- Email: [email protected]
The EEG Society
Electroencephalography and clinical Neurophysiology , 1990, 76: 16P-18P Elsevier Scientific Publishers Ireland, Ltd.
16P
EEG 89681
Society proceedings
The E E G Society London,
7 October 1989
Hon. Meetings Secretary." Dr. G. Barrett The National Hospital, Queen Square, London WC1N 3BG (U.K.) (Received for publication: 21 November 1989)
1. Electrophysiology of median nerve lesions in Colles' fracture: a retrospective study. - IC Nagendran and J.G.B. MacLean (St. Bartholomew's Hospital, London) Thirty-five patients with Colles' fracture referred over a period of 12 months for suspected median nerve lesions were investigated electrophysiologically. Seven patients were excluded from the study (5 = other diagnoses, 1 = site of lesion uncertain, and 1 = median lesion probably unrelated to Colles' fracture). Of the 28 remaining patients, 15 showed 'definite' evidence of carpal tunnel syndrome (CTS). Three had 'possible' CTS. Six had median nerve dysfunction at wrist but not CTS. The latter showed primarily axonal changes. In 4, the studies were normal. When the electrophysiological findings of the Colles' fracture associated CTS group were compared with a control group of 15 consecutive patients with CTS due to other causes, the former showed a significantly older age distribution and more severe sensory and motor abnormalities. In the majority, symptoms appeared within a few weeks to a few months after removal of the plaster cast. It is possible that many patients may spontaneously recover without surgical decompression. The possible aetiological factors of this distinct entity are discussed.
2. Physiogenesis and psychogenesis in the post-concussional syndrome. - R.J. MeClelland, G.W. Fenton, A. Montgomery, G. MacFlynn and W. Rutherford (Queen's University, Belfast) The study consisted of a follow-up investigation of 45 consecutive admissions to the City's main A and E Unit with minor head injury (PTA < 24 h). Central conduction times (CCTs) measured by auditory brain-stem potentials were assessed at day 0, 6 weeks and 6 months post injury. A standardised psychiatric assessment (Present State Examination) was made on each case at 6 weeks. At 6 weeks follow-up 36% of the group were diagnosed as cases compared with only 4% cases in a matched control group
( P <0.001). Half the group had significant delays in CCT ( P < 0 . 0 0 1 ) , providing strong evidence of disturbances in brain-stem function. Nevertheless other variables appeared to be contributing even at this stage of recovery. Age and gender: 86% of women over 25 years were diagnosed as cases compared with only 18% of males under 25 ( P < 0.001). Chronic social difficulties were 4 times more frequent among cases than non-cases ( P < 0.001). At 6 months follow-up 3 outcome patterns could be discerned. Over half had recovered within 6 weeks. A minority had persistent symptoms over the 6 months - all with abnormal CCTs. Almost one-third had symptom exacerbation - all had normal CCTs. Psychosocial variables were also evident: chronic cases were on average 10 years older ( P < 0.01) and had twice the level of chronic social difficulties as the recovered group ( P < 0.01). Chronicity was not significantly affected by the presence of compensation claims.
3. EEG and somatosensory evoked potentials (SSEPs) during hypoglycaemia. - R.C. Pottinger, S.A. Amiel, D. Cunnah, H. Archibald, G. Chusney, P.F. Prior and E.A.M. Gale (St. Bartholomew's Hospital, London) To assess cerebral tolerance of hypoglycaemia, EEG (visual and Fourier analysis) and median nerve SSEPs were recorded during 2 m U / k g / m i n insulin clamps in 17 neurologically normal subjects grouped as follows: group I - 4 strictly controlled, insulin-dependent diabetics and 2 insulinoma patients (HbA1 5.5+0.5%); group II - 5 poorly controlled diabetics (HbA1 9+0.5%); group III - 6 normal subjects. Blood glucose (BG) was reduced stepwise from 5 m m o l / l until appearance of EEG delta activity not due to drowsiness, or BG 1.6 mmol/l, and then restored. Lesser hormonal (adrenaline, noradrenaline, cortisol, growth hormone) and 'autonomic' (sweating, palpitations) responses, starting at lower BG, characterised group I, where intermittent frontal rhythmic delta activity (FIRDA) was universal at BG 1.9+0.1 retool/1. In spite of similar BG nadirs, only 2 group II and no group III subjects developed
0013-4649/90/$03.50 © 1990 Elsevier Scientific Publishers Ireland, Ltd.
BRITISH EEG SOCIETY FIRDA, although occasional irregular delta/theta bursts or enhanced posterior temporal slow waves occurred. SSEPs showed slower peripheral conduction in group II but no significant alterations with hypoglycaemia in any group. Despite delayed hormone responses and symptomatic tolerance, hypoglycaemic EEG abnormality is more severe in subjects with strict diabetic control and/or previous hypoglycaemia.
4. EMG 'turns' analysis using short analysis epoch~ results on patients. - A. Forster and P. Fitch (Royal Infirmary, Dundee, and The National Hospital, London) Variations of Willisons' 'turns' analysis remain among the most effective ways of quantifying the EMG interference pattern from a concentric needle electrode. Turns (potential reversals of the signal exceeding 100 gV) and amplitude counts (potential changes exceeding 100 ttV) obtained over 50 msec epochs were plotted against each other on-line. 250 sequential epochs were analysed per site while force was varied; 5 sites were normally examined in vastus medialis and biceps during routine EMG studies. Provisional findings were shown in 1985:157 patients have now been examined using this method. Relatively increased amplitude was found in most cases of neurogenic disorders including 18 motor neurone disease cases. 14 myopathies produced the expected relative increase in turns. In some neurogenic processes the quantitation may appear 'myopathic' from the contribution of polyphasic units to the interference pattern. Guillain Barr6 can produce 'myopathic' quantitative changes. Where patients were reviewed quantitated changes mirrored clinical change. 11 patients with a diagnostic label of 'ME syndrome' showed no clear abnormality. This simple version of the tried and tested 'turns' analysis may have much to offer: it provides an immediate, objective means of interference pattern quantitation.
5. Pituitary hormones and magnetic cortical stimulation. - W.L. Merton, S. Thomas and S.G. Boyd (The Hospital for Sick Children, London) Serum prolactin (PRL) is known to rise after certain types of seizure and following ECT, but not after transcutaneous electrical cortical stimulation (Boyd and De Silva, JNNP, 1986, 49: 954). As magnetic cortical stimulation is better tolerated, many more stimuli are likely to be given. Pituitary hormones were therefore measured in 17 healthy adults before and up to 1.5 h after a train of (10-50) magnetic cortical stimuli at or just above threshold. 11 subjects acted as their own control on a separate occasion. 6 subjects had EEG recordings before and after stimulation. No change in growth hormone (GH) was seen in male subjects. However, in females a rise in GH was seen in 5/11 subjects with stimulation and 2/11 during the control investi-
17P gation. In all but one of these the rise was evident before stimulation. In 2 subjects the GH increase was accompanied by a similar change in PRL and cortisol. No significant change was seen in serum LH, FSH or TSH. The EEG was unaltered by stimulation. The pattern of GH rise seen in some female subjects suggests a stress-related response or a physiological pulse of secretion rather than a direct effect of magnetic stimulation.
6. Human excitatory post-synaptic potentials studied by changes in motoneurone firing probability following magnetic brain stimulation. - K.R. Mills, S.J. Boniface and M. Schubert (Radcliffe Infirmary, Oxford) The duration and shape of the primary peak in a poststimulus time histogram (PSTH) are related to the rising phase of the underlying excitatory post-synaptic potential (EPSP). Fifteen tonically active low threshold motor units in the first dorsal interosseous muscle of 9 healthy subjects were studied during the random delivery of 100-500 transcranial magnetic stimuli at the vertex. The primary peaks in the PSTHs from 13 motor units were found to consist of 1-3 sub-peaks. The mean interval between the first and second sub-peak ranged from 0.8 to 1.7 msec. This interval is comparable to the time between the successive discharges of pyramidal tract neurones produced by a single electrical stimulus delivered to the cortical surface in the baboon (KerneU and Wu Chien-Ping, J. Physiol. (Lond.), 1967, 191: 653). Construction of a cumulative sum curve from the sub-peak regions of the PSTH, normalized to the total number of evoked discharges, produced an estimate of the expected shape of the rising phase of the underlying compound EPSP. Support from the MRC is gratefully acknowledged.
7. Inter-relationship between EEG, pattern VEP and CSF findings in referrals for suspected MS. - E. Poole and W. Payne (Radcliffe Infirmary, Oxford) EEG equipment may be used for VER studies but EEG tends to be neglected, despite its possible functional significance, particularly as to future developments/therapy (and alternative diagnoses). EEG findings over a 3 year period have been analysed (634 referrals) and related to VER and CSF results. EEG was graded as to background and episodic disturbances on a long-standing departmental 1-9 scale (background and dysrhythmic scales); 67% showed at least slight abnormalities in either scale (30% in both). Incidence of abnormal or suspect VERs was 40%; this increased only slightly when EEG was abnormal (maximal when both grades abnormal - 47%). CSF changes typical of MS occurred in 40% of the 298 studied; this incidence also increased only shghtly in relation to EEG abnormalities (up to 50% when both grades abnormal). Incidence of abnormal VERs in this CSF group was 44%; combined CSF/VER abnormahties occurred in 25% but rose to 34% when both EEG
18P
grades were abnormal - only 17% when both normal, giving an abnormal EEG loading ratio of 2.0 ( P < 0.01). Similar ratios occurred in clinical subgroups but with significant differences. EEG provides a predictive input, probably as one of fairly independent facets - more easily dismissed were this not an unsolved chronic disease.
8. The contribution of median nerve SEPs in the functional assessment of the cervical spinal cord in syringomyelia: a study of 24 patients. - D. Restuceia and F. Maugui6re (Neurological Hospital, Lyon, France) Since the early study of Halliday and Wakefield (JNNP, 1963, 26: 211), it has generally been agreed that scalp SEPs are normal in patients with dissociated loss of pain and temperature sensations. U p to now a few patients with abnormal spinal N13 and preserved scalp P14 and N20 have been reported in the literature, but there is no firm evidence, based on group data, that this dissociation can be related to any form of dissociated sensory loss. We studied median nerve SEPs in 24 patients with syringomyelia documented by CT scan or MRI. For the recording of the cervical N13 we used a Cv6-anterior cervical montage, which cancels the potentials generated above the foramen m a g n u m and enhances the amplitude of N13. Scalp far-field and early cortical SEPs were recorded using a non-cephalic reference electrode. Since N13 pathological changes are mainly changes in amplitude we have measured the N13/P9 amplitude ratio in normals and patients and found that it was a reliable index to quantify the amplitude decrease of N13. Absent or reduced N13 was observed in 40 median nerve SEPs (83%) in conjunction with normal P14 and N20 in 30 SEPs. Thus the dissociated loss of the cervical N13 was identified as the most conspicuous SEP feature in syringomyelia. A significant correlation was found between abnormal N13 and loss of pain and temperature sensations, whereas P14 abnormalities correlated well only with loss of joint and touch sensations. In the median nerve territory sensations were either normal (6 cases) or lost only for pain and temperature (24 cases) when SEPs showed abnormal N13 and normal P14. Though it does not reflect directly the post-synaptic activity of spinal cells receiving their inputs from A a and C fibres the N13 potential proved to be a reliable index of spinal cord grey matter dysfunction in syringomyelia.
9. Non-invasive recording of the pattern electroretinogram with an eyelid electrode. - S.R. Butler, S.H. Curry and P.D. Newton (Burden Neurological Institute, Bristol) The pattern reversal electroretinogram (PERG) has a smaller amplitude than that evoked by flash stimulation and
SOCIETY P R O C E E D I N G S
certain methodological compromises have been necessary to record it. Recordings from skin electrodes placed on the cheek at the lower margin of the orbit involve no discomfort and are non-invasive but they yield responses with relatively poor signal to noise levels for practical numbers of stimuli. Recordings from the sclera, such as those made with gold foil, D T L fibre or carbon fibre electrodes inserted under the lower eyelid, resolve clear wave forms with small numbers of stimuli. However, these invasive techniques are vulnerable to eye movement artefact or cause discomfort in some subjects. We have recorded the P E R G from a carbon electrode weighing less than 50 mg which adheres firmly to the lower eyelid by the viscosity and surface tension of a drop of electrode jelly. It is flexibly connected to the headbox lead by a very thin wire which enables eyelid movements to occur without displacing the electrode. This non-invasive technique yields a P E R G whose amplitude approaches that obtained with direct scleral contact while remaining comparatively insensitive to eye movement artefacts. The electrode is easily applied and comfortable for the patient.
10. Electrophysiological evidence of stimulus detection in a 'blind' hemifield. - G. Barrett, S.J.M. Smith and A.D. Towell (The National Hospital, London)
A 28-year-old m a n reported being able to detect movement in the left half-field of vision following a right occipital lobectomy for t u m o u r removal some 3 years earlier. Pattern reversal visual evoked potentials confirmed loss of the normal response to stimulation of the left half-field, whereas reponses to right half-field stimulation were normal. Performance and event-related potentials (ERPs) were recorded in a visual attention task. A 4-square black-and-white check pattern (check size 1 ° 10') was presented for 1 sec with its edge 3 ° 4 0 ' from a central fixation cross in either the left or right half-field. O n half the trials the pattern reversed at 5 Hz, whereas on the other half the pattern remained static. One of the 4 half-field/pattern conditions was designated as a target in each run with the subject pressing a reaction-time key on target presentation. The subject's performance in detecting stimuli in the 'blind' hemifield showed that he could clearly distinguish the reversing pattern from the static ( P < 0.001). ERPs consisted primarily of a large negative-positive complex maximal at a right central electrode. The duration of the negativity and latency of the positivity were longer for decisions about stimuli in the 'blind' field compared with equivalent decisions in the intact field.
16P
EEG 89681
Society proceedings
The E E G Society London,
7 October 1989
Hon. Meetings Secretary." Dr. G. Barrett The National Hospital, Queen Square, London WC1N 3BG (U.K.) (Received for publication: 21 November 1989)
1. Electrophysiology of median nerve lesions in Colles' fracture: a retrospective study. - IC Nagendran and J.G.B. MacLean (St. Bartholomew's Hospital, London) Thirty-five patients with Colles' fracture referred over a period of 12 months for suspected median nerve lesions were investigated electrophysiologically. Seven patients were excluded from the study (5 = other diagnoses, 1 = site of lesion uncertain, and 1 = median lesion probably unrelated to Colles' fracture). Of the 28 remaining patients, 15 showed 'definite' evidence of carpal tunnel syndrome (CTS). Three had 'possible' CTS. Six had median nerve dysfunction at wrist but not CTS. The latter showed primarily axonal changes. In 4, the studies were normal. When the electrophysiological findings of the Colles' fracture associated CTS group were compared with a control group of 15 consecutive patients with CTS due to other causes, the former showed a significantly older age distribution and more severe sensory and motor abnormalities. In the majority, symptoms appeared within a few weeks to a few months after removal of the plaster cast. It is possible that many patients may spontaneously recover without surgical decompression. The possible aetiological factors of this distinct entity are discussed.
2. Physiogenesis and psychogenesis in the post-concussional syndrome. - R.J. MeClelland, G.W. Fenton, A. Montgomery, G. MacFlynn and W. Rutherford (Queen's University, Belfast) The study consisted of a follow-up investigation of 45 consecutive admissions to the City's main A and E Unit with minor head injury (PTA < 24 h). Central conduction times (CCTs) measured by auditory brain-stem potentials were assessed at day 0, 6 weeks and 6 months post injury. A standardised psychiatric assessment (Present State Examination) was made on each case at 6 weeks. At 6 weeks follow-up 36% of the group were diagnosed as cases compared with only 4% cases in a matched control group
( P <0.001). Half the group had significant delays in CCT ( P < 0 . 0 0 1 ) , providing strong evidence of disturbances in brain-stem function. Nevertheless other variables appeared to be contributing even at this stage of recovery. Age and gender: 86% of women over 25 years were diagnosed as cases compared with only 18% of males under 25 ( P < 0.001). Chronic social difficulties were 4 times more frequent among cases than non-cases ( P < 0.001). At 6 months follow-up 3 outcome patterns could be discerned. Over half had recovered within 6 weeks. A minority had persistent symptoms over the 6 months - all with abnormal CCTs. Almost one-third had symptom exacerbation - all had normal CCTs. Psychosocial variables were also evident: chronic cases were on average 10 years older ( P < 0.01) and had twice the level of chronic social difficulties as the recovered group ( P < 0.01). Chronicity was not significantly affected by the presence of compensation claims.
3. EEG and somatosensory evoked potentials (SSEPs) during hypoglycaemia. - R.C. Pottinger, S.A. Amiel, D. Cunnah, H. Archibald, G. Chusney, P.F. Prior and E.A.M. Gale (St. Bartholomew's Hospital, London) To assess cerebral tolerance of hypoglycaemia, EEG (visual and Fourier analysis) and median nerve SSEPs were recorded during 2 m U / k g / m i n insulin clamps in 17 neurologically normal subjects grouped as follows: group I - 4 strictly controlled, insulin-dependent diabetics and 2 insulinoma patients (HbA1 5.5+0.5%); group II - 5 poorly controlled diabetics (HbA1 9+0.5%); group III - 6 normal subjects. Blood glucose (BG) was reduced stepwise from 5 m m o l / l until appearance of EEG delta activity not due to drowsiness, or BG 1.6 mmol/l, and then restored. Lesser hormonal (adrenaline, noradrenaline, cortisol, growth hormone) and 'autonomic' (sweating, palpitations) responses, starting at lower BG, characterised group I, where intermittent frontal rhythmic delta activity (FIRDA) was universal at BG 1.9+0.1 retool/1. In spite of similar BG nadirs, only 2 group II and no group III subjects developed
0013-4649/90/$03.50 © 1990 Elsevier Scientific Publishers Ireland, Ltd.
BRITISH EEG SOCIETY FIRDA, although occasional irregular delta/theta bursts or enhanced posterior temporal slow waves occurred. SSEPs showed slower peripheral conduction in group II but no significant alterations with hypoglycaemia in any group. Despite delayed hormone responses and symptomatic tolerance, hypoglycaemic EEG abnormality is more severe in subjects with strict diabetic control and/or previous hypoglycaemia.
4. EMG 'turns' analysis using short analysis epoch~ results on patients. - A. Forster and P. Fitch (Royal Infirmary, Dundee, and The National Hospital, London) Variations of Willisons' 'turns' analysis remain among the most effective ways of quantifying the EMG interference pattern from a concentric needle electrode. Turns (potential reversals of the signal exceeding 100 gV) and amplitude counts (potential changes exceeding 100 ttV) obtained over 50 msec epochs were plotted against each other on-line. 250 sequential epochs were analysed per site while force was varied; 5 sites were normally examined in vastus medialis and biceps during routine EMG studies. Provisional findings were shown in 1985:157 patients have now been examined using this method. Relatively increased amplitude was found in most cases of neurogenic disorders including 18 motor neurone disease cases. 14 myopathies produced the expected relative increase in turns. In some neurogenic processes the quantitation may appear 'myopathic' from the contribution of polyphasic units to the interference pattern. Guillain Barr6 can produce 'myopathic' quantitative changes. Where patients were reviewed quantitated changes mirrored clinical change. 11 patients with a diagnostic label of 'ME syndrome' showed no clear abnormality. This simple version of the tried and tested 'turns' analysis may have much to offer: it provides an immediate, objective means of interference pattern quantitation.
5. Pituitary hormones and magnetic cortical stimulation. - W.L. Merton, S. Thomas and S.G. Boyd (The Hospital for Sick Children, London) Serum prolactin (PRL) is known to rise after certain types of seizure and following ECT, but not after transcutaneous electrical cortical stimulation (Boyd and De Silva, JNNP, 1986, 49: 954). As magnetic cortical stimulation is better tolerated, many more stimuli are likely to be given. Pituitary hormones were therefore measured in 17 healthy adults before and up to 1.5 h after a train of (10-50) magnetic cortical stimuli at or just above threshold. 11 subjects acted as their own control on a separate occasion. 6 subjects had EEG recordings before and after stimulation. No change in growth hormone (GH) was seen in male subjects. However, in females a rise in GH was seen in 5/11 subjects with stimulation and 2/11 during the control investi-
17P gation. In all but one of these the rise was evident before stimulation. In 2 subjects the GH increase was accompanied by a similar change in PRL and cortisol. No significant change was seen in serum LH, FSH or TSH. The EEG was unaltered by stimulation. The pattern of GH rise seen in some female subjects suggests a stress-related response or a physiological pulse of secretion rather than a direct effect of magnetic stimulation.
6. Human excitatory post-synaptic potentials studied by changes in motoneurone firing probability following magnetic brain stimulation. - K.R. Mills, S.J. Boniface and M. Schubert (Radcliffe Infirmary, Oxford) The duration and shape of the primary peak in a poststimulus time histogram (PSTH) are related to the rising phase of the underlying excitatory post-synaptic potential (EPSP). Fifteen tonically active low threshold motor units in the first dorsal interosseous muscle of 9 healthy subjects were studied during the random delivery of 100-500 transcranial magnetic stimuli at the vertex. The primary peaks in the PSTHs from 13 motor units were found to consist of 1-3 sub-peaks. The mean interval between the first and second sub-peak ranged from 0.8 to 1.7 msec. This interval is comparable to the time between the successive discharges of pyramidal tract neurones produced by a single electrical stimulus delivered to the cortical surface in the baboon (KerneU and Wu Chien-Ping, J. Physiol. (Lond.), 1967, 191: 653). Construction of a cumulative sum curve from the sub-peak regions of the PSTH, normalized to the total number of evoked discharges, produced an estimate of the expected shape of the rising phase of the underlying compound EPSP. Support from the MRC is gratefully acknowledged.
7. Inter-relationship between EEG, pattern VEP and CSF findings in referrals for suspected MS. - E. Poole and W. Payne (Radcliffe Infirmary, Oxford) EEG equipment may be used for VER studies but EEG tends to be neglected, despite its possible functional significance, particularly as to future developments/therapy (and alternative diagnoses). EEG findings over a 3 year period have been analysed (634 referrals) and related to VER and CSF results. EEG was graded as to background and episodic disturbances on a long-standing departmental 1-9 scale (background and dysrhythmic scales); 67% showed at least slight abnormalities in either scale (30% in both). Incidence of abnormal or suspect VERs was 40%; this increased only slightly when EEG was abnormal (maximal when both grades abnormal - 47%). CSF changes typical of MS occurred in 40% of the 298 studied; this incidence also increased only shghtly in relation to EEG abnormalities (up to 50% when both grades abnormal). Incidence of abnormal VERs in this CSF group was 44%; combined CSF/VER abnormahties occurred in 25% but rose to 34% when both EEG
18P
grades were abnormal - only 17% when both normal, giving an abnormal EEG loading ratio of 2.0 ( P < 0.01). Similar ratios occurred in clinical subgroups but with significant differences. EEG provides a predictive input, probably as one of fairly independent facets - more easily dismissed were this not an unsolved chronic disease.
8. The contribution of median nerve SEPs in the functional assessment of the cervical spinal cord in syringomyelia: a study of 24 patients. - D. Restuceia and F. Maugui6re (Neurological Hospital, Lyon, France) Since the early study of Halliday and Wakefield (JNNP, 1963, 26: 211), it has generally been agreed that scalp SEPs are normal in patients with dissociated loss of pain and temperature sensations. U p to now a few patients with abnormal spinal N13 and preserved scalp P14 and N20 have been reported in the literature, but there is no firm evidence, based on group data, that this dissociation can be related to any form of dissociated sensory loss. We studied median nerve SEPs in 24 patients with syringomyelia documented by CT scan or MRI. For the recording of the cervical N13 we used a Cv6-anterior cervical montage, which cancels the potentials generated above the foramen m a g n u m and enhances the amplitude of N13. Scalp far-field and early cortical SEPs were recorded using a non-cephalic reference electrode. Since N13 pathological changes are mainly changes in amplitude we have measured the N13/P9 amplitude ratio in normals and patients and found that it was a reliable index to quantify the amplitude decrease of N13. Absent or reduced N13 was observed in 40 median nerve SEPs (83%) in conjunction with normal P14 and N20 in 30 SEPs. Thus the dissociated loss of the cervical N13 was identified as the most conspicuous SEP feature in syringomyelia. A significant correlation was found between abnormal N13 and loss of pain and temperature sensations, whereas P14 abnormalities correlated well only with loss of joint and touch sensations. In the median nerve territory sensations were either normal (6 cases) or lost only for pain and temperature (24 cases) when SEPs showed abnormal N13 and normal P14. Though it does not reflect directly the post-synaptic activity of spinal cells receiving their inputs from A a and C fibres the N13 potential proved to be a reliable index of spinal cord grey matter dysfunction in syringomyelia.
9. Non-invasive recording of the pattern electroretinogram with an eyelid electrode. - S.R. Butler, S.H. Curry and P.D. Newton (Burden Neurological Institute, Bristol) The pattern reversal electroretinogram (PERG) has a smaller amplitude than that evoked by flash stimulation and
SOCIETY P R O C E E D I N G S
certain methodological compromises have been necessary to record it. Recordings from skin electrodes placed on the cheek at the lower margin of the orbit involve no discomfort and are non-invasive but they yield responses with relatively poor signal to noise levels for practical numbers of stimuli. Recordings from the sclera, such as those made with gold foil, D T L fibre or carbon fibre electrodes inserted under the lower eyelid, resolve clear wave forms with small numbers of stimuli. However, these invasive techniques are vulnerable to eye movement artefact or cause discomfort in some subjects. We have recorded the P E R G from a carbon electrode weighing less than 50 mg which adheres firmly to the lower eyelid by the viscosity and surface tension of a drop of electrode jelly. It is flexibly connected to the headbox lead by a very thin wire which enables eyelid movements to occur without displacing the electrode. This non-invasive technique yields a P E R G whose amplitude approaches that obtained with direct scleral contact while remaining comparatively insensitive to eye movement artefacts. The electrode is easily applied and comfortable for the patient.
10. Electrophysiological evidence of stimulus detection in a 'blind' hemifield. - G. Barrett, S.J.M. Smith and A.D. Towell (The National Hospital, London)
A 28-year-old m a n reported being able to detect movement in the left half-field of vision following a right occipital lobectomy for t u m o u r removal some 3 years earlier. Pattern reversal visual evoked potentials confirmed loss of the normal response to stimulation of the left half-field, whereas reponses to right half-field stimulation were normal. Performance and event-related potentials (ERPs) were recorded in a visual attention task. A 4-square black-and-white check pattern (check size 1 ° 10') was presented for 1 sec with its edge 3 ° 4 0 ' from a central fixation cross in either the left or right half-field. O n half the trials the pattern reversed at 5 Hz, whereas on the other half the pattern remained static. One of the 4 half-field/pattern conditions was designated as a target in each run with the subject pressing a reaction-time key on target presentation. The subject's performance in detecting stimuli in the 'blind' hemifield showed that he could clearly distinguish the reversing pattern from the static ( P < 0.001). ERPs consisted primarily of a large negative-positive complex maximal at a right central electrode. The duration of the negativity and latency of the positivity were longer for decisions about stimuli in the 'blind' field compared with equivalent decisions in the intact field.