PosterPDF Available

A wristband assessment of accelerometry and autonomic activity of epileptic patients

Authors:

Abstract

In this work, the performance of an automated seizure detection system based on ACM and EDA features measured from the wrist was presented using clinical data collected from a total of 53 patients having two types of seizures. The classifier we tested allows a high seizure detection rate for GTC and FOCM seizures that had never been trained by the model (Se=100% on 12 new seizures) while maintaining an acceptable false alarm rate of 0.93/day on average over 164 days. Furthermore, it is efficiently integrated into a hardware platform to provide real-time alarms of seizures while they are occurring. In the future, the model will be tested on data collected outside the clinic, where the test conditions are expected to be much more challenging.
!"#$%&"'(
)!"#$%&'"%()! )#$*+*$, -./01 *2 ) 34,2*&!&5*#)! 2*56)! %"7!"#$*65 $4"
)#$*+*$, &7 $4" 28")$ 5!)6'2 '%*+"6 9, $4" :,(3)$4"$*# ;"%+&<2
:,2$"(= ./0 )6' 8%*2$ )##"!"%)$*&6 -0>?1 (")2<%"("6$2 #)6 9"
<2"' $& )<$&()$*#)!!, *'"6$*7, 5"6"%)!*@"' $&6*#A#!&6*# 2"*@<%"2
-BC>:21 *+,=
D<% 3%*&% 8&%E *-./, 24&8"' $4)$ <2*65 FG 7")$<%"2 7%&( ./0 )6'
0>?H )6' ):<33&%$ I"#$&% ?)#4*6" )33%&)#4H *$ 8)2 3&22*9!" $&
&9$)*6 ) '"$"#$&% -:I?JFG18*$4 2"62*$*+*$, -:"1 &7 KLM)6' 7)!2"
)!)%( %)$" -N0O1 &7 P=GP "+"6$2Q'), &6 ) 2"$ &7 FR BC>:2 7%&( SR
"3*!"32, 3)$*"6$2=
T$ *2 *(3&%$)6$ $& #&6$*6<" $& $"2$ $4"2" ("$4&'2 &6 (&%" 3)$*"6$2H
(&%" 2"*@<%"2H (&%" 7)!2"A)!)%(A"!*#*$*65 )#$*+*$*"2H )6' &+"% !&65"%
3"%*&'2 &7 $*("=
010"2%3&$+4151617&%3"#$+581!(9"'$"+51:1:$9'$%3$&$+51!1;$<"3=+5-4
+>1.(3)$*#)HUT6#HU>)(9%*'5"HU?0U)6'U?*!)6HUT$)!,HU888="(3)$*#)=#&(UUUU->1?TCU?"'*)UV)9HU?)22)#4<2"$$2UT62$*$<$"U&7UC"#46&!&5,HU>)(9%*'5"HU?0UU
W"()*!2XU##Y"(3)$*#)=#&( &%U%3Y"(3)$*#)=#&(
!(?(3(&<(@
*+, Z&4 "$ )!=H [0<$&6&(*# #4)65"2 8*$4 2"*@<%"2 #&%%"!)$" 8*$4 3&2$*#$)! ..B 2<33%"22*&6[ !"#$%&%'(H
)*-PF1H SR\R]SR^\HPGSP=
*-, B= O"5)!*) "$ )!=H _06 *(3%&+"' 8%*2$A8&%6 #&6+<!2*+" 2"*@<%" '"$"#$&% 9)2"' &6 )##"!"%&("$%, )6'
"!"#$%&'"%()! )#$*+*$, 2"62&%2[H +,"$-./0 12-&"23( 4%.-"5( /00#/& ,""5-0' 6789H)92 6&=FGK\HPGSL=
*/, N= D6&%)$* "$ )!=H _T(3%&+"("6$ &7 ) #&6+<!2*+" 2"*@<%" '"$"#$&% %"!,*65 &6 )##"!"%&("$"% )6'
"!"#$%&'"%()! )#$*+*$, #&!!"#$"' #&6$*6<&<2!, 9, ) 8%*2$9)6'[H 12-&"23( :-2"&-0" ;%0<"$"0."HPGS\=
:(#A%=@
B6 #&!!)9&%)$*&6 8*$4 $&3 4&23*$)!2H 8" #&!!"#$"' #!*6*#)!!, !)9"!!"'
2"*@<%" ')$) <2*65 +*'"& ..B -+A..B1H #&62*2$*65 &7 SKP
%"#&%'*652 $)E"6 7%&( LF 3)$*"6$2 8")%*65 )6 .(3)$*#) .F &% .`
8%*2$ 2"62&% %"#&%'*65 ./0 )6' FA)a*2 0>? -6$9C3( +D"E1=
C4" ')$) 8"%" )6)!,@"' &77A!*6" <2*65 3%&3%*"$)%, 2&7$8)%"
-.(3)$*#)H T6#=1 $& #!")6 $4" ')$) )6' "a$%)#$ 2*56)! 7")$<%"2 &6 )
SG 2"#&6'2 8*6'&8 "+"%, P=L 2"#&6'2 -&+"%!)3X ^LM1 -6$9C3(
+D2E1=
;"a$H :I?JFG 8)2 "(3!&,"' *6 &%'"% $& "+)!<)$" $4" :" )6' N0O
&6 $4*2 $"2$ 2"$= C4" &3$*()! '"#*2*&6 $4%"24&!' $& '*2#%*(*6)$"
9"$8""6 2"*@<%" )6' 6&6A2"*@<%" "3&#42 8)2 2"!"#$"' 9, (")62
&7 %"#"*+"% &3"%)$*65 #4)%)#$"%*2$*# #<%+" )6)!,2*2 &6 )3%*&%A
#&!!"#$"' $%)*6*65 2"$ -FR BC>:2 7%&( SR 3)$*"6$2 &+"% `F '),21=
!(@C'#@
C4" :I?JFG #!)22*7*"% 8)2 $"2$"' &6 ')$) 7%&( 7*+" 6"8
3)$*"6$2 4)+*65 (&%" '*+"%2" 2"*@<%"2H <2*65 6&6" &7 $4"*% ')$)
*6 $4" $%)*6*65 2"$= C4" 6"8 $"2$ ')$) #&62*2$"' &7 F8G0 "&= H
?%<"' I%#%3 D670:E @($JC3(@ ?3%I KL"#$(&#@ %M(3 /NF
A%C3@ -SL=^L '),21= C& 7<%$4"% %)*2" $4" '*77*#<!$,H 8" )!2&
*6#!<'"' $"2$ ')$) &7 6&6A2"*@<%" 8")%)9!" 2"62&% %"#&%'*652
7%&( HF (L$'(L@O L"#$(&#@ &+"% ) $&$)! &7 /KPH A%C3@ -S`R=L
'),21 *6 &%'"% $& (*(*# $4" %")!*2$*# <$*!*@)$*&6 &7 $4" 8")%)9!"
'"$"#$&% &+"% "#%#"' %? +PH ="O@=
C4" #!)22*7*"% 8)2 )9!" $& 3%&+*'" )6 )!)%( 9"7&%" $4" 2"*@<%"
4)' 7*6*24"' *6 +QQR&7 $4" #)2"2H 8*$4 )(")6 '"!), *6 $4" SP
#)2"2 &7 -S @(<%&=@ -:/bSP 2"#1= 02 ) 3"%#"6$)5" &7 $4"
2"*@<%" '<%)$*&6H $4" '"$"#$*&6 &##<%%"' &6 )+"%)5" )7$"% FKM
&7 $4" 2"*@<%" 4)' &##<%%"'= -6$9C3( -1
0%&<'C@$%&@
B6$4*2 8&%EH $4" 3"%7&%()6#" &7 )6 )<$&()$"' 2"*@<%" '"$"#$*&6 2,2$"(
9)2"' &6 0>? )6' ./0 7")$<%"2 (")2<%"' 7%&( $4" 8%*2$ 8)2 3%"2"6$"'
<2*65 #!*6*#)! ')$) #&!!"#$"' 7%&( )$&$)! &7 LF 3)$*"6$2 4)+*65 $8& $,3"2 &7
2"*@<%"2=
GA( <'"@@$?$(3 T( #(@#(= "<A$(M(@ "A$9A @($JC3( =(#(<#$%& 3"#( 7&% BC>
)6' ND>? 2"*@<%"2 $4)$ 4)' 6"+"% 9""6 $%)*6"' 9, $4" (&'"! -:"bSGGM&6
SP 6"8 2"*@<%"21 TA$'( I"$&#"$&$&9 "& "<<(L#"2'( ?"'@( "'"3I 3"#( &7
G=KFQ'), &6 )+"%)5" &+"% S\` '),2=
N<%$4"%(&%"H$4" ("$4&' *2 "77*#*"6$!, *6$"5%)$"' *6$& )4)%'8)%" 3!)$7&%( $&
3%&+*'" %")!A$*(" )!)%(2 &7 2"*@<%"2 84*!" $4", )%" &##<%%*65=
T6 $4" 7<$<%"H $4" (&'"! 8*!! 9" $"2$"' &6 ')$) #&!!"#$"' &<$2*'" $4" #!*6*#H
84"%" $4" $"2$ #&6'*$*&62 )%" "a3"#$"' $& 9" (<#4 (&%" #4)!!"65*65=
=" /$" '$/5"<#& 5% >/0 ?$-"@,/0A B% C - D%@@"0E",2"$A ;&/#3 F"-03C"$'"$A F/0- 4/$E-3A /0@ 5G"-$ 5"/,3 /5
!HIA J%35%0 ;G-&@$"03 K%32-5/&A /0@ J$-'G/, L=%,"03 K%32-5/&A J%35%0A /0@ 5% M%0/5G/0 J-@N"&& /0@
G-3 5"/,3 /5 1,%$( /0@ ;G-&@$"03 K"/&5G./$" %< +5&/05/ K%32-5/&A +5&/05/A <%$ 5G"-$ '"0"$%#3 3#22%$5
.%&&".5-0' 3"03%$ @/5/ /0@ &/C"&-0' O-@"%P11Q3R
6$9C3( +> D"E ./0 )6' FA)a"2 0>? 2*56)!2 &7 )3)$*"6$ %"#&%'"' '<%*65 )
5"6"%)!*@"' $&6*#A#!&6*# 2"*@<%" -BC>21 8*$4 )8%*2$A8&%6 '"+*#"=D2E
:#4"()$*# 8&%E7!&8 &7 $4" BC>2 '"$"#$&%=
6$9C3( -> V)$"6#, &7
'"$"#$*&6 -'"!), 9"$8""6
..B &62"$ &7 2"*@<%" )$
$*(" G)6' 7*%2$ "3&#4
#!)22*7*"' )2 2"*@<%"1 7&%
")#4 $"2$ BC> 2"*@<%"
U1T3$@#2"&=1"@@(@@I(&#1%?1"<<('(3%I(#3O "&=1"C#%&%I$<1"<#$M$#O1
%?1(L$'(L#$<1L"#$(&#@
N.0CcO.U
.dCO0>CTD; BC>:U/.C.>CDO
0>?
./0
N.0CcO.U
:.C
;DUBC>:
BC>:U-0V0O?1
D"E
D2E
C4" :I?JFG )<$&()$*#)!!, %"#&56*@"' SPQSP &7 $4" 6"8
2"*@<%"2 -6$9C3( /D"E1H )#4*"+*65 7&% S\` '),2 &7 %"#&%'*65 7%&(
LF 3)$*"6$2H ) @(&@$#$M$#O %? +QQR"&= 6U! %? Q>S/
(M(&#@V="O=D7 $4" K/ (L$'(L@O L"#$(&#@H`G 3)$*"6$2 -(&2$ &7
$4"(1 )+"%)5"' !"22 $4)6 &% "e<)! $& S 7)!2" )!)%(Q'),=D7
$4"2"H SR 4)' 6& 7)!2" )!)%(2 -6$9C3( /D2E1= :*a 4)' )N0O
4*54"% $4)6 PH 84*!" 2"+"6 4)' 7%&( SAP 7)!2" )!)%(2Q'),=
D"E
D2E
6$9C3( /> D"EC4" 4*2$&5%)( 24&82 $4" 6<(9"% &7 3)$*"6$2 4)+*65 ")#4
7)!2" )!)%( %)$" 3"% '),=D2E C4" 7*+" 6"8 3)$*"6$2 4)' 7%&( SA` 2"*@<%"2
")#4H )!! 3"%7"#$!, '"$"#$"'H "+"6 $4&<54 $4" )<$&()$"' ("$4&' 4)' 6&$
9""6 3%"+*&<2!, 24&86 )6, ND>? "+"6$2 &% )6, ')$) 7%&( $4"2" 3)$*"6$2=
... There are reports of highly accurate seizure detection systems using dedicated wristworn devices that have been helped by using electrodermal activity (EDA) sensors [22,25,26], which are currently unavailable in mainstream wearable devices. ...
... For instance, older smartphone devices might not feature accelerometers with optimal characteristics for detecting falls. Moreover, accurate heartrate and EDA sensors are not commonly found in smartwatch devices, even though their inclusion would be beneficial for detecting alarming states [22,25]. Nevertheless, with the increasing quantity and quality of sensors in modern devices, this is becoming less of an issue. ...
Conference Paper
Epilepsy and falls incur a great social and economic cost globally. Automatically detecting their occurrence would help mitigate the myriad of issues that arise from not receiving assistance after such an event. Despite existing research showing the potential advantages in using the ever-improving sensor technology incorporated within commercially available smartphone and smartwatch devices for human activity recognition, most available solutions for fall and seizure detection are still offered with dedicated hardware, which is often more expensive and less practical. This paper presents a comparison and evaluation of algorithms for detecting convulsions and falls, separately and combined, using smartphone and smartwatch devices. With a dataset of ordinary activities and simulated falls and convulsions, recorded by 15 test subjects, we found the devices a viable option for the successful detection of the activities, achieving accuracy rates between 89.7% and 98.5% with C4.5 decision tree algorithms.
Poster
Full-text available
Wrist acceleration (ACM), measured through a wearable device, has been used to automatically identify the motion of convulsive seizures such as generalized tonic-clonic seizures (GTCSs). ACM-based seizure detectors can trigger with many repetitive movements, and are affected by high false alarm rates, which might prevent daily life use. Recently, EDA and ACM features have been used to improve the classification performance of a seizure detector, reducing the false alarm rate in the detection of 16 GTCSs recorded from 7 patients. Here we present the results of further improvements, better matched to the real-time computational capacity of wearable devices, building a seizure detector from a higher number of patients and GTCSs (i.e., 38 GTCSs from 18 patients).
Conference Paper
Full-text available
Measurement of wrist acceleration (ACM) by means of wearable devices has been exploited to automatically detect ongoing motor seizures in patients with epilepsy (Epilepsy and Behav 2011, 20, 638-641; Epilepsia 2013, 54(4), e58-e61). Nevertheless, such seizure detectors can show high false alarm rates in active patients, which might hinder their use in daily life. Electrodermal activity (EDA) is a physiological signal reflecting sympathetic activity. Large ipsilateral EDA responses are elicited via direct stimulation of several subcortical regions (Int J Psychopysiol 1996, 22, 1-8). Measuring a combination of EDA and ACM has been previously shown to enhance specificity, i.e. to reduce the false alarm rate, in detection of secondary generalized tonic-clonic seizures (GTCS) (Epilepsia 2012, 53(5), 93-7). Nevertheless, the aforementioned approach requires further improvements in generalization capability and in further reducing false alarm rate for use in the widest variety of daily activities. Accordingly, in this contribution we report the performances of four ACM and EDA-based seizure detectors fed with different feature sets, trained on a higher number of seizures than in our previous work (Epilepsia 2012, 53(5), 93-7). See more at: https://www.aesnet.org/meetings_events/annual_meeting_abstracts/view/2327131#sthash.GtSp9aOu.dpuf
Article
Sudden unexpected death in epilepsy (SUDEP) poses a poorly understood but considerable risk to people with uncontrolled epilepsy. There is controversy regarding the significance of postictal generalized EEG suppression as a biomarker for SUDEP risk, and it remains unknown whether postictal EEG suppression has a neurologic correlate. Here, we examined the profile of autonomic alterations accompanying seizures with a wrist-worn biosensor and explored the relationship between autonomic dysregulation and postictal EEG suppression. We used custom-built wrist-worn sensors to continuously record the sympathetically mediated electrodermal activity (EDA) of patients with refractory epilepsy admitted to the long-term video-EEG monitoring unit. Parasympathetic-modulated high-frequency (HF) power of heart rate variability was measured from concurrent EKG recordings. A total of 34 seizures comprising 22 complex partial and 12 tonic-clonic seizures from 11 patients were analyzed. The postictal period was characterized by a surge in EDA and heightened heart rate coinciding with persistent suppression of HF power. An increase in the EDA response amplitude correlated with an increase in the duration of EEG suppression (r = 0.81, p = 0.003). Decreased HF power correlated with an increase in the duration of EEG suppression (r = -0.87, p = 0.002). The magnitude of both sympathetic activation and parasympathetic suppression increases with duration of EEG suppression after tonic-clonic seizures. These results provide autonomic correlates of postictal EEG suppression and highlight a critical window of postictal autonomic dysregulation that may be relevant in the pathogenesis of SUDEP.
Improvement of a convulsive seizure detector relying on accelerometer and electrodermal activity collected continuously by a wristband
  • F Onorati
F. Onorati et al., "Improvement of a convulsive seizure detector relying on accelerometer and electrodermal activity collected continuously by a wristband", Epilepsy Pipeline Conference, 2016.