Recen Inno a ions in Mecha onics (RIiM) Vol. 2. (2015). No. 1-2.
DOI: 10.17667/ iim.2015.1-2/12.
1
Technological conside a ions in measu ing wea he -
de i ed epilep ic seizu es o dogs
Ádám Pin é
MOGI Depa men
Budapes Uni e si y o Technology
and Economics
Budapes , Hunga y
adam.pin e @icicom.hu
K isz ián Samu
MOGI Depa men
Budapes Uni e si y o Technology
and Economics
Budapes , Hunga y
[email p o ec ed]me.hu
Kálmán Czeibe
Depa men o Ana omy and
His ology
Facul y o Ve e ina y Science
Budapes , Hunga y
czeibe .kal[email p o ec ed]u
Abs ac —As in e e y echnical discipline, he objec i e,
ep oducible ins umen al measu emen s a e indispensable. I
is no di e en a in es iga ing he e ec o wea he on li ing
o ganism. The pa allel measu emen o he a mosphe ic
en i onmen and he physiological pa ame e s is no el and
indi idual, as well as he in es iga ion o epilep ic seizu es
depending on wea he si ua ions. Se e al ques ions came up in
connec ion wi h ou equipmen designed o s udying dogs
epilep ic phenomenon, such as ixing, wa e esis an ,- and
obus implemen a ion, he p oblem o da a s o age,
measu emen schedule o he powe supply o ins ance.
Keeping hese in mind du ing he de elopmen p ocess we go
an appa a us, which since is an in eg al pa o ou u he
esea ches.
Keywo ds—e ec s o wea he , epilepsy, ins umen al
measu emen s
I. STATE OF THE ART
As demons a ed in ou p e ious pape s [1][2], he
scien i ic wo ld pays close a en ion o in es iga e he
physiological e ec s o he wea he , bu s ill no awa e o
any equipmen – besides ou de ices – which measu es
pa allely he physiological cha ac e is ics o an indi idual
and he indi idual's a mosphe ic mic oen i onmen . In
addi ion, in such a way o examine he impac o wea he on
epilep ic seizu es a e comple ely no el [3][4]. This pape
discusses expe imen al ins umen de elopmen p oblems
ela ed o his esea ch a ea.
II. INTRODUCTION
Today i is accep ed as a ac ha he wea he – and i s
changes – has a signi ican impac on li ing o ganisms, as
bo h in e na ional li e a u e [5-8] and ou p e ious
publica ions [3][4] poin ed ou . In hese publica ions we
discussed some measu ing echnology ques ions as ecu ing
p oblems, such as he schedule o he measu emen , he
sampling equency, o he wea abili y and com o o he
equipmen s. We ha e pe o med measu emen s bo h on
human and animal (especially dogs) indi iduals – om
which mu ually use ul esea ch esul s can be hoped due o
he high deg ee on analogy [9] –, bu he implemen a ion o
he measu emen s aise e y di e en kinds o p oblems. In
his pape we ocus on dogs' epilep ic seizu e measu emen
di icul ies.
I is necessa y o discuss b ie ly he backg ound o
epilep ic symp oma ology, so i can be seen, ha s udies
should be pe o med only wi h indi idualized
cha ac e is ics, and indi idual assessmen should be applied
in e e y case. Dogs’ epilep ic seizu es show a high deg ee
o co ela ion wi h he human's, and seizu e classi ica ion is
based on a simila manne p o ided by he ILAE
(In e na ional League Agains Epilepsy) [10]. In he
pe spec i e o his pape – and also he de eloped es
equipmen – i is ele an o poin ou ha an epilep ic
seizu e can be pa ial ( ocal) o gene alized, and in he
majo i y o cases, mon hs o e en yea s be o e a g and mal
seizu e – in ol ing loss o consciousness and gene alized
con ulsions – ocal symp oms may show. In addi ion,
seizu es may no a ec only he musculoskele al sys em,
he e a e also seizu es a ec ing isce al (wi h au onomous
symp oms) and senso y sys em (eg. s inging, bu ning,
ingling sensa ions e c.), o hey esul he impai men o
consciousness (eg. hallucina ions and empo a y a en ion
de ici diso de s). The objec i e assessmen can be achie ed
in hese cases usually only wi h p ope ideo-EEG
moni o ing. Consequen ly ou s udy is p ima ily aimed a
animals whose seizu es induce mo o ic mo emen changes
(i.e. seizu es e ec ing muscle mo emen , hus easie o
de ec ins umen ally): hese can be onic i s – wi h
con inuous con ac ion –, muscle con ac ions wi h g a e o
less equency (myoclonus, ei he locally o sp eading o e
he whole body), mixed ( onic-clonic) o a onic seizu es as
well.
III. QUASI-TIME-POINT EVENTS – EPILEPSY
Ou discussed esea ch a ea is o examine he epilep ic
seizu es in dogs gene a ed by a mosphe ic mic o and mac o-
en i onmen al impac . P e ious esul s ha e demons a ed
he exis ence o his ela ionship [3][4]. I is impo an ha
each animal ha e indi idual seizu e-pa e n: an animal may
ha e only one kind o seizu e, bu i o en happens ha is
shows se e al ocal symp oms beside gene alized
Recen Inno a ions in Mecha onics (RIiM) Vol. 2. (2015). No. 1-2.
DOI: 10.17667/ iim.2015.1-2/12.
2
con ulsion, and he empo al dis ibu ion o hese seizu es
a e di e en . In long- e m epilepsy ea men one o he
mos impo an ac o is he seizu e equency, i.e. how
equen ly epilep ic i s occu (=how long is he seizu e- ee
o in e ic al pe iod). An epilep ic e en gene ally ollows he
nex o de : changes in beha io be o e he seizu e
(p od oma sign eg. changes in ac i i y), hen he epilep ic i
(ic al signs), and a e ha pos ic al pe iod. In some case he
la e is much longe in ime compa ed o he i han he
e ec i e seizu e (i could ake hou s o e en days), so i
could be easily mis aken wi h he ac ual seizu e, since as
majo i y o epilep ic e en s las only o minu es (o jus 10-
15 seconds), so owne o he animal only could no ice – i
hey exis (eg. agg ession, ea , psychosis, compulsi e
diso de s, e c) – he pos ic al signs, no he i i sel . Ano he
challenge in cons uc ing a de ice which is capable o
egis e hese e en s in co ela ion wi h wea he ac i i y
while me eo ological changes akes place h ough se e al
days (o weeks) we ha e o look o epilep ic i s wi h
minu es o e en seconds (and hus esul ing only a small
ac ion in he obse a ion pe iod).
To sol e he p oblem o managemen and s o age o he
big ou pu da a o long- e m measu emen s he p edic abili y
o wea he changes can be e y help ul. In case o a gi en
indi idual – possess su icien knowledge – we can
de e mine wha ype o wea he si ua ion we should s a ou
measu emen s o inc ease he sampling equency. In he
i s phase o he examina ion o a selec ed animal, when
we do no ha e p io knowledge abou he indi idual ye , i
is necessa y o collec in o ma ion co e ing he en i e
spec um o wea he .
IV. COMPLEX, INEGRATED MEASUREMENTS
As se e al physiological, en i onmen al, me eo ological
pa ame e s ha e o be measu ed in sync, a high deg ee o
in eg a ion is equi ed, which aises a numbe o di icul ies.
Because o he mainly long- e m and high- esolu ion
measu emen s, ansmission and s o age o la ge amoun s o
da a is equi ed. The wi eless implemen a ion – see below –
sends he da a h ough a 2.4 [GHz] communica ion channel
o he docking uni so he amoun o da a o be ansmi ed is
inc eased in p opo ion o he communica ion a ic, and
ha means ex a ene gy needs. The s o age and he
p ocessing o he big amoun o da a is an addi ional
p oblem in case o mon h- e m s udy o nume ous
indi iduals. The local da a-p ocessing could be a solu ion o
his, bu he s udied me eo ological p ocesses can no be
conside ed poin -like and also no independen om each
o he , hus i would need se ious memo y and compu ing
capaci y, which is oo ene gy in ensi e. Fo as much as he
measu emen o he ac i i y1 (as one o he physiological
pa ame e s) is quasi-con inuous, while he wea he
pa ame e s a e enough o measu e in minu e in e als, we
need o ocus on his pa ame e a he solu ion o da a
amoun p oblem. The ollowing solu ion s a egy has been
de eloped (schema ically):
 We ind he lowes sampling equency – based on
Shannon's heo em [11] –, which does no dis o
he use ul signal mo e han a h eshold we de ine.
Fo his, apa om he ex eme mo emen s a 25
[Hz] sampling equency is sa is ac o y;
 A scala so wa e has been de eloped, which,
unde a gi en le el o ac i i y keeps he
mic ocon olle in he ul a-low-powe mode, hen
as he mo emen amoun g ows – wha we de ec
wi h ha dwa e in e up s – is se s he measu emen
dynamic ange o he le el o ac i i y, and a e a
educ ion in he ac i i y is swi ches back o he
IDLE s a e;
 O he senso s du ing no mal beha io a e u ned on
only wi h small ime du y a io and a e he
measu emen se ies (which happen a he same
ime) hey swi ches back o s andby;
 In case o abno mal ac i i y de ec ion o a d as ic
change in he measu ed pa ame e s – depending on
he se ings – measu e equency can be inc ease
au oma ically.
These solu ions can op imize he use o ene gy – which has
an impo an ole in wea abili y h ough educing he
necessa y size o he ba e y – and educes he amoun o
da a as well. Da a downloaded om he docking uni is hen
subjec ed o u he p ocessing – o example we can s o e
mo emen -pa e n in o ma ion o ac i i y le el ins ead o
accele a ion alues – o educe he amoun o da a o be
1 h ee-axis gy oscope and absolu e accele a ion
Figu e 2. Occu ence o an examined animal's epilep ic
seizu es in 2013 ( oge he wi h he iden i ied
me eo ological on s in hose pe iods)
Recen Inno a ions in Mecha onics (RIiM) Vol. 2. (2015). No. 1-2.
DOI: 10.17667/ iim.2015.1-2/12.
3
s o ed. The a mos amoun o ou pu da a a ises on ha
e sion o he de ice which eco ds ECG-signal as well,
especially in con inuous ope a ion mode. In his case he
mos impo an is o de e mine he equi ed measu emen
sec ions, which is cu en ly unde de elopmen .
I also posed a se ious challenge o design he de ice –
wi h ene gy sou ces, a numbe o senso s, wi eless module
and cen al mic ocon olle uni – in such a deg ee o
in eg a ion le el ha is s ill should be small and com o able
o wea . To his, beside se ious echnical expe ience, he use
o smalles size componen s in he indus y con ibu ed. We
used:
 0402 size code passi e componen s;
 QFN package mic ocon olle uni ;
 Su ace Moun (SMD) 2.4 [GHz] an enna;
 mul i-le el PCB implemen a ion a MEMS senso
modules;
 Li-Polyme ba e y.
The use o hese echnologies is clea ly equi es o en e low
solde ing, wha – being a small p oduc se ies – we did
ou sel es using he appa a us o ICI In e ac i e Co. L d.
V. FIXING, ERGONOMICS AND SIGNAL LEAD
Since in case o dogs we can no expec ha he wea e
o he de ice pays pa icula a en ion o i s sa e y, he
implemen a ion and he ixing mus be obus . In addi ion, i
is an equally impo an aspec ha he de ice does no
impede he animal in he no mal way o li ing, no o
dis u b he indi idual wi h i s p esence e en in case o a
long- e m use. Va ious echnological aspec s a ose du ing
he design and es ing (schema ically):
 In he accele a ion measu emen – i we a e
in e es ed also in he mo emen pa e n – he
di ec ion o he senso (so ha we know which way
is "up") is impo an , and o de e mine he co ec
amoun o mo emen is essen ial ha he senso
mo e oge he wi h he animal's mass cen e . I is
ob ious ha he animal will p oduce qui e di e en
accele ome e da a when he senso is loca ed on i s
back o on one o he lowe legs;
 Measu ing he ambien empe a u e i is impo an
ha he lead o he empe a u e senso p obe poin
he ou side as mo e as i is possible o a oid
measu ing he body empe a u e o he animal. The
empe a u e senso s mus be loca ed on he mobile
de ice, as he docking uni is almos always being
indoo , while he dog is no necessa ily, and we
need he a mosphe ic mic oen i onmen o he
animal, as i s o ganism exposed di ec ly o ha ;
 Measu ing he ambien ligh – which is also
impo an in epilepsy examina ions [12][13] –
s ands simila o he c i e ia abo e;
 In case o he ECG e sion he ches elec odes
should be p o ec ed agains majo shi , o he wise
he quali y o he signal lead d as ically
de e io a es.
Beside hese he ollowing gene al c i e ia should be
conside ed:
 The de ice need o be placed so he dog can no
each i and o example no able o chew i ;
 The de ice mus be wa e - esis an , wi h wa e -
esis an echa ging abili y. The o me we sol ed
wi h epoxy molding, he la e wi h gil , co osion-
ee connec o s;
 Due o he la ge amoun o da a he mobile uni
should send he da a h ough wi eless
communica ion o he docking uni so he de ice
ixed on he animal can be eally mobile while he
da a can obse ed in eal- ime. These wo ea u es
a e e y impo an , so du ing he design p ocess
we e main ained wi h a high p io i y.
Based on he demands abo e using ha ness seemed o be
an ob ious solu ion. I is he mos s abile way o ixing a
de ice on a dog. This way, we ha e he chance o se he
senso s in he bes possible loca ion ( his in luences mos
he ECG elec odes' posi ion-s abili y), conside ing also he
ac o s occu s in he animal's ac i i y.
The la e no only helps o il e he noise, bu also –
aking in o accoun he speci ic cha ac e is ics o he
pa icula animal’s epilep ic symp oms – may help o ind
he bes loca ion whe e o place he accele ome e (eg. in he
case o ocal con ulsions o head and neck muscles i can be
placed on neck ing, o in case o limb mo emen he p ope
side o he animal)
Figu e 4. ECG mobile de ice's main PCB
Recen Inno a ions in Mecha onics (RIiM) Vol. 2. (2015). No. 1-2.
DOI: 10.17667/ iim.2015.1-2/12.
4
Figu e 5. Two o en used ha ness ype
sou ce: h p://www.pe d eam.hu
VI. RESULTS
As a esul o he echnical conside a ions ou lined abo e
we ge an easily wea able ins umen , which – besides i is
capable o se e al days o con inuous ope a ion wi h a
single cha ge – p o ides eliable measu emen esul s. The
measu emen s o he ambien ligh and empe a u e a e
sa is ac o y, he ac i i y measu emen is excellen . The ECG
signal's lead du ing se ies o es s on animals in es ing
posi ion b ough e y good esul s, bu in case o mo ing,
ac i e animals u he imp o emen s a e needed o inc ease
he s abili y o he elec odes. Ou u he goal is o c ea e an
expe imen al ool can gain da a mo e e icien ly and
accu a ely, hus also sui able o clea iden i ica ion o he
ype o epilep ic seizu es.
Figu e 6. Mobile de ice (ECG ype) and docking uni
Re e ences
[1] Ádám Pin é , K isz ián Samu, “De elopmen o a complex
mic oen i onmen - and hea moni o ing equipmen o measu ing he
e ec s o he wea he on human o ganism”, CERiS'13, pp. 176-180,
2013, ISBN 978-963-313-086-5
[2] Ádám Pin é , K isz ián Samu, “Technological solu ions o
compensa e humanme eo ological® measu emen e o s wi h
measu ing human ac i i y and mo emen ”, CERiS'14, pp. 30-37,
2014, ISBN 978-963-313-086-5
[3] Kálmán Czeibe , Ádám Pin é , Ane I ády, K isz ián Samu,
“Rela ionship be ween me eo ological ac o s and epilep ic i s in a
Jack Russell Te ie ” Magya Álla o osok Lapja 136:(8) pp. 461-
471. (2014)
[4] Kálmán Czeibe , Ádám Pin é , Ane I ády, K isz ián Samu, “E ec
o he me eo ological ac o s on he i s o an epilep ic dog. Case
epo ” Magya Álla o osok Lapja pp. 149-156. (2014)
[5] S.L. Molloy, S. Mihal che a “Ex emeWea he E en s”, Re e ence
Module in Ea h Sys ems and En i onmen al Sciences, Volume 1:
Vulne abili y o Human Heal h o Clima e, 2013, Pages 3–16.
[6] R Julian P es on “Unce ain ies in es ima ing heal h isks associa ed
wi h exposu e o ionising adia ion”, Jou nal o Radiological
P o ec ion, 2013.
[7] João Vasconcelos “The impac o win e cold wea he on acu e
myoca dial in a c ions in Po ugal”, En i onmen al Pollu ion,
Volume 183, Decembe 2013, Pages 14–18.
[8] S.L. Molloy “Ex eme Wea he E en s and Human Heal h”,
In e na ional Encyclopedia o Public Heal h, 2008, Pages 536–544.
[9] McCandless, D. W. “Epilepsy. Animal and Human Co ela ions”,
Sp inge , 2012
[10] de Risio, L. – Pla , S “Canine and Feline Epilepsy: Diagnosis and
Managemen ” CABI, 2014
[11] Huba An al, PhD, Gabo Halasz, PhD “Technical Measu emen s”,
ISBN 963 420 744 8, Pages 29-30.
[12] Senanayake, N “Re lex epilepsies: expe ience in S i Lanka”, Ceylon
Med J. 1994 Jun;39(2):67-74.
[13] Chang Yeob Lee – Ji-Ye Jeon – Yong Won Cho – Hye-Jin Moon “A
Case o Tempo al Onse Pa ial Seizu e Induced by Pho ic S imuli”, J
Epilepsy Res. Jun 2014; 4(1): 18–20.