Can we predict the onset of seizures? Behnaam Aazhang J.S. - PowerPoint PPT Presentation

Can we predict the onset of seizures? Behnaam Aazhang J.S. Abercrombie Professor Electrical and Computer Engineering Rice University

Can we predict the onset of seizures? • Let’s step back with a few more fundamental questions.

How can engineers contribute to medicine? • understanding various disorders • developing therapies • patient-specific • episode-specific • scalability • cost

engineers • problem solving with constraints • developing tools • sense and measure • nano-electronics • control—modulation, stimulation, pacing • machine learning and data analytics

example • pacemakers

example • pacemakers • Can we modulate our neurological circuit? • 86 billion neurons • 10 micron diameter • 100 Hz clock speed • 100 trillion synapses

Rice neuroengineering initiative Get Data (Nanotechnology) Szablowski St. Pierre Xie Luan Robinson Seymour Raphael Veerarag- Kemere havan Interpret and Use Data (Signal Processing) O’ Malley Pitkow Aazhang Patel Baraniuk Allen Hardware Algorithms

Rice neuroengineering initiative Get Data (Nanotechnology) Szablowski St. Pierre Xie Luan Robinson Seymour Raphael Veerarag- Kemere havan Interpret and Use Data (Signal Processing) O’ Malley Pitkow Aazhang Patel Baraniuk Allen Hardware Algorithms

What am I excited about? • Can we predict the onset of seizures?

What am I excited about? • Can data analytics predict and prevent the onset of seizures in epileptic patients?

epilepsy • unprovoked and recurring seizures • seizure • no standard definition • abnormally hyper-excited neuronal activities

epilepsy • celebrities

epilepsy • 1% of world’s population • causes: stroke, tumors, infection, genetic, developmental,… • 1/3 of patients do not respond to medication • resection!!!!! • deep brain stimulation?

ictal the challenge

the challenge inter-ictal

the challenge pre-ictal

approach • patient and episode specific • identify the seizure onset zone • understand the dynamics of the underlying system • predict seizures • modulate (stimulate) to prevent the onset of seizure

epilepsy • identify seizure onset zone RPBT1 RAMY2 RAH1 seizure zone Seizure Start Time 0 10 20 30 Time (s)

epilepsy • identify seizure onset zone RPBT1 RAMY2 RAH1 seizure onset zone Seizure Start Time 0 10 20 30 Time (s)

epilepsy • identify seizure onset zone RPBT1 RAMY2 causality RAH1 Seizure Start Time 0 10 20 30 Time (s)

causality • one time series forecasting another • economics • transportation • … • n. wiener (1956), c. granger (1969), h. marko (1973) • j. massey (1990), g. kramer (1998), • c. quinn, et. al. (2011)

a little background • directed information and causality N X 1 ; Y n | Y n − 1 I ( X N → Y N I ( X n 1 ) = ) 1 1 n =1 • directional with temporal information Y N X N ≡ ( Y 1 , Y 2 , . . . , Y N ) 1 ≡ ( X 1 , X 2 , . . . , X N ) 1

N X I ( X N → Y N I ( X n 1 ; Y n | Y n − 1 1 ) = ) a little background 1 1 n =1 • mutual information of time series N X 1 ; Y n | Y n − 1 I ( X N 1 ; Y N I ( X N 1 ) = ) 1 n =1 • no temporal and no causal information Y N X N ≡ ( Y 1 , Y 2 , . . . , Y N ) 1 ≡ ( X 1 , X 2 , . . . , X N ) 1

<latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> <latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> <latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> <latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> I ( X N 1 ; Y N 1 ) = H ( Y N 1 ) − H ( Y N 1 | X N 1 ) a little background • directed information of time series I ( X N 1 → Y N 1 ) = H ( Y N 1 ) − H ( Y N 1 || X N 1 ) • where N X H ( Y N 1 || X N H ( Y n | Y n − 1 , X n 1 ) = 1 ) 1 n =1

<latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> <latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> <latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> <latexit sha1_base64="UvF6lyb0fNWp7Abw2mz+Q50GlL0=">ACJnicbVBLSwMxGMz6rPW16tFLsAgtaNmIoJdC0UtPpYJ9SB9LNk01NJtdkqxQtvtrvPhXvHioiHjzp5hte1DrQMgwMx/JN17ImdKO82ktLa+srq1nNrKbW9s7u/befkMFkS0TgIeyJaHFeVM0LpmtNWKCn2PU6b3vA69ZuPVCoWiFs9CmnXx/eCDRjB2kiuXark71zUq47HsJXeBViCHRX5bixKOnF1QRWoIkIOIZpMBanKIEn07AouHbOKTpTwEWC5iQH5qi59qTD0jkU6EJx0q1kRPqboylZoTJNuJFA0xGeJ72jZUYJ+qbjxdM4HRunDQSDNERpO1Z8TMfaVGvmeSfpYP6i/Xir+57UjPbjsxkyEkaCzB4aRBzqAKadwT6TlGg+MgQTycxfIXnAEhNtms2aEtDflRdJ46yInCK6Oc+Vr+Z1ZMAhOAJ5gMAFKIMKqIE6IOAJvIAJeLOerVfr3fqYRZes+cwB+AXr6xsRFKHS</latexit> I ( X N 1 ; Y N 1 ) = H ( Y N 1 ) − H ( Y N 1 | X N 1 ) a little background • directed information of time series I ( X N 1 → Y N 1 ) = H ( Y N 1 ) − H ( Y N 1 || X N 1 ) causal conditional entropy • where N X H ( Y N 1 || X N H ( Y n | Y n − 1 , X n 1 ) = 1 ) 1 n =1

back to seizures • causal relation among electrodes • directed information • model free—data driven • k-nearest neighbor density estimation • identify time series with largest directed information RPBT1 RAMY2 → ˆ f X,Y → ˆ H ( X ) , ˆ H ( X, Y ) → ˆ I ( X → Y ) RAH1 Seizure Start Time 0 10 20 30 Time (s)

seizure onset zone • causal influence—directed connectivity • a graph with electrodes as nodes and directed information as edge • pre-ictal (period prior to seizure) RAH1 RPH4 RMOF10 RPBT1 LAH12 RAH13 LAH8 RPBT11 electrode RMOF5 RAH5 RAMY7 LAH5 RAINS4 RAH2 RAMY6 RAMY12 RPH3 RAMY5 RAH3 RPH2 RAMY11 RPH10 RAH10 RAH6 RAMY10 RAMY9 RAMY2 RAINS3 RAMY4 RAMY3