- Email: [email protected]
Spectral analysis of the EEG (alpha rhythm) and activity in the left hemisphere: The effects of lateral gaze
NOTE SPECTRAL ANALYSIS OF THE EEG (ALPHA RHYTHM) AND ACTIVITY THE LEFT HEMISPHERE: THE EFFECTS OF LATERAL GAZE BERTRAND
Clinique
DETOFFOL.ALAIN
Neurologique,
IN
AUTRET. ERIC DECIOVAN~‘Iand SYLVIE Roux
CHU
(Received 6 December
Bretonneau,
37044 Tours.
Cedex. France
1989; accepred IO April 1990)
Abstract-This study tests the effect of maintaining right and left lateral gaze during a writing task which preferentially implicates the left hemisphere using an asymmetry parameter calculated from the spectral power of the alpha rhythm (RP- LP/RP+ LP) in a right-handed patient undergoing the same experimental regimen nine times. A six derivatton EEG was recorded. Maintaining left lateral gaze (toward the active hemisphere) removes the lateralization found durmg writing while staring straight ahead whereas maintaining right lateral gaze (toward the side opposite the active hemisphere) results in slightly lower values which are however, not significantly different from those obtained during staring straight ahead. This study adds an eiectrophysiologic aspect to Kinsbournes’s paradigm on gaze position and hemtspheric activation.
INTRODUCTION KINSBOURNE has demonstrated
that eye movements reflected the asymmetric distribution of activation between the cerebral hemispheres [lo. 1 I] and NE~XAUER er al’s. 1131 study has given an EEG validation of this hemispheric asymmetry model. Subsequently, a number of authors have been able to verify the inverse hypothesis: maintaining lateral gaze is accompanied by activation of the contralateral hemisphere. Indeed. significant improvement in tash performance involving a hemisphere has been obtained when gaze is maintained toward the contralateral half of the visual field as compared to maintained straight gaze or gaze toward the ipsilateral half of the visual field [6 8. I?]. During rest. it has previouslv been shown that maintaining lateral gaze produces a reduction in spectral power in the contralateral cerebral hem;sphere [I]: this can be interpreted as regecting hemispheric activation. The present study of EEG spectral analysis tries to detect an electrophysiological manifestation ofthe variation in activatron ofa hemisphere induced by eye position when that hemisphere is implicated in a cognitive task. This investigation IS part of other studies dealing with modifications in cerebral electrogenesis during cognitive activities prcfercnttally involving one of the cerebral hemispheres. There is no consensus on whether or not electrical lateralizatmn exists when a cerebral hemtsphere is preferentially solicited during a cognitive task. According to DAVIMN and EHRLICHMAN [3], preferential involvement of one hemisphere clearly results in reduced spectral power over that hemisphere. However. GEVlN.5 et ul.[5] constder that the lateraiized electrical modifications which accompany cognitive activities are due to non-specific sensorimotor activttics. We have recently developed a method of EEG spectral analysis which measures both the effects of task complexity on spectral power (variation in the log of spectral power) and the preferenttal involvement of one ofthe cerebral hemispheres in the task considered (variation in an asymmetry index RP-LP RPt LP). Accordtngly. when a gtven task prcferenttally solictts the left hemisphere, the power diminishes over that hemisphere while the asymmetry mdex Increases: thrs has been shown during right hand writing in a right-handed subject [I 51. Inversely. when a given task prcfcrcntially solicits the right hemisphere, both the power and the asymmetry index decrcasc over that hemisphere. Using this method. wc have been able to show that sensorimotoractrvrties and pure neuropsychological tasks each play a role m the variatmns in EEG spectral power [ 151. In this study we looked at the variatton+tn a specific parameter: asymmetry in alpha rhythm spectral power. A patient w’as exposed to four different expcrrmental condrtrons: rest, writtng while looking straight ahcad. wrrtrng while gazing to the right. and finally writmg vvhile gazing to the left. WC tested the CITCCIofmaintainingIatcral garc on left hemispheric power after it had been prcvrously “activated” by a wrtttng task. WC hypothcstscd that maintaining right lateral gaze during a writing task mcrca\es the asymmetry paramctcr in comparison to writtng vvhile staring straight ahead. wherea\ marntarnlng left lateral gaze dccrcasc\ this same paramctcr.
SUBJECT
AND METHODS
A 2%kear-old right-handed [ 131 man uithout medicatlonh recclied nlnc EEGs one e\er! other mc,rmng .II precisely the same time. He &as placed in the semi-rechnlng pos~tmn on a bed located an a part~all! soundproofed. dlml! ht room He underwent four different sequence>. each lastlns 2 mm InstructIons wcrc g~\cn \erhall! bct\\ecn the sequences. Eye movements were controlled b) an electro-oculol2ram. The pattent’s chcc Nere clohcd during the sequences. One rest sequence (Rest) alternated uith three acll\e sequences comprised h! one sequence tn\ol\~ng writmg a memorized text with the right hand while loohing straight ahcad (WI. a hccond \V w)ucncc \\hdc maintaming right lateral saze IWRLG I. and tinall!. one sequence durmg left lateral gaze IFVLLG I Sequence progress and patlent performance uere controlled. A latln square nas utlllzed 111order to cllmlnatc an! effect due to passmg order. Six symmetrIcal reference-vertex derlratmns were simultaneou\l\ rccordcd using carefully placed silver chloride electrodes pasted to the scalp: F4Cr F3Cz. ChCz CSCr. P4c‘z P.JCz. (IO 20 International system). The EEG uas amplified usin! an ALVAR polygraph (3 dh: 0.5 2.5 hHz). Sampling \+a\ accomplished in real time at a frequent! of 100 Hz using an anaio g dleltal converter coupled to a hlINCl I 03 and stored on a disc. Before samphng. stgnals were filtered hj anti-aliasmg kiters u hich had a cutting frequent! of 30 H/ and a 24 dB octave attenuatton frequency. The InformatIon was treated using a deferred mode. Succwl\c elementar! 2.56 set spectrums were calculated and each anal!sls period was filtered h> temporal rnultlpllcat~~~n usmp a Hanmng window. An mtcrmedlate spectrum was determlned from the average offourelemcntary spectrumh. We considered Intermediate spectrums having a cgreatcr than 200 !IV’ Hz power in frequency hand 0 I Hz to hc artefacts and dlscarded them. Twelve intermediate spectrums were used to determme a final average spectrum A hlch served to define three parameters: P lthe average power in /IV’ Hz during one sequence). log P and tinall! an asymmetry index (AI1 RP-LP RP+LP. The alpha rhythm wah studled: 7.03 12.89 HZ.
Variance analysis of the average log P values and the asymmetry Index for each sequence and each sglmclric derivation was performed and. hereafter. the Greenhouse Gel>sercorrectlon was applied [Y]. When the slgniticancc attained the P x0.05 threshold. theasymmetry index valuesofeach actliesequence werecompared to rest scqucnw while each staring straight ahead sequence has compared to lateral gaze sequences using the Newman Keulb teht.
RESULTS The results of the variance analqsls of lop P and the absolute spectral powers (P) in {IV’ Hz are shov, n 111Tahlc I. The ANOVA of log P after Greenhouse Geisser correctlon lead to a significant difference in F3Cr dcrivatlon and approach the significance threshold in P3Cz tP=O.Oh). In general. P decreased tn the left hemisphere during acti\c tasks in comparison to rest (except for WRLG and WLLG m PKz). The values of the asymmetry index (AI) arc summarized in Table 1 for each symmetrical dcrllation pair and each activit!. The variance analysis of the Al values. F (3. 24). was significant after Greenhouse Gewer correction 111two ofthe thrccdcrlvations (see the lirst line m Table 2).
Table I. Results of the log P ANOVA and spectral value\ of the r rhythm for each derivation and each sequence. *P
ANOVA t: values
Log P (F 3.241
F4C7
F3C7
CflC.7
I .5Y
5.13’ 0.641
I 52
2 42
2.6’)
13.7 13.Yi I (1 15.1 I
Y.32 7.X3 Y.31 XYI
56.33 63.46 hh.Xh 63.IY
CSCZ
P4Cz
P3CI 3.1 I 0.X I I
I ~rluc\rrf I’ Rest W WRLG WLLG
5.33 4.Yi 5.61 >.>i
5.52 3.17 4.YY 5.03
24.6’) 23.37 2X.5X 2x.47
(al In comparison to rest. Al is slpnificantly Increawd tpowr 1, wcakcr over the Icft hcmt\phcre) during the nrlting task while staring \tralght ahead. I’,() 01 in F4(‘1f 3. I’~005 in 1’4C‘/f’3. th) Comparison between WRLG rest. 41 ih\i~nilicantly tncrca\ed in lc4C‘1F3 1/‘<0,05)during WRLG, hut nonslfnlficant in P4CzP3.
881
NOTE Table 2. derivation correction associated
Values of the AI (asymmetry index) for each symmetrical pair and each sequence. *P
ANOVA AI F (3.24) E values significance after GGC
C6CzC5
P4CzP3
5.20’ 0.508 P
4.06 0.475 ns
5.37* 0.833 P < 0.05
Values 01 AI -0.01 0.08 0.06 0.04
Rest W WRLG WLLG
(c) Comparison between (d) Comparison between (e) Comparison between significant in F4CzF3. (f) Comparison between
F4CzF3
WLLG/rest. W/WRLG. W/WLLG.
0.38 0.45 0.40 0.38
None of the differences were significant. None of the differences were significant. Al was significantly increased in P4CzP3
WRLG/WLLG.
None of the differences
(PcO.05)
during
W. but non
were significant.
In summary, AI for the WRLG sequence is no different from W and is significantly stronger than for the rest sequence in F4CzF3 derivation. AI during the WLLG sequence is significantly weaker for W in P4CzP3 derivation and is no different from the rest sequence. These results are consistent with the hypothesis upon which the study was conducted.
DISCUSSION The method we used concerning spectral analysis. montage selection, tasks and analysis parameters has been discussed elsewhere [I 51. We studied a single subject who was repeatedly exposed to the sameexperimental regimen because EEG values and interhemispheric asymmetrtes show a great deal of individual variation [ 133. Furthermore. TUCKEKrr al. [ 171 have emphasized the interest of EEG single subjects’s studies. We selected an electrophysiological parameter. the asymmetry index R P - LP:RP + LP which gives information on the relative values of spectral power of each hemisphere with respect to the other. A zero index signifies that the power is the same over the two hemispheres, a positive one means that the power is stronger over the right hemisphere or weaker over the left hemisphere. There was a significant difference in AI during writing while staring straight ahead as opposed to during rest and we studied the influence oflateral gaze on this parameter. We limited our analysis to the alpha rhythm since the spectral modifications induced by cognitive activity has been studied most thoroughly in this frequency band [5] which is the most sensitive to the effects of lateral gaze [ 131. We chose writing a memorized text with the right hand for the right-handed patient since this tash is easy to reproduce and because it is generally admitted that. on the basis of a classic lesion model 141. it preferentially solicits the left hemisphere. The Greenhouse Geisser correction is required in repeated-measures ANOVA to avoid false positives results. The “epsilon” values are calculated and giv*en in tables. The observed Fratios (F 3. 24) are referred to as F (3r:. 241:). If only F (31:. 24~) exceeds the conservative critical values. the null hypothesis should be rejected. This study shows that right lateral gazeduring writingdoes not modify the Al valuesin relation tostaringstraight ahead while maintaining left lateral gaze signiticantly reduces AI. which approaches the value obtained during rest. Our hypothesis that AI would be reduced during left lateral gaze has been confirmed. One may postulate that the right hemisphere is”activated” by mamtaining left lateral gaze. tending to balance the power in the two hemispheres. and therefore producing an AI nearly equal to that seen durmg rest. However. the first part ofour hypothesis has not been confirmed: AI values during right lateral gaze are lower than those found during staring straight ahead. withouf being signilicantly different from the latter. This ma\ he due to the experimental conditions chosen. during which lateral gaze does not etlect performance. Indeed. review of the literature has shown that maintaining right lateral gaze during a tash involving the left hemisphere was intended to increase the task performance. Performance was measured by studying reaction times or hy the lateralization score durmg thedichotic listening test [h. 7.8. 173. There is increased performance in relation to hram stem lesions whether lateral gaze is voluntary or pathological [2]. The demonstrated hemispheric activation may he related IO preferential attention focalization to stimuli m the contralateral hcmispace [ 161. However. this study does not address performance modification during a tash according to gaze position or modilication in gaze position during changes in “hemispheric activation”. like in the
classical Kmsbourne paradigm. Rather. it deals wth the variation of a lateralizatlon parameter H hlch has hecn Induced by a wrtting actiwty. Independent of all cognitive activity, It has prebiously been shown that maIntaIned lateral gaze reduced alpha power m the contralateral hemisphere in relation to staring straight ahead [I]. Measuring the activation of a hemisphere by usmg an electrophysiological lateralization parameter based on an analysis of the background EEG only gives a very general impresslon of the activation state. Aasoclatinp two let”1 hemispheric “activating” tasks (right lateral gaze and writing) does not produce a simple additive effect. When the left hemisphere is activated by wrttmg it has reached ceiling and rtght lateral gaze cannot actrrate II an> further. Thus. this study adds an electrophyslological aspect to the importance oflateral gaze in the activation ofthe cerebral hemispheres.
REFERENCES 1. AUTRET. A.. ALVERT. L.. LAFFONT. F. and LARMA\IX. P. Electroencephalographic spectral pwer and lateralized motor activities. Elccrrorncepll. clitl. ,Yeuroph~.sio/. 60, 2% 236. 1985. ?. BELIX, C.. PERR~ER.D.. CAMBIER. J. and LARMASLX P. Influence de la diviatlon pathologique du regard par lesion du tronc &t-bra1 sur I’kquilibre inter-hemlspherique ttudie par le test d’kcoute dichotique non verbal Nruropsyho/ocja 26, 753-758. 1988. 3. DAVIUSON. R. J. and EHRLICHMAS. H. Lateralized cogmtive processes and the electroencephalogram. .%rww 207, 1005-1006. 1980. 4. GARO~~ER.E. ~~ir,~dawwtal.s (!/ A’ruroloyr. T/w Forebrain (‘ewhrcrl Spwialixrror~, pp. 4’7&430. 6th Edn. W. B. Saunders Company. Philadelphia, 1975. 5. GEVN. A. S.. DOYLE. J. C.. SCHAEFER.R. E.. CALLAWA’I..E. and Y~AC;ER.C. Lateralized cognitive processes and the electroencephalogram. Science 207, 1006~1007. 1980. 6. GROSS. Y.. FRAXKO. R. and LEWIN. I. EtTects of voluntary eke movements on hemispheric activity and choice of cogmttve mode. Neurops!,cho/oyia 16, 653--657. 1978. 7. HONORLJ. Posture oculaire et attention sklective i des stimuli cutan&. N~ur~jps!choloclicc 20, 717 730. 1981. 8. HONORE, J., BOURDEAU D’HuI. M. and SPARKOW. L. Reduction of cutaneous reaction time by directing eyes towards the source of stimulation. h’ercrop.\!~c,/~~~/~)y~u 27, 367 371. 1989. 9. KESELMAN.H. J. and ROGAX. J. C. Repeated measures Ftests and psychophysmloglcal research: controlling the number of false positives. P.s!,~/loph?‘si~)/~~~/~, 17. 499 503. 1980. 10. KINSBOURNE, M. Eye and head turnmg lndtcates cerebral laterallzation. .S~~irnw 176, 539 541, 1972. Il. KISSBOLRNE. M. and HIS~O~-K. M. AsymmetrIes of dual task performance. In Cerehrul ffcwi.splww .l.~~~rm~rr,~Methods. Theor!, and Applicarions. J. HELLICX (Editor). Academic Press. New Yorh. 1983. 12. LEMPERT. H. and KI~S~OLR~E. M. The effect of visual guidance and hemlspacc on lateralized vocal-manual interference. Neurops~choloyia 23, 691-695. 19X5. 13. NELBAUER. A.. SCHULTER, G. and PFURTSCHFLLFK.G Lateral eye movement5 as an indication of hcmi~pherlc preference. In!. J. Psychoph~~siol. 6, 177 1X4. 198X. 14. OLwIELt). R. C. Edirdwyh ffamfrw.s.~ /wcwror~~. .~c,lo.op\~c,hf,/~,c/itr 1, 97 1f 3. 1971. I j. IX TOFFOL. B., AUTR~T. A.. MAKI;AaI. S. and Rot,?.. S. Influence of lateralized scnsorl-motor and neuropsychological activities on electroenccphalographlc spectral power. Elcc,rr~,crrc,c,pll. diu. .b’c,~r~~,ph!,.\~r~/.75, 200~ 206. 1990. 16. TRESSOLI)I. P. E. Visual hemispace differences retlcct hemisphere asymmetries. .~cl~~~Jp.\?c,h~,l~,Ui~r 25, 625 636. 19x7. 17. TLXXER. D. M.. DAWSOY. S. L.. ROTH. D L. and PI UI.A\I). J. G Regional changes in EEG power and coherence during cognition: intensive study ol” two Indl\iduals. B&r ,Vcwo.wr.99, 564 577. 19X5.
Clinique
DETOFFOL.ALAIN
Neurologique,
IN
AUTRET. ERIC DECIOVAN~‘Iand SYLVIE Roux
CHU
(Received 6 December
Bretonneau,
37044 Tours.
Cedex. France
1989; accepred IO April 1990)
Abstract-This study tests the effect of maintaining right and left lateral gaze during a writing task which preferentially implicates the left hemisphere using an asymmetry parameter calculated from the spectral power of the alpha rhythm (RP- LP/RP+ LP) in a right-handed patient undergoing the same experimental regimen nine times. A six derivatton EEG was recorded. Maintaining left lateral gaze (toward the active hemisphere) removes the lateralization found durmg writing while staring straight ahead whereas maintaining right lateral gaze (toward the side opposite the active hemisphere) results in slightly lower values which are however, not significantly different from those obtained during staring straight ahead. This study adds an eiectrophysiologic aspect to Kinsbournes’s paradigm on gaze position and hemtspheric activation.
INTRODUCTION KINSBOURNE has demonstrated
that eye movements reflected the asymmetric distribution of activation between the cerebral hemispheres [lo. 1 I] and NE~XAUER er al’s. 1131 study has given an EEG validation of this hemispheric asymmetry model. Subsequently, a number of authors have been able to verify the inverse hypothesis: maintaining lateral gaze is accompanied by activation of the contralateral hemisphere. Indeed. significant improvement in tash performance involving a hemisphere has been obtained when gaze is maintained toward the contralateral half of the visual field as compared to maintained straight gaze or gaze toward the ipsilateral half of the visual field [6 8. I?]. During rest. it has previouslv been shown that maintaining lateral gaze produces a reduction in spectral power in the contralateral cerebral hem;sphere [I]: this can be interpreted as regecting hemispheric activation. The present study of EEG spectral analysis tries to detect an electrophysiological manifestation ofthe variation in activatron ofa hemisphere induced by eye position when that hemisphere is implicated in a cognitive task. This investigation IS part of other studies dealing with modifications in cerebral electrogenesis during cognitive activities prcfercnttally involving one of the cerebral hemispheres. There is no consensus on whether or not electrical lateralizatmn exists when a cerebral hemtsphere is preferentially solicited during a cognitive task. According to DAVIMN and EHRLICHMAN [3], preferential involvement of one hemisphere clearly results in reduced spectral power over that hemisphere. However. GEVlN.5 et ul.[5] constder that the lateraiized electrical modifications which accompany cognitive activities are due to non-specific sensorimotor activttics. We have recently developed a method of EEG spectral analysis which measures both the effects of task complexity on spectral power (variation in the log of spectral power) and the preferenttal involvement of one ofthe cerebral hemispheres in the task considered (variation in an asymmetry index RP-LP RPt LP). Accordtngly. when a gtven task prcferenttally solictts the left hemisphere, the power diminishes over that hemisphere while the asymmetry mdex Increases: thrs has been shown during right hand writing in a right-handed subject [I 51. Inversely. when a given task prcfcrcntially solicits the right hemisphere, both the power and the asymmetry index decrcasc over that hemisphere. Using this method. wc have been able to show that sensorimotoractrvrties and pure neuropsychological tasks each play a role m the variatmns in EEG spectral power [ 151. In this study we looked at the variatton+tn a specific parameter: asymmetry in alpha rhythm spectral power. A patient w’as exposed to four different expcrrmental condrtrons: rest, writtng while looking straight ahcad. wrrtrng while gazing to the right. and finally writmg vvhile gazing to the left. WC tested the CITCCIofmaintainingIatcral garc on left hemispheric power after it had been prcvrously “activated” by a wrtttng task. WC hypothcstscd that maintaining right lateral gaze during a writing task mcrca\es the asymmetry paramctcr in comparison to writtng vvhile staring straight ahead. wherea\ marntarnlng left lateral gaze dccrcasc\ this same paramctcr.
SUBJECT
AND METHODS
A 2%kear-old right-handed [ 131 man uithout medicatlonh recclied nlnc EEGs one e\er! other mc,rmng .II precisely the same time. He &as placed in the semi-rechnlng pos~tmn on a bed located an a part~all! soundproofed. dlml! ht room He underwent four different sequence>. each lastlns 2 mm InstructIons wcrc g~\cn \erhall! bct\\ecn the sequences. Eye movements were controlled b) an electro-oculol2ram. The pattent’s chcc Nere clohcd during the sequences. One rest sequence (Rest) alternated uith three acll\e sequences comprised h! one sequence tn\ol\~ng writmg a memorized text with the right hand while loohing straight ahcad (WI. a hccond \V w)ucncc \\hdc maintaming right lateral saze IWRLG I. and tinall!. one sequence durmg left lateral gaze IFVLLG I Sequence progress and patlent performance uere controlled. A latln square nas utlllzed 111order to cllmlnatc an! effect due to passmg order. Six symmetrIcal reference-vertex derlratmns were simultaneou\l\ rccordcd using carefully placed silver chloride electrodes pasted to the scalp: F4Cr F3Cz. ChCz CSCr. P4c‘z P.JCz. (IO 20 International system). The EEG uas amplified usin! an ALVAR polygraph (3 dh: 0.5 2.5 hHz). Sampling \+a\ accomplished in real time at a frequent! of 100 Hz using an anaio g dleltal converter coupled to a hlINCl I 03 and stored on a disc. Before samphng. stgnals were filtered hj anti-aliasmg kiters u hich had a cutting frequent! of 30 H/ and a 24 dB octave attenuatton frequency. The InformatIon was treated using a deferred mode. Succwl\c elementar! 2.56 set spectrums were calculated and each anal!sls period was filtered h> temporal rnultlpllcat~~~n usmp a Hanmng window. An mtcrmedlate spectrum was determlned from the average offourelemcntary spectrumh. We considered Intermediate spectrums having a cgreatcr than 200 !IV’ Hz power in frequency hand 0 I Hz to hc artefacts and dlscarded them. Twelve intermediate spectrums were used to determme a final average spectrum A hlch served to define three parameters: P lthe average power in /IV’ Hz during one sequence). log P and tinall! an asymmetry index (AI1 RP-LP RP+LP. The alpha rhythm wah studled: 7.03 12.89 HZ.
Variance analysis of the average log P values and the asymmetry Index for each sequence and each sglmclric derivation was performed and. hereafter. the Greenhouse Gel>sercorrectlon was applied [Y]. When the slgniticancc attained the P x0.05 threshold. theasymmetry index valuesofeach actliesequence werecompared to rest scqucnw while each staring straight ahead sequence has compared to lateral gaze sequences using the Newman Keulb teht.
RESULTS The results of the variance analqsls of lop P and the absolute spectral powers (P) in {IV’ Hz are shov, n 111Tahlc I. The ANOVA of log P after Greenhouse Geisser correctlon lead to a significant difference in F3Cr dcrivatlon and approach the significance threshold in P3Cz tP=O.Oh). In general. P decreased tn the left hemisphere during acti\c tasks in comparison to rest (except for WRLG and WLLG m PKz). The values of the asymmetry index (AI) arc summarized in Table 1 for each symmetrical dcrllation pair and each activit!. The variance analysis of the Al values. F (3. 24). was significant after Greenhouse Gewer correction 111two ofthe thrccdcrlvations (see the lirst line m Table 2).
Table I. Results of the log P ANOVA and spectral value\ of the r rhythm for each derivation and each sequence. *P
ANOVA t: values
Log P (F 3.241
F4C7
F3C7
CflC.7
I .5Y
5.13’ 0.641
I 52
2 42
2.6’)
13.7 13.Yi I (1 15.1 I
Y.32 7.X3 Y.31 XYI
56.33 63.46 hh.Xh 63.IY
CSCZ
P4Cz
P3CI 3.1 I 0.X I I
I ~rluc\rrf I’ Rest W WRLG WLLG
5.33 4.Yi 5.61 >.>i
5.52 3.17 4.YY 5.03
24.6’) 23.37 2X.5X 2x.47
(al In comparison to rest. Al is slpnificantly Increawd tpowr 1, wcakcr over the Icft hcmt\phcre) during the nrlting task while staring \tralght ahead. I’,() 01 in F4(‘1f 3. I’~005 in 1’4C‘/f’3. th) Comparison between WRLG rest. 41 ih\i~nilicantly tncrca\ed in lc4C‘1F3 1/‘<0,05)during WRLG, hut nonslfnlficant in P4CzP3.
881
NOTE Table 2. derivation correction associated
Values of the AI (asymmetry index) for each symmetrical pair and each sequence. *P
ANOVA AI F (3.24) E values significance after GGC
C6CzC5
P4CzP3
5.20’ 0.508 P
4.06 0.475 ns
5.37* 0.833 P < 0.05
Values 01 AI -0.01 0.08 0.06 0.04
Rest W WRLG WLLG
(c) Comparison between (d) Comparison between (e) Comparison between significant in F4CzF3. (f) Comparison between
F4CzF3
WLLG/rest. W/WRLG. W/WLLG.
0.38 0.45 0.40 0.38
None of the differences were significant. None of the differences were significant. Al was significantly increased in P4CzP3
WRLG/WLLG.
None of the differences
(PcO.05)
during
W. but non
were significant.
In summary, AI for the WRLG sequence is no different from W and is significantly stronger than for the rest sequence in F4CzF3 derivation. AI during the WLLG sequence is significantly weaker for W in P4CzP3 derivation and is no different from the rest sequence. These results are consistent with the hypothesis upon which the study was conducted.
DISCUSSION The method we used concerning spectral analysis. montage selection, tasks and analysis parameters has been discussed elsewhere [I 51. We studied a single subject who was repeatedly exposed to the sameexperimental regimen because EEG values and interhemispheric asymmetrtes show a great deal of individual variation [ 133. Furthermore. TUCKEKrr al. [ 171 have emphasized the interest of EEG single subjects’s studies. We selected an electrophysiological parameter. the asymmetry index R P - LP:RP + LP which gives information on the relative values of spectral power of each hemisphere with respect to the other. A zero index signifies that the power is the same over the two hemispheres, a positive one means that the power is stronger over the right hemisphere or weaker over the left hemisphere. There was a significant difference in AI during writing while staring straight ahead as opposed to during rest and we studied the influence oflateral gaze on this parameter. We limited our analysis to the alpha rhythm since the spectral modifications induced by cognitive activity has been studied most thoroughly in this frequency band [5] which is the most sensitive to the effects of lateral gaze [ 131. We chose writing a memorized text with the right hand for the right-handed patient since this tash is easy to reproduce and because it is generally admitted that. on the basis of a classic lesion model 141. it preferentially solicits the left hemisphere. The Greenhouse Geisser correction is required in repeated-measures ANOVA to avoid false positives results. The “epsilon” values are calculated and giv*en in tables. The observed Fratios (F 3. 24) are referred to as F (3r:. 241:). If only F (31:. 24~) exceeds the conservative critical values. the null hypothesis should be rejected. This study shows that right lateral gazeduring writingdoes not modify the Al valuesin relation tostaringstraight ahead while maintaining left lateral gaze signiticantly reduces AI. which approaches the value obtained during rest. Our hypothesis that AI would be reduced during left lateral gaze has been confirmed. One may postulate that the right hemisphere is”activated” by mamtaining left lateral gaze. tending to balance the power in the two hemispheres. and therefore producing an AI nearly equal to that seen durmg rest. However. the first part ofour hypothesis has not been confirmed: AI values during right lateral gaze are lower than those found during staring straight ahead. withouf being signilicantly different from the latter. This ma\ he due to the experimental conditions chosen. during which lateral gaze does not etlect performance. Indeed. review of the literature has shown that maintaining right lateral gaze during a tash involving the left hemisphere was intended to increase the task performance. Performance was measured by studying reaction times or hy the lateralization score durmg thedichotic listening test [h. 7.8. 173. There is increased performance in relation to hram stem lesions whether lateral gaze is voluntary or pathological [2]. The demonstrated hemispheric activation may he related IO preferential attention focalization to stimuli m the contralateral hcmispace [ 161. However. this study does not address performance modification during a tash according to gaze position or modilication in gaze position during changes in “hemispheric activation”. like in the
classical Kmsbourne paradigm. Rather. it deals wth the variation of a lateralizatlon parameter H hlch has hecn Induced by a wrtting actiwty. Independent of all cognitive activity, It has prebiously been shown that maIntaIned lateral gaze reduced alpha power m the contralateral hemisphere in relation to staring straight ahead [I]. Measuring the activation of a hemisphere by usmg an electrophysiological lateralization parameter based on an analysis of the background EEG only gives a very general impresslon of the activation state. Aasoclatinp two let”1 hemispheric “activating” tasks (right lateral gaze and writing) does not produce a simple additive effect. When the left hemisphere is activated by wrttmg it has reached ceiling and rtght lateral gaze cannot actrrate II an> further. Thus. this study adds an electrophyslological aspect to the importance oflateral gaze in the activation ofthe cerebral hemispheres.
REFERENCES 1. AUTRET. A.. ALVERT. L.. LAFFONT. F. and LARMA\IX. P. Electroencephalographic spectral pwer and lateralized motor activities. Elccrrorncepll. clitl. ,Yeuroph~.sio/. 60, 2% 236. 1985. ?. BELIX, C.. PERR~ER.D.. CAMBIER. J. and LARMASLX P. Influence de la diviatlon pathologique du regard par lesion du tronc &t-bra1 sur I’kquilibre inter-hemlspherique ttudie par le test d’kcoute dichotique non verbal Nruropsyho/ocja 26, 753-758. 1988. 3. DAVIUSON. R. J. and EHRLICHMAS. H. Lateralized cogmtive processes and the electroencephalogram. .%rww 207, 1005-1006. 1980. 4. GARO~~ER.E. ~~ir,~dawwtal.s (!/ A’ruroloyr. T/w Forebrain (‘ewhrcrl Spwialixrror~, pp. 4’7&430. 6th Edn. W. B. Saunders Company. Philadelphia, 1975. 5. GEVN. A. S.. DOYLE. J. C.. SCHAEFER.R. E.. CALLAWA’I..E. and Y~AC;ER.C. Lateralized cognitive processes and the electroencephalogram. Science 207, 1006~1007. 1980. 6. GROSS. Y.. FRAXKO. R. and LEWIN. I. EtTects of voluntary eke movements on hemispheric activity and choice of cogmttve mode. Neurops!,cho/oyia 16, 653--657. 1978. 7. HONORLJ. Posture oculaire et attention sklective i des stimuli cutan&. N~ur~jps!choloclicc 20, 717 730. 1981. 8. HONORE, J., BOURDEAU D’HuI. M. and SPARKOW. L. Reduction of cutaneous reaction time by directing eyes towards the source of stimulation. h’ercrop.\!~c,/~~~/~)y~u 27, 367 371. 1989. 9. KESELMAN.H. J. and ROGAX. J. C. Repeated measures Ftests and psychophysmloglcal research: controlling the number of false positives. P.s!,~/loph?‘si~)/~~~/~, 17. 499 503. 1980. 10. KINSBOURNE, M. Eye and head turnmg lndtcates cerebral laterallzation. .S~~irnw 176, 539 541, 1972. Il. KISSBOLRNE. M. and HIS~O~-K. M. AsymmetrIes of dual task performance. In Cerehrul ffcwi.splww .l.~~~rm~rr,~Methods. Theor!, and Applicarions. J. HELLICX (Editor). Academic Press. New Yorh. 1983. 12. LEMPERT. H. and KI~S~OLR~E. M. The effect of visual guidance and hemlspacc on lateralized vocal-manual interference. Neurops~choloyia 23, 691-695. 19X5. 13. NELBAUER. A.. SCHULTER, G. and PFURTSCHFLLFK.G Lateral eye movement5 as an indication of hcmi~pherlc preference. In!. J. Psychoph~~siol. 6, 177 1X4. 198X. 14. OLwIELt). R. C. Edirdwyh ffamfrw.s.~ /wcwror~~. .~c,lo.op\~c,hf,/~,c/itr 1, 97 1f 3. 1971. I j. IX TOFFOL. B., AUTR~T. A.. MAKI;AaI. S. and Rot,?.. S. Influence of lateralized scnsorl-motor and neuropsychological activities on electroenccphalographlc spectral power. Elcc,rr~,crrc,c,pll. diu. .b’c,~r~~,ph!,.\~r~/.75, 200~ 206. 1990. 16. TRESSOLI)I. P. E. Visual hemispace differences retlcct hemisphere asymmetries. .~cl~~~Jp.\?c,h~,l~,Ui~r 25, 625 636. 19x7. 17. TLXXER. D. M.. DAWSOY. S. L.. ROTH. D L. and PI UI.A\I). J. G Regional changes in EEG power and coherence during cognition: intensive study ol” two Indl\iduals. B&r ,Vcwo.wr.99, 564 577. 19X5.