平成13（2001）年度　一般研究２実施報告書

課題番号	13−共研−2023			専門分類		3
研究課題名	てんかん脳波の時空間解析
フリガナ 代表者氏名	ウチダ スナオ 内田 直			ローマ字		Uchida Sunao
所属機関	(財)東京都医学研究機構
所属部局	東京都精神医学総合研究所
職　　名	副参事研究員
所在地
TEL				FAX
E-mail
URL
配分経費	研究費	0千円	旅　費		0千円		研究参加者数	4　人

研究目的と成果（経過）の概要

Mesial temporal lobectomy has now become a popular treatment for temporal lobe
epilepsy.For this operation,it is necessary to detect the location of epileptic focus.For this
purpose,electrocorticogram(ECoG)is an efficient implement.However visual observations of
epileptic waveforms or inter-ictal epileptic discharges do not always provide enough information
for the determination of epileptic focus.In this study,we tried to determine epileptic
hemisphere by analyzing basic activities of medial temporal lobe ECoG using linear stochastic
time series analysis.
Non-linear time series analysis based on deterministic chaos theory has been widely applied
to epileptic EEG and ECoG,mainly correlation dimension(D2)and Lyapunov exponent.
Lehnertz et al demonstrated that neural complexity is reduced on focal side(neural complexity
loss)comparing to non-focal side and the location of foci can be detected by ECoG data(analyzed
data length were about 30-45 minutes)[1].Neural complexity loss is defined through D2 in their
work.
However non-linear deterministic method,D2,takes much computation cost because
correlation integral has to be computed by changing embedding dimensions,and the saturation
is evaluated empirical visual inspection,it is rather subjective proceeding and not easy to
automated.Our interest is discrimination of focal side and also studying the changing of
dynamical properties in all-night recorded ECoG data,therefore these short points in
deterministic non-linear analysis are disadvantage for our purpose.In this work it was shown
that even though linear model,if it is sensibly applied to actual data,it fills our requirement.
Dynamical characterization was attempted for the detection of focal side with employing
multivariate autoregressive(AR)model.This model doesn't need so much computation cost,and
the all of estimation proceeding for parameters is done automatically and objectively.Through
these parameters the system can be identified.For the evaluation of the dynamical property
impulse response functions of estimated multivariate AR model were computed.In this work
the discrimination of focal side was attempted by taking advantage of the difference of damping
speed of impulse response function between focal and non-focal side.Though the methods based
on deterministic chaos theory are mainly for interpretation of system,the dynamical properties
of system can be evaluated more positively by simulation through stochastic model.
Spike count is generally high during NREM sleep(Fig.1-b),and higher in the epileptic
hemisphere.This patient has focus in right interior temporal lobe.The impulse response
function in the focal hemisphere dumps faster than its in other side(Fig.1-a).The difference is
larger during REM sleep when epileptic discharge is less frequently.The same difference of
impulse response function was observed also during wake state before and after all-night
sleep.Note that these differences were observed in epileptic discharge free basic
activities.Thus,this method give an extremely important information in the determination of
epileptic hemisphere even in cases epileptic discharges were not necessarily enough to
determine laterality.
[1]Lehnertz K,Elger CE.Spatio-temporal dynamics of the primary epileptogenic area in
temporal lobe epilepsy characterized by neuronal complexity loss.Electroencephalogr Clin
Neurophysiol 1995;95:108-117.
Fig.1(a)Fluctuation of attenuation coefficient averaged every minute,(b)number of spikes per
minute.(Solid line and dotted line are corresponding to left and right side of interior temporal
lobe respectively.)(c)Sleep stage

当該研究に関する情報源（論文発表、学会発表、プレプリント、ホームページ等）

●Babloyantz A,Destexhe A.Low-dimensional chaos in an instance of epilepsy.
Proc Natl Acad Sci USA 1986;83:3513-3517.
●Bergey GK,Franaszczuk PJ.Epileptic seizures are characterized by changing
signal complexity.Clin Neurophysiol 2001;112:241-249.
●Chokroveerty S.Sleep and Epilpsy.in:Chokroveerty,S.,ed.Sleep disorders
medicine.Boston:Butterworth-Heinemann,1994:429-454
●El Bahh B,Lespinet V,Lurton D,Coussemacq M,Le Gal La Salle G,Rougier A.
Correlation Between Grnule Cell Dispersion,Mossy Fiber Sprouting,and
Hippocampal Cell Loss in Temporal Lobe Epilepsy.Epilepsia
1999;40:1393-1401.
●Frank GW,Lookman T,Nerenberg MAH,Essex C.Chaotic time series
analyses of epileptic seizures.Physica D 1990;46:427-438.
Grassberger P,Procaccia I.Characterization of strange attractors.Phys Rev Lett
1983;50:346-349.
●Houser CR.Granule cell dispersion in the dentate gyrusof humans with
temporal lobe epilepsy.Brain Res 1990;535:195-204.
●Jing H,Takigawa M.Comparison of human ictal interictal and normal
non-linear component analyses.Clin Neurophysiol 2000;111:1282-1292.
●Kim JH,Guimaraes PO,Shen MY,Masukawa LM,Spencer DD.Hippocampal
neuronal density in temporal lobe epilepsy with and without gliomas.Acta
Neuropathol 1990;80:41-45.
●Le Van Quyen M,Martinerie J,Navarro V,Baulac M,Varela F.Characterizing
Neurodynamic Changes Before Seizures.J Clin Neurophysiol 2001;18:191-207.
Lehnertz K,Elger CE.Spatio-temporal dynamics of the primary epileptogenic
area in temporal lobe epilepsy characterized by neuronal complexity loss.
Electroenceph and clin Neurophysiol 1995;95:108-117.
●Lurton D,Sundstrom L,Brana C,Bloch B,Rougier A.Possible mechanisms
inducing granule cell dispersion in humans with temporal lobe epilepsy.
Epilepsy Res 1997;26:351-361.
●Miwakeichi F.Characterization of Spike & Wave Signals in epileptic EEG:A
non-linear non-parametric time series approach.in:ed.Ph.D.Thesis,
Department of Statistical Science,School of Mathematical and Physical
Science,The Graduate University for Advanced Studies.2001:
●Montplaisir J,Laveerdiere M,Saint-Hilaire J,Rouleau I.Nocturnal sleep
recording in focal epilepsy:A study with depth electrodes.J Clin
Neurophysiol 1987;4:383-388.
●Nakamura H,Akaike H.Statistical identification of optimal control of thermal
power plants.Automatica 1981;17:143-155.
●Otomo T,Nakagawa T,Akaike H.Statistical approach to computer control of
cement rotary kilns.Automatica 1972;8:35-48.
●Pitka(umlaut)nen A,Nissinen J,Lukasiuk K,et al.Association between the
density of mossy fiber sprouting and seizure frequency in experimental and
human temporal lobe epilepsy.Epilepsia 2000;41(Suppl.6):S24-S29.
●Sloviter RS.The functonal organization of the hippocampal dentate gyrus and
its relevant to the pathogenesis of temporal lobe epilepsy.Ann Neurol
1994;35:640-654.
●Theiler J.On the evidence for low-dimensional chaos in an epileptic
electroencephalogram.Phys Lett A 1995;196:335-341.
●Uruno K,O'Conner MJ,Masukawa LM.Alterations of inhibitory synaptic
responses in the dentate gyrus of temporal lobe epileptic patients.
Hippocampus 1994;4:583-593.

研究会を開催した場合は、テーマ・日時・場所・参加者数を記入してください。

研究参加者一覧
氏名	所属機関
尾崎 統	統計数理研究所
平井 伸英	(財)東京都医学研究機構
三分一 史和	東京都精神医学総合研究所