
2
.145
6.193
.003
In he g oups
3.100
132
.023
To al
3.391
134
Sche e es
S ep leng h [m]
(I) S udy g oup
(J) S udy g oup
Mean di e ence (I-J)
E o
Sig.
Con idence le el 95%
Minimum limi
Maximum limi
1
2
-.023061
.032143
.773
-.10264
.05652
3
-.107813*
.031726
.004
-.18636
-.02927
2
1
.023061
.032143
.773
-.05652
.10264
3
-.084752*
.033449
.044
-.16756
-.00194
3
1
.107813*
.031726
.004
.02927
.18636
2
.084752*
.033449
.044
.00194
.16756
*. A di e ença média é signi ica i a no ní el 0.05.
The dis ibu ion o S ance phase [%] is he same in he
S udy g oup ca ego ies
.057
Re ain null hypo hesis
The dis ibu ion o Swing phase [%] is he same in he
S udy g oup ca ego ies
.044
Rejec null hypo hesis
The dis ibu ion o Veloci y [m/s] is he same in he
S udy g oup ca ego ies
.008
Rejec null hypo hesis
The dis ibu ion o Cadence [s eps/min] is he same
in he S udy g oup ca ego ies
.033
Rejec null hypo hesis
The dis ibu ion o SD s ep leng h [m] is he same in
he S udy g oup ca ego ies
.011
Rejec null hypo hesis
The dis ibu ion o SD s ep ime [s] is he same in he
S udy g oup ca ego ies
.266
Re ain null hypo hesis
The dis ibu ion o SD eloci y [m/s] is he same in
he S udy g oup ca ego ies
.008
Rejec null hypo hesis
The dis ibu ion o AS s ep leng h [m] is he same in
he S udy g oup ca ego ies
.006
Rejec null hypo hesis
The dis ibu ion o AS s ep ime [s] is he same in he
S udy g oup ca ego ies
.027
Rejec null hypo hesis
The dis ibu ion o AS eloci y [m/s] is he same in he
S udy g oup ca ego ies
.158
Re ain null hypo hesis

112
APPENDIX B – Cus om-made cue pe cep ion Ques ionnai e
Ques ions
How pe cep ible was he lase ?
Wi h wha equency did you use he lase cues?
How pe cep ible was he ib a ion du ing he ial?
How pe cep ible was he ib a ion in he i s 20s o he ial?
1 - No pe cep ible a all/did no use a all; 5 – Always pe cep ible/used i always