Top Banner
SEIN Medical Physics Department Stimulation-based state monitoring – new developments S. Kalitzin , D.Velis , F. Lopes da Silva Epilepsy Institute of The Netherlands (SEIN), Medical Physics Department Epilepsy Institute of The Netherlands (SEIN), Department of Clinical Neurophysiology Swammerdam Institute of Life Sciences, University of Amsterdam, The Netherlands
23

Stimulation-based state monitoring – new developments

Jan 12, 2016

Download

Documents

Jake

Stimulation-based state monitoring – new developments S. Kalitzin  , D.Velis  , F. Lopes da Silva   Epilepsy Institute of The Netherlands (SEIN), Medical Physics Department  Epilepsy Institute of The Netherlands (SEIN), Department of Clinical Neurophysiology - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Stimulation-based state monitoring – new developmentsS. Kalitzin, D.Velis, F. Lopes da Silva

Epilepsy Institute of The Netherlands (SEIN), Medical Physics Department Epilepsy Institute of The Netherlands (SEIN), Department of Clinical Neurophysiology Swammerdam Institute of Life Sciences, University of Amsterdam, The Netherlands

Page 2: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Phase clustering index (PCI)

Complex Fourieramplitudes

kkf

kkf

f Z

ZPCI

||

ffkkffk

kf APCIZPCIZ ||||

Repetitive stimulus

ktR

S. Kalitzin, J. Parra, D. Velis, F. Lopes da Silva, Enhancement of phase clustering in the EEG/MEG gamma frequency band anticipates transition to paroxysmal epileptiform activity in epileptic patients with known visual sensitivity, IEEE-TBME, v.49, 11 p 1279-1286, 2002

kfZ

Triggered response

qqkf

kkfq

f

Z

ZPCI 1

||

Page 3: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Question 1: how to stimulate and measure the response ?

Decision on stimulation sequence

Biphasic Monophasic

BiCAPCyclic Alternating Pattern (CAP)

Page 4: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Non-linear response to direct electric stimulation

What is a linear response ?

)()()( 2121 SbRSaRbSaSR

How can we test this ?

)(0)()()( 1111 noiseSRSRSSR

Page 5: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Averaged triggered response analysis

Page 6: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Page 7: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

sfff

sf QxTZ

Linear Response Model

)2/1(

2

2

21

||

f

f

ssf

ssf

f D

S

Z

ZPCI

s

sff QS 22 )( 2222

ffkkff xTZD

Page 8: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

For linear response model: total PCI=0

CAP – polarity alternation

sff

sf

sf QxTZ

sff

sf

sf QxTZ

fss

f xx )1(

Page 9: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

sffff

sf QxRxTZ )(

BICAP: Separating PCI of the non-linear response

sff

sfff

sf

sf

sf

sf

QxU

QxRxRZZWZ

)(

))()((1

)2/1(

2

2

1

f

fBICAPf U

SPCI

)2/1(

2

22

21

||

f

ff

ssf

ssf

f R

DS

Z

ZPCI

Page 10: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Stimulation in conducting fluid (NaCl)

Page 11: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Stimulation frequency

Example from patient case –MONOCAPStimulation frequency invariant response

Page 12: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Application to interictal localization of the SOSCase: left TLE

1)(max PCIPCIrPCI ff

Page 13: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Robustness to stimulation frequency

Page 14: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Question 2: where to stimulate and where to measure the response ?

(1)Averaged rPCI(2) Collective PCI

Page 15: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Processing features of the collective response to stimulation in one location

s

fcs

fcs

fcc

s

fcs

fcf

cs ZZC

ZRZ

1121 ,,,

,

,

4 22

2

))(())(~

(

)()(

fCTrfCTcxPCI

CTrRPCI

fcollective

f

ffcollective

Triggered response and correlation

Collective phase clusteringand bi-local phase coherence

fc

fc

fcc

fcc RRCC 2,, 12121

~ Partial correlation

Page 16: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Application to interictal localization of the SOSCase: (predominantly) right TLE

Page 17: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Collective Impact estimations

Page 18: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

rPC

I of

non

-lin

ear

resp

onse

Speculation: going beyond locality (bilateral case)

Page 19: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

What about Seizure prediction ?

Page 20: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Page 21: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Page 22: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Page 23: Stimulation-based state monitoring – new developments

SEINMedical Physics Department

Summary & Conclusions:

1. rPCI feature successfully detected the SOS in six new TLE cases. 2 left, 2 right 2 bilateral.

2. BICAP stimulation solves the problem of high PCI due to linear conductance.

3. For interictal localization tasks, the best spatial coding so far is at the stimulation site and the best response measure is the collective response.

4. Bi-local rPCI spread studies deserve attention.

5. Seizure prediction in time remains a statistical observation at best. Electrode size, recording sites and stimulation protocol might be critical. Robustness is therefore crucial.