Brain Topography, Volume 2, Number 3, 1990 221

Brain Mapping Analysis in Patients with Hepatic Encephalopathy

Teresa Sagal~s, Victor Gimeno, M. Dolores de la Calzada, Francesc Casellas, M. Dolors Maci~, and M. Villar Soriano

Summary: Topographical analysis of cerebral electrical activity was performed in 44 patients with hepatic encephalopathy. These patients were classified in 5 groups according to clinical criteria. Eight healthy subjects were used as a control group. All were studied in an awake, eyes closed, condition and some [Control Group (CG), Group 0 (GO), Group I (G1) and Group 2 (G2)] also in an awake, eyes open, condition. The awake, eyes closed, maps showed marked differences in the power spectral density (PSD) of the different bands, when comparing normal subjects with patients with several degrees of hepatic encephalopathy. These differences were related to the degree of clinical involvement, mainly in the alpha and delta PSD bands. The combination of a decreased alpha PSD, increased delta PSD, and decreased mean dominant frequency (MDF) allowed a dear discrimination between the different clinical groups. The differences observed between awake, eyes closed, and awake, eyes open, conditions were especially helpful to discriminate between CG subjects and GO, G1 and G2 patients.

Key words: brain mapping; hepatic encephalopathy; power spectral density; PSD; mean dominant frequency; MDF.

Introduction

An essential element in the clinical evaluat ion of hepatic encephalopathy (HE) is the availability of objec- tive methods to quantify the degree of involvement of the central nervous system. Classifications have been based on the clinical scores suggested by Opolon et al. (1976) or on the use of psychometr ic tests, such as those of Conn et al. (1977). Since the early 1950's, several studies (Bickford and Butt 1955, Foley et al. 1950, Parsons-Smith et al. 1957) have analyzed the e lectroencephalogram (EEG) features of patients in hepatic coma or with other forms of HE. Al though EEG recordings constitute an objective ap- proach to the evaluat ion of these patients, the simple analysis of conventional tracings necessarily carries a considerable amount of variability and lack of precision. As expected, these techniques were soon fol lowed by the advent of more elaborate methods, such as f requency analysis (Hawkes et al. 1970), spectral analysis (Yaar et al. 1981, Garcia de Leon et al. 1983, Van der Rijt et al. 1984, 1985 ) and visual evoked potentials (Casellas et al. 1985). The experience of our group in using the methods of

Servicio de Neurofisiologia Clinica. Hospital General Vall d'- Hebron. Barcelona.

(Accepted for publication: February 28, 1990. ) Correspondence and reprint requests should be addressed to: Dr.

T. Sagal6s, Servicio de Neurofisiologia Clinica, Hospital General Vall d'Hebron, Pg. Vall d'Hebron s/n. 08035, Barcelona, Spain.

Copyright © 1990 Human Sciences Press, Inc.

topographic analysis of the EEG signal (Brain m a p p i n g ) is here applied to the s tudy of the cerebral activity of patients with HE.

Method

Fourty-four patients with chronic liver disease and HE were studied and classified in 5 groups according to the clinical criteria of Parsons-Smith et al. (1957) and Zieve (1982) (Table 1). Clinical data were reviewed inde- pendent ly by two physicians. Group 0 (GO) was com- prised of 9 patients with a mean age of 45 years (range 29-66) with no symptoms of HE, but with clinical and histological data of chronic liver disease. Group 1 (G1) included 11 patients with a mean age of 59 years (range 52-69). Group 2 (G2) had 13 patients with a mean age of 57 years (range 30-76). Group 3 (G3) had 6 patients with a mean age of 62 years (range 32-78). Group 4 included 5 patients with a mean age of 61 years (range 57-65). A group of 8 heal thy subjects, mean age 35 (range 26-50), was used as a control group (CG).

A twenty-channel Alvar-Reega EEG machine was used for recording EEG data f rom Fpl , Fp2, F3, F4, F7, F8, T3, T4, T5, T6, C3, C4, P3, P4, O1 and 02. Disc electrodes were placed according to the 10-20 International System. Linked earlobes were used as reference. The pass band of the amplif ier was set be tween 0.5 and 35 Hz (6 dB/Oct) . Data were processed by means of a PDP 11/34 a n d a H P V e c t r a ES. S a m p l i n g f r e q u e n c y w a s

222 Sagales et al,

Table I: Degrees of hepatic encephalopathy

HE 0: No clinical signs.

HE 1: Euphoria or apathy, lowered perception.

Short attention span.

Inverted sleep patterns.

Tremor.

HE 2: Accentuated grade I signs.

Obvious personality changes.

Flapping tremor.

HE 3: Bizarre behaviour.

Obvious confusion.

Desorientation.

Somnolence, Stupor

HE 4: Coma

128/sec/channel. Thirty artifact-free epochs of 4 second duration were recorded and analysed; that means a total of 120 seconds of cerebral electrical activity free from artifacts. A computerized artifact rejection mechanism, together with visual control, was used in the selection of the epochs. Eye blinks, muscle activity, vascular pulses and electrode artifacts were detected and rejected, as well as periods of drowsiness. Delta and triphasic waves were identified by eye. Screening was done by two senior authors thoroughly experienced in routine electroen- cephalography and special emphasis was put on the exclusion of drowsiness. Electrical activity brain maps were elaborated by means of a method created in our laboratory, which is similar to that used by others (Des- medt et al 1985). The spectra were divided into four different bands: delta (0.5-3.75 Hz), theta (4-7.75 Hz), alpha (8-11.75 Hz) and beta (12-32 Hz). A subdivision within the alpha band was also made: alpha I (8-9.75 Hz) and alpha 2 (10-11.75 Hz).

The mean dominant frequency (MDF), the total power spectral density (PSD) and the PSD percentage of the different bands and alpha fractions were calculated for each recording area. In view of the limited number of patients in each group, non-parametric tests were used in the statistical analysis of the results. These tests were the multiple range test of Newman-Keuls (SNK) (com- parisons of differences between subjects) and the Wil- coxon p a i r e d - s a m p l e tes t ( compa r i son b e t w e e n recordings in the same subject). Fp1, C3, T3 and O1 areas

were selected for statistical analysis between adjacent groups.

Recordings were performed under the following con- ditions:

A. Awake, eyes closed: During the recording the sub- jects were resting with closed eyes. This test was per- formed on all subjects (CG, GO, G1, G2, G3, G4).

B. Awake, eyes open: During the recording the sub- jects were in a resting position, but they were asked to look at a central point placed at a distance of 3 meters. This test was carried out only in those conscious patients able to collaborate (CG, GO, G1, G2).

Results

A. Awake, Eyes closed, Condit ion

Control Group

As expected, the greatest PSD was recorded in the alpha band, particularly in the occipital areas, with a mean percentage of 64.05 + 17.9 SD in O1. Much lower PSD percentages were recorded in the beta, theta and delta bands (Figures I and 4).

Group 0

Compared with the CG, GO patients showed a decrease of the alpha band PSD that was more evident in the occipital areas (O1:33.6 %, p < 0.05) (Figures I and 4). The overall topographic distribution of cerebral activity was not noticeably altered. The MDF decreased, mainly in the occipital regions (from 10.1 + 0.3 Hz (CG) to 7.8 + 0.7 Hz (GO); p < 0.02) (Table II). There was a marked increase in the power percentage of the beta band, with statistically significant differences in the anterior areas (Figures I and 4).

Group 1

G1 patients, compared with CG, showed an increase in the percentage of delta PSD (over 25% in the occipital areas with a maximum of 65 % in the frontal regions; p < 0.02) and in the theta band (more evident in the central, 35% p < 0.001, and occipital areas, 45% p < 0.05) (Figure 2). When they were compared with GO, G1 patients showed a diffuse increase in delta PSD, mainly in the parieto-occipital areas, and a tendency to increased theta power in the precentral regions. There was also a decrease in alpha power, more evident in the centro-tem- poral areas. The MDF was also decreased (O1:6.8 + 0.6), although the differences did not achieve statistical sig- nificance (Table II).

Brain Mapping Anatysis and HE 223

@ @ , @

CONTROL

~ .~ -~-:~ }:<~.~,~ ~ ~ ~<C~<~C~.~ ~ .~ ~ ~.~ <~ ~ . ~ ~ ~ ~ L ~

~ ~

,

: @ •

~ O b ~ 0

~ i ~

i ~ . ~.~ ~ ~ ~J, ~ ~=~ ~ ~ . ~ - ~ ~ , ~

~ ~ . ~ ~ - ~ ~ ~ % ~ ~ ~ ~

• ~ . ~ ~ ~ ~ ~ , ~ ~ ~

~ ~ ~,= ~=

Figure I. (top) Brain EEG map of a representative Control Group subject (left) and pooled topographic distribution of the mean CG values (right). (bottom) Brain EEG map of a representative patient in Group 0 (lefl), pooled topographic distributions of the mean GO values (middle), and distribution maps of the Newman-Keuls test p values (CG vs GO), (right). Color entries in increasing order <from lowest (5) to highest (75) percentage> are: Brown, green, blue, purple, pink, red, orange and yellow,

O@ • t . ~ ~ ~ ~ . ~ . ~ ~ . ~ ~ ,~ ~ ~=~=~. : ~ . ~ . ~ , = ~ : ~ . ~

~ ~

.~-~.~o~,. ~.

~ ~ ~< ~ . , ~ ~ ~ ~ , ~ p~ : ,~ . ~ , . ~ ~ . ~ ~ ~ ~ t ~ ~ ~ ~ ~ ~ , ~ ¢ ~ 7 ~ ~ ~

~ ~ I

Figure 2. (top) Brain EEG map of a representative patient in Group 1 (left), pooled topographic distributions of the mean G1 values (middle), and distribution maps of Newman-Keuls test p values (CG vs G1), (right). (bottom) Brain EEG map of a representative patient in Group 2 (left), pooled topographic distributions of the mean G2 values (middle), and distribution maps of the Neuman-Keuts test p values (CG vs G2), (right). Color entries are the same as in Figure 1.

224 Sagal6s et al.

6

i i 0 O

GROUP ~

~ " ~ E

~ . ~ - ~ ~ ~ , 8 ~ ' ~ ~ ~ , ~ ~ p,~,O~ ~ ~ ~,0.~9~ ~ , ~ - ~ ~ ~ , ~ ~ ~ ~ , ~ ~ ~ ~ , ~ - ~ ~ . < ~

~ , ~ ~

~ 4

S C A L E

,~,~ . . ~ ~ ~ - ~ ~

, ~ , ~ ~ ~ ~ . ~ ~ ~O~}2 ~ ~ ~ . O ~ ~ . ~ ~ ~ ~ . ~ . ~ ~ O ~ : ~ , ~ ~ ~

f ~ , ~ ~

Figure 3. (top) Brain EEG map of a representative patient in Group 3 (left), pooled topographic distributions of the mean G3 values (middle), and distribution maps of the Newman-Keuls test p values (CG vs G3), (right). (boltom) Brain EEG map of a representative patient in Group 4 (left), pooled topographic distributions of the mean G4 values (middle), and distribution maps of the Newman-Keuls test p values (CG vs G4), (right), Color entries are the same as in figure 1.

Fpl C 3 T 3 0 1

C 0 I 2 3 4 ¢ 0 1 2 3 4 ¢ 0 I 2 3 4 C 0 I 2 3 4

1 o~ ~ ~ ~ ~ ~

¢ 0 1 2 3 4 C 0 1 2 3 4 ¢ 0 1 2 3 4 ¢ 0 1 2 3 4

0 * C 0 I > 3 4 ¢ 0 I 2 3 4 C O 1 2 3 4 C~O 1 2 3 4

~8° 1 BETA

o, J ~ ~ ~ ~ . ~ ~ , ~ ~-~_,~_~ C 0 1 2 3 4 C O 1 2 3 4 C O 1 2 3 4 C 0 1 2 3 4

Figure 4 Histogram of the different power spectral density percentage bands for each group at Fpl, C3, T3 and Ot +_ SE, Dots indicate statistical level of significance (one: p < 0.05, %,~o: p < 0.005) when comparing a group with the previous one,

Brain Mapping Analysis and HE 225

Group 2

The more relevant data in group 2 were an increase in PSD of the delta and theta bands, and a decrease in the PSD of alpha and beta bands. This was found when compared with either CG or with G1 (Figures 2 and 4). The MDF was also decreased: it was 4.4 + 0.6 Hz in O1 (p < 0.02 when compared to the previous group) (Table II). Triphasic waves were identified in the recordings of two patients in this group.

Groups 3 and 4

When comparing G3 and G4 data with CG, a great increase in the percentage of delta PSD in all cerebral areas (65 and 75 % in O1 in G3 and G4, respectively) and a low percentage of alpha and beta PSD were observed. These findings were evident over all the brain, and dif- ferences were always statistically significant (p < 0.001) (Figure 3). A decrease in the MDF, and in the PSD of the alpha and beta bands, together with an increase in delta PSD, were detected when comparing each of these groups with the preceding one (G3 vs G2, and G4 vs G3). These differences only occasionally achieved a statistical- ly significant level, possibly due to the small number of patients in each group. (Figure 4, Table II). Triphasic waves were observed in 2 of the 6 patients of G3 and in 4 of the 5 patients in G4.

B.Power Differences Between awake, Eyes closed and Open, Conditions.

Four different groups participated in this study: GO, G1, G2 and CG. Each patient was explored twice, awake with eyes closed and awake eyes open, and changes induced by recording with eyes open were evaluated in each group.

Control Group

As expected, changes in PSD between both tests were readily observed, although there was some variability between subjects. Recording with eyes open resulted in a generalized power decrease in the alpha and beta bands, all over the cerebral areas. The alpha band chan- ges were particularly evident in the occipital region (p < 0.001). These differences were observed in both alpha subbands (alpha I and alpha 2). It was also in the occipital areas where the decrease of beta band power was more evident, and statistically significant. There was an in- crease in delta band power in the central and temporal areas (p < 0.05) in the majority of subjects, and in the occipital regions (p < 0.005) in all subjects (Figure 5).

Group 0

A decrease in alpha band power was noted in group 0

I Fpl

* *

- 1 0 0

1 O0

1 0 0

I-- Z 0

1 0 0

~ o

w a_

° C C ~ 4 " r A O L

C__~OL.P

Io

-- 100 [ D E L T A T H E T A ~ EKETA

Figure 5: Histogram of the percentage of subjects that experienced a change in PSD of the different bands, in different scalp areas, between awake eyes closed and awake eyes open conditions. (statistical significance: * p < 0.05, and **p < 0,005)

patients. However, this change was limited to the alpha 1 fraction (p < 0.05). The theta band showed decreased values in the temporal and in the central and occipital areas, where differences were statistically significant (p < 0.005). There was an increased theta PSD in the frontal areas in 83% of the patients. A generalized decrease in delta band power was observed, except in the frontal areas where it was clearly increased (p < 0.05 in Fpl) (Figure 5).

226 Sagal6s et al.

Table I1: Mean Dominant Frequency in the Control Group and in patients with different degrees of hepatic en- cephalopathy at Fp 1, C3, T3 and O1. SN K test was used in the statistical analysis, p values always refer to comparison with the previous group (* p < 0.05, ** p < 0.01, *** p < 0.001).

MEAN DOMINANT FREQUENCY (Hz)

(Mean + S.E.)

Cerebral areas Fpl C3 13 O1 Overall

Control Group 8.1_+0.3

(n=8)

Hepatic encephalopathy deree 0 7.3_+0.7

(n=9)

Hepatic encephalopathy degree I 5.1+0.6"

(n=11)

Hepatic encephalopathy degree 2 4.6+0.6

(n=13)

Hepatic encephalopathy degree 3 3.4+0.4

(n=6)

Hepatic encephalopathy degree 4 2.9_+0.2

(n=5)

8.5+0.4 9.3_+0.4 10.1+0.3 8.9_+0.9

7.5+0.8 8.0_+0.9 7.9_+0.7* 7.6_+0.3***

6.4+0.4 7.0_+0.9 6.8+0.6 6.3+0.9**

4.7+0.6* 5.5_+0.9 4.4_+0.6* 4.8_+0.4***

3.5+0.0 3.8_+0.0 3.4+0.1 3.5+0.2*

3.0+0.2 3.3_+0.3 2.9+0.2 3.2_+0.1

Groups I and 2

The progressive clinical deterioration of the cerebral functions of patients in groups I and 2 resulted in abnor- mal patterns of response to awake, eyes open, cerebral activity. A marked increase of theta and alpha power in the frontal region was the main feature in G1 patients, whereas patients in G2 showed increased power in all bands on the whole brain. In these patients, theta and alpha PSD were particularly increased in the temporal region (p < 0.005) (Figure 5).

Discussion

Brain mapping has proved of value in locating lesions in vascular (Culebras et al. 1986), and epileptic patients (Nuwer 1988, Wong et al. 1986, Gueguen and Gaches 1986), and in evaluating changes in the background EEG related to drug effects (Itil el al. 1981, Herrmann and Schaerer 1986). Interesting results have also been ob- tained using brain mapping in the assessment of patients with Alzheimer's disease (Duffy et al. 1984) and in depressive syndromes (Pockberger et al. 1985). In our study, brain mapping has revealed gradual differences between a control group and patients with different degrees of hepatic encephalopathy. These differences have been observed even in patients without mental,

personality or neuromuscular activity disturbances. As shown in Figure 4, brain mapping abnormalities in awake, eyes closed, patients are evident in all cerebral regions and involve almost all PSD bands, although the decrease in alpha and increase in delta band PSD seem most appropriate for the quantification of these changes. Duplicate recordings (awake, eyes closed and eyes open) further help to differentiate these patients from control subjects. The main feature in the brain mapping of the patients with grade 0 hepatic encephalopathy, i.e. those without mental, personality or neuromuscular activity disturbances, when compared with a control group of normal subjects was a decrease in occipital alpha PSD and MDF, and also an increase in the beta band PSD, in accordance with the findings of Samson-Dollfus et al. (1987). On the other hand, control subjects reacted to eyes open with a decrease in alpha 1. and alpha 2 activity, whereas this was restricted to the alpha 1 subband in group 0 patients. Moreover, there was an increase in occipital delta in control subjects when they kept their eyes open, whereas group 0 patients showed a decrease in occipital delta together with an increase in frontal delta activity. Furthermore, an important decrease of theta PSD as well as decreased values of delta PSD in central, temporal and occipital areas was also noted in GO patients.

An increase in delta and theta PSD with a marked

Brain Mapping Analysis and HE 227

decrease in alpha and beta PSD, as well as a decrease in MDF, in G1 awake, eyes closed, patients allowed us to differentiate them from control subjects and even from GO patients. When studying the PSD differences between both awake recordings, patients in G1 showed abnormal reactivity, with an increase in the delta, theta and alpha PSD in the frontal areas.

Brain mapping of G2 patients in awake, eyes closed, distinguished them from normal subjects and even from G1 patients, the main features being an increase in delta and theta PSD with a decrease in alpha and beta PSD and in the MDF. An increase in temporal alpha and beta PSD in awake, eyes open, was also a distinguishing feature in this group.

The progressive nature of the differences when com- paring patients in GO, G1 and G2 with the control group deserves some comment. A relatively young control group was selected in view of the mean age of the patients in group 0, i.e, those with clinical and histological data of chronic liver disease but without symptoms of HE, in which differences with the control group could be of greatest interest. There were no statistically significant differences in the age of subjects in CG and GO (Mann Witney's U Test) although brain mapping analysis did show significant changes. Moreover, while there were no statistically significant differences between the ages of G1 patients and those of the previous and following groups, comparison of brain maps between G1 and the other two, did show important differences. Also, it is worth noticing that an overall reduction in EEG amplitude for all fre- quency bands has been reported as the main topographic EEG change with normal aging (Breslau et al. 1989), while we have observed that increases in delta activity were most prominent in the sequence of HE group studied. Thus, although some questions may arise as to the clinical significance of comparisons with the control group, age differences per se do not seem to explain our results.

Brain maps in G3 and G4 patients were clearly abnor- mal, and although the number of subjects studied may have been too small to achieve statistically significant results, increases in delta PSD with decreased alpha and beta PSD, together with a decrease in MDF, were seen parallel to clinical deterioration in these two groups.

Using spectral analysis of EEG data in temporo-occipi- tal areas, van der Rijt et al. (1984,1985) have suggested that the percentage of theta activity is able to differentiate between groups 0 and 1, and that a difference between groups 2, 3 and 4 taken together can be established by the percentage of delta activity. Our results show that alpha and delta PSD are good discriminators in all groups, particularly if MDF is also considered. On the other hand, we have observed gradual changes in brain reactivity to the awake, eyes open, condition in the different groups studied. Although different bands were affected, changes

in alpha activity in the temporal areas seemed to be the most discriminative feature among the groups studied. Finally, it is worth commenting that, as shown in Figures 1-3, the color representation of the topographic analysis of EEG activity permits an immediate impression of the CNS abnormality that permits a clearly understandable and easily comparable clinical follow-up.

References

Bickford, R.G. and Butt, H.R. Hepatic coma: The electroen- cephalographic pattern. J. Clin. Invest., 1955, 34: 790-799.

Breslau, J., Starr, A., Sicotte, N., Higa, J. and Buchsbaum, M.S. Topographic EEG changes with normal aging and SDAT. Electroenceph. Clin. Neurophysiol., 1989, 72: 281-289.

Casellas, F., Sagal6s, T., de la Calzada, M.D., Acarino, A., Guarner, L. and Vargas, V. Potenciales evocados visuales en la encefalopatia hepfitica. Med. Clin. (Barc), 1985, 85: 139-142.

Conn, H.O. Trailmaking and number-connection tests in the assessment of mental state in portal systemic en- cephalopathy. Dig. Dis., 1977, 22: 541-550.

Culebras, A., Kline, M.D., Ross, G.S., Hodge, C.H. and Cruz, A. Quantitative EEG mapping of cerebral ischemia. Neurology, 1986, 36, Suppl 1,321.

Desmedt, J.E. and Bourget, M. Color imaging of parietal and frontal somatosensory potential fields evoked by stimula- tion of median or posterior tibial nerve in man, Electroen- ceph. Clin. Neurophysiol., 1985, 62: 1-17.

Duffy, F.H., Albert, M.S. and McAnulty, G. Brain electrical activity in patients with presenile and senile dementia of the Alzheimer type. Ann. Neurol., 1984, 16: 439-448.

Foley, J.M., Watson, C.W. and Adams, R.D. Significance of the electroencephalographic changes in hepatic coma. Trans. Am. Neurol. Assoc., 1950, 75: 161-165.

Garcia de Leon, M., De la Torre, J.M., Montero, J., Nieto, F.J., Bello, G., Castillo, E., De la Torre, S., Duran, R., Suarez, E., Lizarraga, P.M. and Hernandez, A. Electroencefalografia computarizada en la encefalopatia hepatica, Actas II Simp. Ing. Biomed., 1983, 137-140, Madrid.

Gueguen, B. and Gaches, J. La cartographie EEG dans les epilep- sies a crisis partielles. Rev. Electroenceph. Neurophysiol. Clin. 1986, 16:217-28

Hawkes, C.H., Macpherson, A.I.S., Pryor, H. and Townsend, H.R.A. The value of EEG frequency analysis in hepatic en- cephalopathy. J. Roy. Coll. Surg. Edinb., 1970, 15, 151-156.

Herrmann, W.M. and Schaerer, E. Pharmaco-EEG analysis to describe the projection of drug effects on a functional cerebral level in humans. In: Lopes da Silva F.H., Storm van Leenwen, W., R6mond, A. (Eds). Clinical applications of computer analysis of EEG and other neurophysiological signals. Amsterdam: Elsevier 1986:385-445 (Handbook of Electroencephalography and Clinical Neurophysiology, revised series: vol 2).

Itil, T.M. The discovery of psychotropic drugs by computer- analyzed cerebralbioelectrical potentials (CEEG). Drug Dev. Res. 1981, 1:373-407.

Nuwer, M.R. Quantitative EEG: II. Frequency Analysis and Topografic Mapping in Clinical Setting. J. Clin. Neurophysiol., 1988, 5:45-85

228 Sagal6s et al.

Opolon, P., Rapin, J.R., Huguet, C., Granger, A., Delorme, M-L., Boschat, M. and Sausse, A. Hepatic failure coma (HFC) treated by polyacrylonitrile membrane (PAM) hemodialysis ((HD). Trans. Amer. Soc. Artif. Intern. Organs., 1976, 22: 701-711.

Parsons-Smith, B.G., Summerskill, W.H.J., Dawson, A.M. and Sherlock, S. The electroencephalogram in liver disease. Lan- cet, 1957, ii: 867-871.

Pockberger, H., Petsche, H., Rappelsberger, P., Zidek, B. and Zapotoczky, H.G. On-going EEG in depression : a topographic spectral analytical pilot study. Electroenceph. Clin. Neurophysiol., 1985, 61:349-358.

Samson-Dollfus, D., Delmer,C., Fazi, S., Dreano, E., Fodil, D., Blasquez, R. and Terrisse, V. Etude des rythmes de base chez le subject normal et pathologique par la cartographie EEG. Rev. EEG Neurophysiol. Clin. 1987, 17:79-90

Van der Rijt, C.C.D., Schalm, S.W., De Groot, G.H. and Vlieger, M.E. Objective measurement of hepatic encephalopathy by means of automated EE G analysis. Electroenceph. Clin. Neurophysiol., 1984, 57: 423-426.

Van der Rijt, C. and Schalm, S.W. Quantitative EEG analysis and survival in liver disease. Electroenceph. Clin. Neurophysiol., 1985, 61: 502-504.

Wong, P.K.H., Gregory, D., Bencivenga, R. and Farrell K. Typi- cal and atypical benign rolandic epilepsy: statistical analysis of spike maps. Epilepsia, 1986, 27: 641.

Yaar, J., Shapiro, M.B. and Pottala, E. W. Spectral analysis of the EEG in hepatic encephalopathy treated with levodopa. Electroenceph. Clin. Neurophysiol., 1981, 52: 617-625.

Zieve, L. Hepatic encephalopathy in Diseases of the Liver. In: Sciff, L and Sciff, E.R. (Eds). J.B. Lippincott Co, Toronto. 1982, 433-459.