The e ec o pe cu aneous ansluminal angioplas y o
supe icial emo al a e y on pulse wa e ea u es
Mikko Pel okangasa,∗, Velipekka Suominenb, Dami Vakhi o b, Ja mo Ve hoa,
Janne Ko honenc, Jukka Lekkalaa, An i Vehkaojaa, Niku Oksalab,d
aBioMediTech Ins i u e and Facul y o Biomedical Sciences and Enginee ing, Tampe e
Uni e si y o Technology, Tampe e, Finland. Email add esses: [email p o ec ed].
Pos al add ess: Facul y o Biomedical Sciences and Enginee ing, Tampe e Uni e si y o
Technology, Ko keakoulunka u 3, FI-33720 Tampe e, Finland. el. +3583311511.
bDi ision o Vascula Su ge y, Depa men o Su ge y, Tampe e Uni e si y Hospi al,
Tampe e, Finland.
cDi ision o In e en ional Radiology, Depa men o Radiology, Tampe e Uni e si y
Hospi al, Tampe e, Finland
dFinnish Ca dio ascula Resea ch Cen e Tampe e, Su ge y, Facul y o Medicine and Li e
Sciences, Uni e si y o Tampe e, Tampe e, Finland.
Abs ac
We aimed o analyze he e ec s o pe cu aneous ansluminal angioplas y (PTA)
o he supe icial emo al a e y (SFA) on a e ial pulse wa es (PWs). Al-
oge he 24 subjec s i.e. 48 lowe limbs we e examined including 26 ea ed
lowe limbs ha ing abno mal ankle- o-b achial p essu e index (ABI) (ABI<0.9
o ABI>1.3) and 22 non- ea ed lowe limbs. The measu emen s we e con-
duc ed in p e-, pe i- and pos - ea men phases as well as in ollow-up isi
a e 1 mon h. Bo h ABI and oe p essu es measu ed by s anda d equipmen
we e used as e e ence alues. PW-de i ed pa ame e s include a ios o di e en
peaks o he PW and ime di e ences be ween hem as well as aging index. Bo h
ea ed and non- ea ed limbs we e compa ed in p e- and pos - ea men as well
as ollow-up isi condi ions. The esul s we e e alua ed in e ms o s a is ical
es s, Bland-Al man-plo s, ee-ma ginal mul i a e κ-analysis and mul iple lin-
ea eg ession analysis. PTA was ound o cause small changes o he s udied
PW-de i ed pa ame e s o he ea ed limb which we e obse ed immedia ely
a e he ea men , bu he changes we e mo e p onounced in he ollow-up
∗Co esponding au ho
Email add ess: [email p o ec ed] (Mikko Pel okangas)
P ep in submi ed o Compu e s in Biology and Medicine Ma ch 5, 2018
This is he pos -p in e sion o he a icle, which has been
published in Compu e s in Biology and Medicine 2018, 96, 274-282.
The inal publica ion is a ailable a
h ps://doi.o g/10.1016/j.compbiomed.2018.04.003
isi . In addi ion, we obse ed ha he endo ascula ins umen a ion i sel
does no cause signi ican changes o he PW-de i ed pa ame e s. The esul s
show ha PW-analysis could be a use ul ool o moni o ing he ea men -e ec
o he PTA. Howe e , because he p e- ea men di e ences o he ea ed and
non- ea ed limb we e small, u he s udies wi h subjec s ha ing no a e ial
diseases a e equi ed. The s udy demons a es he po en ial o he PW analysis
in moni o ing ascula abno mali ies.
Keywo ds: A he oscle osis, Elec omechanical senso s,
Pho ople hysmog aphy, Pe iphe al a e ial disease, Pulse wa e measu emen s
1. In oduc ion
A he oscle osis may be p esen as s i ening, s enosis o occlusion o he
a e ies. Pe iphe al a e ial disease (PAD) is a speci ic o m o a he oscle o-
sis a ec ing mainly he lowe limbs. PAD may be asymp oma ic, bu o en i
causes symp oms such as in e mi en claudica ion, o condi ions h ea ening
he i ali y o he limb. These condi ions may be c i ical limb ischemia wi h
es pain and issue de ec s and acu e limb ischemia. PAD is also seen as a isk
ac o o acu e ca dio ascula e en s such as a s oke o myoca dial in a c ion.
PAD is ea ed commonly by means o pe cu aneous ansluminal angioplas y
(PTA) in which he s enosed a e y is dila ed in a minimally in asi e endo ascu-
la ea men p ocedu e. Acco ding o he pas expe ience, he PTA esul s in
he dis up ion o he a he oscle o ic plaque and ini ia es he his ological emod-
eling o he a e ial wall [1]. In some cases, inwa d emodeling and a es enosis
occu , esul ing in he wo sening o he symp oms. The pa ency a es o he
PTA a e epo ed o a y be ween 75 %–97 % and 60 %–84 % o one- and
wo-yea ollow-up imes, espec i ely[2]. Commonly, he pa ien s a e subjec ed
o a 1-mon h ollow-up o ensu e he echnical success o he ea men . The
e-examina ion includes he measu emen o he ankle- o-b achial p essu e in-
dex (ABI). Howe e , ABI has se e al limi a ions, such as a ying sensi i i y
and speci ici y om s udy- o-s udy [3] and challenges especially wi h he dia-
2
be ic and mediascle o ic pa ien s [4, 5, 6]. As he ABI measu emen has se e al
d awbacks, imaging me hods such as magne ic esonance angiog aphy (MRA)
o X- ay con as -agen based angiog aphy, could be al e na i es. S ill, MRA
has high cos s and X- ay angiog aphy causes exposu e o adia ion. Sys olic oe
p essu es o oe- o-b achial p essu e index (TBI) ha e also been sugges ed as an
al e na i e o ABI especially o diabe ic pa ien s, bu also hey ha e p oblems
wi h he eliabili y [6]. Many s udies ha e sugges ed pulse wa e (PW) measu e-
men and analysis o inding ascula abno mali ies o abno mal ascula aging
[7, 8, 9, 10, 11, 12].
Ea lie , we ha e s udied he epea abili y o he PW-measu emen s and i
he e exis di e ences in he PW-mo phologies be ween di e en -aged subjec s
as well as be ween heal hy subjec s and a he oscle o ic pa ien s in he signals
eco ded om a ious loca ions [11]. The expe imen al se -up o hese s udies
limi s he esul s only o p o ide in o ma ion on he o e all condi ion o he
a e ial ee wi hou da a on he e ec o he s enosis i sel . In his s udy, we
es he sui abili y o he PW measu emen and analysis o he moni o ing o a
supe icial emo al a e y (SFA) s enosis ea ed by PTA and p o ide in o ma-
ion on he e ec s o PTA on PW. We hypo hesize ha PTA causes quan i a i e
changes in he obse ed PW and ha he e ec o he ea men is obse ed im-
media ely a e he ea men and mo e clea ly in he ollow-up isi wi hin one
mon h due o emodeling ini ia ed by he PTA. We also es i he s udied PW-
analysis me hods a e sui able o diagnos ic use, e.g. i he e a e p e- ea men
di e ences in he PW- ea u es be ween he ea ed and non- ea ed lowe limb.
A bene i o he PW measu emen compa ed wi h he s a ic ABI measu emen
o angiog aphs is ha i p o ides in o ma ion also on he ascula dynamics.
2. Ma e ials and me hods
2.1. Measu emen ha dwa e and senso placemen
The measu emen da a was collec ed in supine posi ion by using elec e
senso s made o elec omechanical ilm (EMFi) (Em i S-se ies, Em i L d, Fin-
3
land) and PPG p obes (S0010A, Shenzhen Med-link, China) connec ed o a
wi eless body senso ne wo k (WBSN) [13]. The sampling equency was 250
Hz o he EMFi signals and 500 Hz o he PPG signals. The EMFi signals we e
in e pola ed o ha e a sampling equency o 500 Hz in u he signal p ocessing.
EMFi-senso s, which a e sensi i e only o dynamic bu no o s a ic loading,
we e placed a he w is on he op o he adial a e y, and bo h ankles on he
op o he pos e io ibial a e y in o de o eco d dynamic p essu e PW signals,
i.e. a signal p opo ional o he a ying AC-componen o pulse p essu e o he
blood p essu e signal.
T ansmission mode PPG p obes ha ing an exci a ion wa eleng h o 905 nm
we e placed on index inge and bo h 2nd oes o eco ding blood olume PW
signals. In addi ion o he PW-signal, bipola ECG (elec oca diog am) was
also measu ed by using con en ional sil e -sil e chlo ide (Ag/Cl) elec odes
placed unde he cla icles and an ECG-de ice compa ible wi h he WBSN [13].
The ECG was u ilized in he ex ac ion o he PWs.
The measu emen s we e conduc ed du ing he p e-, pe i-, and pos - ea men
phases o he PTA and du ing he ollow-up isi a e one mon h (median 33
days, in e -qua ile ange (IQR) 30–36 days). Fo pe i-p ocedu al da a collec-
ion, he senso s we e placed be o e he no mal p epa a ion o he PTA and we e
emo ed wi hin 5–10 minu es a e he PTA. The senso placemen was simila
in he ollow-up isi , bu he du a ion o he measu emen was 5 minu es and
was done be o e he ABI and oe p essu e measu emen s.
The ankle PW-signals we e excluded om he analysis since hey we e dis-
u bed hea ily especially du ing he ea men mainly o wo easons: Fi s ,
he PAD pa ien s commonly ha e ex emely low-ampli ude p essu e PW-signals
and he e o e he measu emen is also e y sensi i e o he co ec posi ioning
o he senso s. Second, he ope a ing able in he PTA oom was na ow, which
o en o ced pa ien ’s lowe limbs in such posi ion ha he s uc u es con aining
EMFi senso s ouched each o he o he ope a ing able, and his caused majo
a i ac s o he signal.
4
2.2. Re e ence alues
The ABI is de ined as a a io o he sys olic blood p essu es measu ed by a
cu om an a m and an ankle, i.e. ABI = Pankle/Pb achial. ABI was compu ed
by di iding he highes ankle p essu e (ei he ADP (a e ia do salis pedis) o
ATP (a e ia ibialis pos e io )) by he highes a m p essu e (le o igh ).
Toe p essu es e e o he sys olic p essu es measu ed om he hallux. Bo h
e e ence alues we e collec ed by using an au oma ed measu emen de ice,
Falcon P o (Viasonix, Is ael). P e- ea men ( isi a polyclinics be o e he
PTA) and ollow-up isi ABI and oe p essu e measu emen s we e conduc ed
as a pa o no mal clinical p ac ice and used as he e e ence alues in he
s udy.
2.3. S udy subjec s
The inclusion c i e ia o he s udy we e he abno mal ABI eading, i.e.
ABI<0.9 o ABI>1.3, ele an symp oms, s enosis in SFA based on mag-
ne ic esonance imaging angiog aphy and a pa ien conside ed as a candida e
o he PTA o he SFA. A pacemake and a possible isk ha he s udy in-
e e es he pa ien ’s ea men p ocess we e conside ed as exclusion c i e ia.
Al oge he 27 olun ee pa ien s unde going PTA o he SFA we e ec ui ed
o he s udy. Th ee o hem we e excluded o he ollowing easons: esea ch
pe sonnel una ailable, a pa ien was ound o be unsui able o PTA du ing
W is :
EMFi
Index inge : PPG
Ankles:
EMFi
Second
oes:
PPG
ECG
Figu e 1: Senso placemen .
5

he imaging and ex emely low pe iphe al pe usion. 14 le and 12 igh lowe
limbs we e ea ed, including wo pa ien s wi h bo h lowe limbs ea ed. Mo e
de ailed da a on he s udy subjec s is shown in Table 1. A clea majo i y o he
included pa ien s me he c i e ion ABI<0.9 — only one o he pa ien s had
p e- ea men ABI highe han 1.3. 15 non- ea ed lowe limbs ou o 22 lowe
limbs (68 %) also had he ABI alue ou side he no mal ange o 0.9<ABI<1.3
a leas in one o he p e- ea men o ollow-up isi ABI measu emen s.
Table 1: Da a desc ibing he s udy subjec popula ion including clinical e e ence alues, ABI
and oe p essu e.
Pa ame e Median (IQR)
Mass (kg) 75.5 (67.0...93.0)
Heigh (cm) 170.0 (165.5...176.0)
BMI (kg/m2) 26.7 (24.4...30.0)
Age (yea s) 71.5 (67.5...76.0)
T ea ed lowe limbs: 26
ABI, p e- ea men 0.61 (0.50...0.75) *
ABI, ollow-up 0.96 (0.82...1.04) #
Toe p essu e, p e- ea men (mmHg) 59.0 (42.0...86.0) **
Toe p essu e, ollow-up (mmHg) 94.5 (76.0...122.0) ##
Non- ea ed lowe limbs: 22
ABI, p e- ea men 0.88 (0.75...1.02) *
ABI, ollow-up 0.94 (0.72...1.08) #
Toe p essu e, p e- ea men (mmHg) 75.5 (53.0...117.5) **
Toe p essu e, ollow-up (mmHg) 99.5 (66.0...122.0) ##
Pa ame e Numbe (%)
Males 16 (66.7 %)
Diabe es 11 (45.8 %)
Dyslipidemia 22 (91.7 %)
Rheuma oid a h i is 3 (12.5 %)
Cu en smoke 2 (8.3 %)
Ex-smoke 11 (45.8 %)
p- alues o he p e- ea men ABI and oe p essu e alues be ween ea ed and non- ea ed limb:
∗:p < 0.001, **: p < 0.1, # and ##: no signigican
BMI: Body-mass index; IQR: in e -qua ile ange.
2.4. E hics and pa ien sa e y
The s udy was app o ed by he local e hical e iew boa d o he hospi al
dis ic (R15107), he Finnish Na ional Supe iso y Au ho i y o Heal h and
6
Wel a e (Val i a, ID 309) and he echnical depa men o he hospi al. The
s udy was egis e ed o a public clinical ials egis e (ClinicalT ials.go ID:
NCT02725307). W i en in o med consen s we e ob ained om he olun ee
pa ien s pa icipa ing in he s udy.
2.5. Pulse wa e analysis
The PW-ex ac ion was implemen ed i s by inding R-peaks om he ECG
by an algo i hm p oposed in [14]. A e ha , local mimima ollowing each
de ec ed R-peak we e ex ac ed om he pulse wa e signals. The pulse wa e
signal be ween he successi e local minima was conside ed as a PW candida e.
The alidi y o he PW candida es was es ed by using adap i e h esholds.
A PW candida e was ejec ed i 1) he slope o he ising edge o he PW
candida e was no posi i e o 50 ms o i i s a e age slope was less han 30% o
he maximum slope; 2) he di e ence o he minimum and maximum alue o
he PW-candida e was ou side ce ain h esholds; o 3) ampli ude-no malized
PW candida e was ou side p ede ined limi s o a leas 20% o i s du a ion.
In p e-p ocessing, he signals we e il e ed wi h a Sa i zky-Golay smoo hing
il e ha ing a window leng h o 91 samples (182 ms) and a polynomial o de o
2. In addi ion, he signals we e o wa d-backwa d il e ed wi h a ini e impulse
esponse low-pass il e ha ing a cu o equency o 10 Hz, ansi ion band o
10–12 Hz, pass band ipple o 0.05 dB, and s op band a enua ion o 100 dB
[15], as implemen ed in [16, 11]. A e p e-p ocessing and ea u e ex ac ion, he
PWs we e cha ac e ized by compu ing al oge he 10 di e en PW-pa ame e s.
2.5.1. PW-cu e de i ed ea u es
The ea u es ex ac ed om he PWs we e o iginally p oposed ei he o
adial p essu e PW-analysis o digi al olume (PPG) PW-analysis as e.g. in
[17, 18]. In ea lie s udies [16, 11], we ex ac ed simila ea u es also om
he lowe limb PWs by using algo i hms ha we e o iginally implemen ed o
uppe limb PWs and p oposed in [15, 19]. In his s udy, we analyzed how hese
ea u es, ex ac ed om lowe limb PWs, a e changed as a esul o he PTA o
7
he SFA. Nume ous s udies ha e p esen ed esul s ha he mo phology o bo h
p essu e and olume pulse wa e depends on he s a us o he ascula u e. The
esis ance and compliance o he a e ial pa hway a ec s he wa e p opaga ion
and e lec ion [9], bu he exac physiological o igin and signi icance o he
di e ences equi es u he s udies especially in case o lowe limbs.
Di e en pa ame e s we e compu ed based on he ampli udes and ime di -
e ences be ween he sys olic and dias olic wa es as in [11] and illus a ed in
Fig. 2. When inding he iducial poin s o P1,P2, and B, he incisu a o
dic o ic no ch di iding he indi idual PW in o sys olic and dias olic pa s was
ound as he las ze o-c ossing om nega i e o posi i e o he 1s de i a i e 0
o he PW in he sea ch window limi ed by a poin 80 ms a e he maximum
o he PW and a poin which co esponds o 65% om he o al leng h o he
PW. In case o he absence o his ze o-c ossing, he loca ion o incisu a was
de ec ed a he loca ion o he highes peak o he 2nd de i a i e 00 ound om
he same in e al [16]. Based on hese ea u es, he ollowing pa ame e s we e
compu ed:
•R1as a a io o he ampli ude o dias olic wa e Band he sys olic max-
imum ( he maximum o P1and P2in Fig. 2), i.e. R1=B/ max(P1, P2),
named as e lec ion index o index- inge PPG in [18],
•R2as a a io o he ampli ude o dias olic wa e Band ea ly sys olic wa e
P1, i.e. R2=B/P1
•R3as a a io o he ampli ude o dias olic wa e Band la e sys olic wa e
P2, i.e. R3=B/P2
•R4as a a io o he ampli ude o la e (P2) and ea ly (P1) sys olic wa e, i.e.
R4=P2/P1, named as pe iphe al augmen a ion index o w is p essu e
pulses e.g. in [17],
•T1as a ime di e ence be ween sys olic maximum (max(P1, P2)) and he
peak o he dias olic wa e (B)
8
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
T ea ed lowe limb
Toe PPG (no malized)
0
0.2
0.4
0.6
0.8
1R1=0.61
R2=0.82
R3=0.61
R4=1.33
T1=0.184 s
T2=0.296 s
T3=0.184 s
B
P1
P2R1=0.54
R2=0.61
R3=0.54
R4=1.14
T1=0.228 s
T2=0.310 s
T3=0.228 s
B
P1
P2R1=0.38
R2=0.41
R3=0.38
R4=1.08
T1=0.326 s
T2=0.386 s
T3=0.326 s
B
P1
P2
Be o e ea men A e ea men Follow-up isi
T2
T1,T3
Time (s)
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Non- ea ed lowe limb
Toe PPG (no malized)
0
0.2
0.4
0.6
0.8
1R1=0.35
R2=0.37
R3=0.35
R4=1.07
T1=0.298 s
T2=0.350 s
T3=0.298 s
B
P1
P2R1=0.38
R2=0.41
R3=0.38
R4=1.06
T1=0.300 s
T2=0.346 s
T3=0.300 s
B
P1
P2R1=0.39
R2=0.43
R3=0.39
R4=1.11
T1=0.296 s
T2=0.358 s
T3=0.296 s
B
P1
P2
Figu e 2: Examples o he PWs, loca ions o iducial poin s and ime in e als, and alues o
he de i ed pa ame e s om one es subjec be o e and a e he PTA as well as du ing he
ollow-up isi . Also T1–T3a e illus a ed in he PW shown in he igh uppe co ne . T1and
T3a e equal in all hese examples, since P2is he maximum o he PW in all he examples.
Each pai o he PWs om ea ed and non- ea ed lowe limb a e eco ded concomi an ly.
•T2as a ime di e ence be ween ea ly sys olic peak (P1) and he dias olic
peak (B)
•T3as a ime di e ence be ween la e sys olic peak (P2) and he dias olic
peak (B)
The peaks o ea ly and la e sys olic wa es P1(Fig. 2) and P2as well as dias olic
wa e Bwe e ound based on he me hod desc ibed in [16, 19]. These peaks a e
o en o e lapped in he PWs eco ded om a he oscle o ic pa ien s, bu hey
a e mo e ob ious in adial o ca o id PWs and in lowe limb PWs eco ded om
heal hy subjec s [16, 11]. Fo hese easons, he iducial poin s we e ound by
implemen ing 5 h-o de de i a i e analysis as in [16, 19].
A e inding he bounda y be ween he sys olic and dias olic pa s, he
ollowing p ocedu e was implemen ed o inding P1and P2[16, 19]: I he sign
o he 5 h de i a i e (5) a he poin co esponding o he sys olic maximum
o he PW was
1. posi i e, his poin was conside ed as a poin o he la e sys olic peak P2
9
and ollow-up isi ha e a leas mode a e accep ance wi h he ABI, and he
κ- alues be ween ABI and R1,R3,T1and T2pa ame e s a e e en highe han
he κ- alue be ween ABI and oe p essu e (0.75). All κ- alues be ween he ABI
and PW-de i ed pa ame e s a e highe han co esponding alues be ween oe
p essu es and PW-de i ed pa ame e s.
Table 3: F ee-ma ginal mul i a e κ alues o he changes be ween di e en alues and ABI
and oe p essu e.
Pa ame e ABI Toe p essu e
R10.82 0.62
R20.73 0.52
R30.91 0.71
T10.73 0.62
T20.82 0.71
T30.82 0.71
Aw 0.52 0.30
A 0.55 0.52
Toe p essu e 0.75
3.6. Mul iple linea eg ession
The esul s o mul iple linea eg ession analysis o he changes be ween
p e- ea men si ua ion and ollow-up isi a e shown in Table 4. Acco ding o
he esul s, he changes a e dependen on he ea men o he SFA bu no on
he RR-in e al excep T2. Su p izingly he oe p essu es a e less dependen on
he ea men o he SFA han PW-de i ed indices and ABI.
3.7. P e- ea men epea abili y
Bo h ICC and CV a e shown in Table 5 o bea - o-bea da a in p e- ea men
si ua ion. ICCs close o 0.9 o highe and median CVs less o equal han 2%
we e obse ed. Bo h indica o s show good epea abili y and a e inline wi h he
ea lie s udy [11].
16

1/2#(p ollow-up+pbe o e)
0.3 0.4 0.5 0.6
p ollow-up-pbe o e
-0.4
-0.2
0
0.2
a) R1
1/2#(p ollow-up+pbe o e)
0.3 0.4 0.5 0.6 0.7 0.8
p ollow-up-pbe o e
-0.5
0
0.5
b) R2
1/2#(p ollow-up+pbe o e)
0.3 0.4 0.5 0.6
p ollow-up-pbe o e
-0.4
-0.2
0
0.2
c) R3
1/2#(p ollow-up+pbe o e) (s)
0.25 0.3 0.35 0.4 0.45
p ollow-up-pbe o e (s)
-0.2
0
0.2
) T1
1/2#(p ollow-up+pbe o e) (s)
0.25 0.3 0.35 0.4 0.45 0.5
p ollow-up-pbe o e (s)
-0.1
0
0.1
0.2
0.3
g) T2
1/2#(p ollow-up+pbe o e) (s)
0.15 0.2 0.25 0.3 0.35
p ollow-up-pbe o e (s)
0
0.2
0.4
h) T3
1/2#(p ollow-up+pbe o e)
0.7 0.8 0.9 1 1.1
p ollow-up-pbe o e
-0.4
-0.2
0
0.2
0.4
d) W is /Toe (Aw )
1/2#(p ollow-up+pbe o e)
0.85 0.9 0.95 1 1.05 1.1 1.15
p ollow-up-pbe o e
-0.5
0
0.5
e) Finge /Toe (A )
1/2#(p ollow-up+pbe o e)
0.5 1 1.5 2
p ollow-up-pbe o e
-0.4
-0.2
0
0.2
0.4
0.6
0.8
i) Re e ence alue: ABI
1/2#(p ollow-up-pbe o e) (mmHg)
0 50 100 150 200
p ollow-up-pbe o e (mmHg)
-100
0
100
j) Re e ence alue: Toe p essu e
Non-ope a ed limb
A e age di e ence (non-ope a ed limb)
95% limi s o ag eemen (non-ope a ed limb)
Ope a ed limb
A e age di e ence (ope a ed limb)
Figu e 4: Bland-Al man plo s o he p e- ea men alues and he alues ound in he ollow-
up isi o each PW-pa ame e and he e e ence alues ABI and oe p essu e. The no mal
ange o ABI is also shown in black e ical dashed lines in panel i).
17
Table 4: Mul iple linea eg ession coe icien s and signi icance le els. The ma ke a e pa-
ame e name e e s o p- alue o he F-s a is ics s. cons an model.
Coe icien s
Pa ame e In e cep ∆RR SFA PTA
R1◦-0.047 -0.164 -0.116•
R2•-0.058 -0.257 -0.141•
R3•-0.041 -0.134 -0.133/
T1•0.013 -0.056 0.054/
T2/0.010 -0.235•0.047•
T3◦0.011 -0.129 0.052•
Aw •0.005 0.047 0.086/
A •-0.004 0.123 0.108/
ABI/-0.053 0.490 -0.210/
Toe p essu e -4.731 19.928 -20.727*
∆RR: change in RR-in e al
∗:p < 0.10, ◦:p < 0.05, •:p < 0.025, /:p < 0.01
Table 5: In a-class co ela ions (ICC) and coe icien s o a ia ion (CV) o p e- ea men
measu emen .
Pa ame e ICC (95% con idence limi ) Median CV (1s ...3 d qua ile) (%)
R10.96 (0.93–0.97) 0.6 (0.4...3.0)
R20.96 (0.93–0.97) 0.5 (0.3...3.0)
R30.96 (0.94–0.97) 0.6 (0.4...2.9)
T10.89 (0.84–0.93) 1.5 (0.9...2.7)
T20.89 (0.83–0.92) 1.5 (0.8...3.5)
T30.91 (0.87–0.94) 2.0 (1.2...4.0)
Aw 0.96 (0.94–0.97) 1.9 (1.4...3.2)
A 0.97 (0.96–0.98) 1.3 (1.0...1.8)
4. Discussion
The p esen ed esul s sugges ha PTA caused signi ican changes in he ea-
u es ex ac ed om he PWs. The majo i y (R1–R3,T1–T3,Aw , and A ) o
he s udied pa ame e s showed s a is ically signi ican di e ences be ween p e-
ea men condi ion and a si ua ion a he ollow-up isi . Only R2,T2–T3, and
A showed di e ences be ween he p e- ea men and immedia e pos - ea men
alues, bu o he PW-de i ed pa ame e s s ill had ends owa ds heal hie con-
di ion. The PW-measu emen is able o ind some immedia e changes, bu he
esul s sugges ha he emodeling o he plaque and a e ial wall wi hin he
18
ollow-up pe iod makes he changes mo e isible. The e o e, he beha iou o
he pos - ea men PW-pa ame e s is simila o he pos - ea men ABI as p o-
posed in [23] and especially i he p e- ea men ABI is less han 0.80 [24].
Acco ding o he esul s, he endo ascula ins umen a ion inse ed in o SFA
does no a ec he alues o he pe i-p ocedu al PW-pa ame e s. The indepen-
dence o he PW-pa ame e s on he inse ion o he endo ascula ins umen s
could allow o moni o changes in PW mo phology caused by he PTA in eal
ime, e en hough he immedia e changes could be small.
4.1. PW-analysis and changes caused by PAD
To he ex en o ou knowledge, he p esen ed s udy is he i s s udy which
compa es he alues o lowe -limb PW-de i ed pa ame e s be ween he p e- and
pos - ea men si ua ions o he PTA o he SFA. Howe e , he esul s a e sup-
po ed by o he s udies concen a ing on he diagnosis o he PAD: Yokoyama
e al. [25] ha e epo ed an inc ease in PW eloci y in diabe ic pa ien s due
o a e ial s i ness, bu dec ease in he PW eloci y in a lowe limb ha ing
PAD symp oms likely due o s enoses which dec ease he blood p essu e and
hus he PW eloci y in he diseased lowe limb. A e he success ul PTA,
he PW eloci y inc eased indica ing he changes in he lowe limb ascula u e.
Simila kind o esul s we e ound in his s udy: he alues o he PW-de i ed
pa ame e s change in he heal hie di ec ion as epo ed ea lie in [11].
The PW-analysis can also e eal he bene icial wide-scale sys emic emo e e -
ec s o he PTA: Jacomella e al. [12] epo ed changes owa ds be e condi ion
wi hin 3-mon h ollow-up pe iod o he PTA o he SFA in a ca o id a e y aug-
men a ion index based on a adial a e y applana ion onome y. In he p esen
s udy, he medians o he dis ibu ions o R1–R3in non- ea ed limbs (Table
2) shi ed owa ds heal hie alues obse ed in [11] du ing he 1-mon h ollow-
up pe iod e en hough he pa ien -wise di e ences be ween p e- ea men and
ollow-up isi si ua ions we e close o ze o (Figs. 3h–j). Howe e , he changes
in he IQRs we e no as d ama ic in Table 2. We assume ha a success ul PTA
es o es physical ac i i y, such as walking, also imp o ing he condi ion o he
19
con ala e al lowe limb and esul ing in inc eased pe usion in he lowe limbs.
Simila changes we e also obse ed in he clinical e e ences alues (Table 1)
and Figs. 3d and 3k) which suppo s he assump ion. The a ea a ios Aw
and A as well as ime in e als T1–T3beha e di e en ly: hei dis ibu ions a
p e-PTA and ollow-up isi si ua ions did no di e om each o he in case o
non-ope a ed lowe limbs (Table 2), indica ing weake dependence on posi i e
sys emic emo e e ec s. Fu he s udies should con i m hese assump ions.
The e a e also s udies [7, 8, 9, 10] epo ing bila e al di e ences be ween he
lowe limbs wi h and wi hou PAD. E s e al. [10] ha e epo ed ime di e ences
in he oo poin s ( he beginnings o he PW) be ween he lowe limb wi h and
wi hou he PAD. Allen e al. [7, 8, 9] ha e ound bila e al di e ences be ween
oe-PPG signals in unila e al PAD pa ien s. A compa ison be ween di e en
ime delays be ween he peak o oo poin s o he PWs om di e en loca ions
and be ween he pa ien s and con ol subjec s is also in ou u u e in e es s.
4.2. Limi a ions o he me hod
In p e ious s udy[11], s a is ically signi ican di e ences we e ound o he
s udy g oups consis ing o pa ien s ha ing a he oscle o ic changes and same-
aged heal hy con ol subjec s. The expe imen al se up in he p e ious s udy
[11] was di e en han in his s udy in which he lowe limbs ha unde wen
SFA o he PTA we e compa ed wi h he lowe limbs wi hou he PTA. The
esul s o he p esen s udy do no show signi ican bila e al di e ences be ween
he ea ed limb and he non- ea men limb i he p e- ea men alues o he
PW-de i ed pa ame e s a e s udied. This possibly limi s he diagnos ic use o
he s udied PW-analysis me hods.
Possible eason is ha 68% o non- ea ed lowe limbs p o ided abno mal
ABI eading a leas in one o he wo measu emen s. I a pa ien has he symp-
oms o PAD in one limb, he e is a high likelihood o a he oscle o ic changes
also con ala e ally and elsewhe e in he ascula u e. The s udied PW-de i ed
pa ame e s may also be sensi i e o all kinds o degene a i e a he oscle o ic
changes including he s i ening o he la ge a e ies such as he ao a, no only
20
he s enoses and occlusions which a e he ypical lesions o he PAD. S ill, de-
pending on he me hod, ei he bila e al compa ison o unila e al measu emen
can p o ide he be e pe o mance, e en hough he di e ences a e small as in
[8, 9].
I he ob ious sou ces o e o in he PPG-measu emen , such as inco ec ly
placed PPG-p obe and mo ion a e ac s, a e excluded, he e a e a ew me hod-
ological limi a ions. Poo pe iphe al pe usion dec eases he ampli ude o he
PPG-signal and may p e en he p oposed analysis. The low-ampli ude PPG-
signal can be a ma ke o PAD, bu can also be caused by asocons ic ion
as he esul o low empe a u e. The body posi ion a ec s he shape o he
PPG-signal and is a opic o u he s udies. Howe e , in he p esen s udy,
all he measu emen s we e conduc ed in supine posi ion and limbs ex en ed.
Ligh om he en i onmen can dis u b he PPG-signals, bu in ou measu e-
men s, egula 50 Hz and 100 Hz equency componen s caused by he mains
and he luo escen ligh s we e he mos signi ican sou ces o dis u bance and
we e elimina ed by he il e ing.
Despi e he limi a ions discussed, he esul s sugges ha PTA has an e -
ec on mos o he ea u es ex ac ed and e alua ed in he p esen s udy. The
analysis o he PW- ea u es may p o ide addi ional in o ma ion in inding he
ea ly signs o PAD o in moni o ing he ea men du ing he ollow-up pe iod
especially in he cases in which he con en ional me hods, ABI and oe p es-
su e measu emen s, p o ide in un eliable o con lic ing esul s. A pa o he
ea u es combined wi h o he heal h da a may be u ilized also as he inpu s o
mul i a ia e analysis me hods o machine lea ning which would be a opic o
u he s udies.
5. Conclusions
S a is ically signi ican di e ences we e ound in he alues o he PW-
de i ed ea u es be ween p e- ea men and immedia e pos - ea men condi-
ion. S onge di e ences in he ea ed lowe limb we e ound be ween he
21

p e- ea men alues and he pos - ea men condi ions a e a 1-mon h ollow-
up pe iod. In bo h cases, he changes in he alues o he ea ed limb di e ed
signi ican ly om he co esponding changes ound o he non- ea ed lowe
limb. In case o he non- ea ed lowe limbs, he a ia ions o he di e ences
be ween di e en measu emen e en s we e ela i ely high bu wi hou a consis-
en end owa ds heal hie o wo se condi ion. The endo ascula ins umen-
a ion inse ed in o SFA did no a ec he PW-pa ame e s. The s udy was no
able o show s a is ically signi ican di e ences in he p e- ea men alues o
he PW-de i ed pa ame e s be ween he ea ed and non- ea ed lowe limbs.
This is possibly explained by he s udy sample: 68 % o he non- ea ed limbs
p oduced an abno mal ABI- eading a leas o one o he wo measu emen s.
This indica es ha many pa ien s had subclinical a he oscle o ic changes also in
he o he lowe limb. P io o he eal u iliza ion o he measu emen me hod,
u he compa ison o he heal hy subjec s and PAD-pa ien s and ha ing s an-
da dized measu emen condi ions should con i m he indings.
Con lic o in e es s a emen
The au ho s decla e no con lic o in e es .
Acknowledgmen
We would like o hank all he olun ee es subjec s o hei aluable
con ibu ion. The pe sonnel in Tampe e Uni e si y Hospi al (Depa men o
Vascula Su ge y) a e acknowledged o hei con ibu ion o collec ing he e e -
ence alues. The wo k was unded by he Doc o al P og amme o he P esiden
o Tampe e Uni e si y o Technology, Finnish Funding Agency o Inno a ion
(TEKES) as a pa o p ojec Vi alSens (decision ID 40103/14), and g an s
om Finnish Cul u al Founda ion/Pi kanmaa Regional Fund/Elli and El i Ok-
sanen’s Fund, Tekniikan edis ¨amiss¨a¨a i¨o and Emil Aal onen Founda ion.
22
Re e ences
[1] R. Wy enbach, A. Gallino, M. Ale ci, F. Mahle , L. Cozzi,
M. Di Valen ino, J. J. Badimon, V. Fus e , R. Co i, E ec s o pe cu-
aneous ansluminal angioplas y and endo ascula b achy he apy on as-
cula emodeling o human emo opopli eal a e y by nonin asi e magne ic
esonance imaging, Ci cula ion 110 (9) (2004) 1156–1161.
[2] J. R. Lai d, Limi a ions o pe cu aneous ansluminal angioplas y and
s en ing o he ea men o disease o he supe icial emo al and popli eal
a e ies, Jou nal o Endo ascula The apy 13 (2 suppl) (2006) II–30–II–40.
[3] D. Xu, J. Li, L. Zou, Y. Xu, D. Hu, S. L. Pago o, Y. Ma, Sensi i i y and
speci ici y o he ankle-b achial index o diagnose pe iphe al a e y disease:
a s uc u ed e iew, Vascula Medicine 15 (5) (2010) 361–369.
[4] V. Suominen, T. Ran anen, M. Vene mo, J. Saa inen, J. Salenius, P e a-
lence and isk ac o s o pad among pa ien s wi h ele a ed abi, Eu opean
Jou nal o Vascula and Endo ascula Su ge y 35 (6) (2008) 709–714.
[5] S. Ala ab, H. Fakh zadeh, F. Sha i i, A. Mos ash i, M. Mi a e in, Z. Badam-
chizadeh, Y. Tagalizadehkhoob, Impac o hype ension on a ious ma ke s
o subclinical a he oscle osis in ea ly ype 2 diabe es, Jou nal o Diabe es
& Me abolic Diso de s 13 (1) (2014) 1.
[6] M. T. Romanos, A. Raspo ic, B. M. Pe in, The eliabili y o oe sys olic
p essu e and he oe b achial index in pa ien s wi h diabe es, Jou nal o
oo and ankle esea ch 3 (1) (2010) 31.
[7] J. Allen, A. Mu ay, Simila i y in bila e al pho ople hysmog aphic pe iph-
e al pulse wa e cha ac e is ics a he ea s, humbs and oes, Physiological
measu emen 21 (3) (2000) 369.
[8] J. Allen, K. O e beck, A. F. Na h, A. Mu ay, G. S ansby, A p ospec i e
compa ison o bila e al pho ople hysmog aphy e sus he ankle-b achial
23
p essu e index o de ec ing and quan i ying lowe limb pe iphe al a e ial
disease, Jou nal o ascula su ge y 47 (4) (2008) 794–802.
[9] J. Allen, C. P. Oa es, T. A. Lees, A. Mu ay, Pho ople hysmog aphy de-
ec ion o lowe limb pe iphe al a e ial occlusi e disease: a compa ison o
pulse iming, ampli ude and shape cha ac e is ics, Physiological measu e-
men 26 (5) (2005) 811.
[10] R. E s, J. Spigulis, I. Kukulis, M. Ozols, Bila e al pho ople hysmog aphy
s udies o he leg a e ial s enosis, Physiological measu emen 26 (5) (2005)
865.
[11] M. Pel okangas, A. A. Telembeci, J. Ve ho, V. M. Ma ila, P. Romsi,
A. Vehkaoja, J. Lekkala, N. Oksala, Pa ame e s ex ac ed om a e ial
pulse wa es as ma ke s o a he oscle o ic changes: pe o mance and e-
pea abili y, IEEE Jou nal o Biomedical and Heal h In o ma icsIn p ess.
DOI: 10.1109/JBHI.2017.2679904.
[12] V. Jacomella, A. Shenoy, K. Mosimann, M. Kohle , B. Amann-Ves i,
M. Husmann, The impac o endo ascula lowe -limb e ascula isa ion on
he ao ic augmen a ion index in pa ien s wi h pe iphe al a e ial disease,
Eu opean Jou nal o Vascula and Endo ascula Su ge y 45 (5) (2013) 497–
501.
[13] M. Pel okangas, A. Vehkaoja, J. Ve ho, M. Huo a i, J. Rning, J. Lekkala,
Moni o ing a e ial pulse wa es wi h synch onous body senso ne wo k,
Biomedical and Heal h In o ma ics, IEEE Jou nal o 18 (6) (2014) 1781–
1787. doi:10.1109/JBHI.2014.2328788.
[14] M. Pel okangas, J. Ve ho, A. Vehkaoja, Nigh - ime ekg and h moni o ing
wi h bed shee in eg a ed ex ile elec odes, In o ma ion Technology in
Biomedicine, IEEE T ansac ions on 16 (5) (2012) 935–942.
[15] S. C. Millasseau, R. P. Kelly, J. M. Ri e , P. J. Chowienczyk, The ascula
impac o aging and asoac i e d ugs: compa ison o wo digi al olume
24
pulse measu emen s, Ame ican jou nal o hype ension 16 (6) (2003) 467–
472.
[16] M. Pel okangas, A. Vehkaoja, J. Ve ho, V. M. Ma ila, P. Romsi, J. Lekkala,
N. Oksala, Age dependence o a e ial pulse wa e pa ame e s ex ac ed
om dynamic blood p essu e and blood olume pulse wa es, IEEE jou nal
o biomedical and heal h in o ma ics 21 (1) (2017) 142–149.
[17] V. Meleno sky, B. A. Bo laug, B. Fe ics, K. Kessle , L. Shi ely, D. A. Kass,
Es ima ion o cen al p essu e augmen a ion using au oma ed adial a e y
onome y, Jou nal o hype ension 25 (7) (2007) 1403–1409.
[18] S. C. Millasseau, J. M. Ri e , K. Takazawa, P. J. Chowienczyk, Con ou
analysis o he pho ople hysmog aphic pulse measu ed a he inge , Jou -
nal o hype ension 24 (8) (2006) 1449–1456.
[19] K. Takazawa, N. Tanaka, K. Takeda, F. Ku osu, C. Ibukiyama, Unde es i-
ma ion o asodila o e ec s o ni oglyce in by uppe limb blood p essu e,
Hype ension 26 (3) (1995) 520–523.
[20] J. Randolph, F ee-ma ginal mul i a e kappa (mul i a e κ ee): An al-
e na i e o leiss’ ixed-ma ginal mul i a e kappa, in: P esen ed a he
Joensuu Lea ning and Ins uc ion Symposium, Vol. 2005, 2005.
[21] K. O. McG aw, S. P. Wong, Fo ming in e ences abou some in aclass co -
ela ion coe icien s., Psychological me hods 1 (1) (1996) 30.
[22] C. Lessells, P. T. Boag, Un epea able epea abili ies: a common mis ake,
The Auk (1987) 116–121.
[23] S. Ray, T. Buckenham, A. Belli, R. Taylo , J. Do mandy, The na u e and
impo ance o changes in oe-b achial p essu e indices ollowing pe cu a-
neous ansluminal angioplas y o leg ischaemia, Eu opean jou nal o as-
cula and endo ascula su ge y 14 (2) (1997) 125–133.
25