366
Eh'ctroencephak~graphy and climcul Neurophysiology, 81 t 1991 ) 366-376 :~: 1991 Elsevier Scientific Publishers Ireland, Ltd. ()924-980X/91/$03.50 ADONIS 09249811X9100097F
ELMOCO 89192
A t t e n ~ t i o n in detection of somatosensory stimufi by transcranial magnetic stimulation Leonardo G. Cohen, Stefania BandineHi, Susumu Sato, Conrad Kufta and Mark Hallett Human Cortical Physiology Unit, Human Motor Control Section, Medical NeurologT Branch, National Institute ~ff"Neurological Disorders and Stroke, National Institutes o f Health, Bethesda, MD 20892 (U.SIA. 1 (Accepted for publication: 4 March 1991 )
Summary
Effects of magnetic stimulation (MS) of the scalp and direct cortical electrical stimulation on detection of an electrical stimulus to the index finger (SI) were studied in 7 normal volunteers and a patient with epilepsy. Detection of somatosensory stimuli was attenuated when MS was delivered 200 msec before S1, was blocked when MS was delivered simultaneously to and 20 msec after SI. and was fully recovered when MS was delivered 200 msec after S1. This effect showed topographic specificity, being produced by scalp stimulation of restricted scalp positions eontralateral to the finger stimulated, was maximal with low intensities of finger stimulation and high intensities of MS (usually over that required for motor threshold), and could also be produced in the absence of motor evoked responses in a peripheral hand muscle. These restdts show that a focal cortical stimulus can briefly attenuate detection of somatosenso~' stimuli before, during, and after cortical arrival of a somatosensory afferent volley. Several different mechanisms probably contribute to this phenomenon.
Key words: Somatosensory stimuli; Transcranial magnetic stimulation
Direct cortical electrical stimulation, as studied during brain surgery, can produce feelings of deafness with delivery over the auditory cortex (Penfield and Rasmussen 1950), vision of lights or blindness with delivery over the calcarine cortex (Penfield 1958), and speech arrest with delivery over Broca's area. Such stimulation can also block detection of a peripheral stimulus with delivery over the somatosensory areas (Libet et al. 1972). Faradic stimulation delivered over the postcentral gyrus was reported to induce paresthesias (Cushing 1909). These studies had the limitations that they could only be done during surgery in patients with brain disease. More recently, it has been shown that a brief electrical or magnetic stimulus applied over the scalp can activate areas of the motor cortex and evoke brief muscle twitches (Merton and Morton 1980; Barker et al. 1985). Such stimuli provide a non-invasive means of studying the speed of conduction in human central motor pathways (Mills et al. 1987). Negative phenomena can also result from this type of stimulation. An appropriately timed magnetic stimulus delivered to the scalp can delay an expected motor response (Day et al. 1989). Similarly, an appropriately timed stimulus delivered over occipital areas can disrupt visual processing (Amassian et al. 1989a). These findings show the possi-
Correspondence to: Leonardo G. Cohen, M.D., Building 10. Room 5N226, NINDS, NIH, Bethesda, MD 20892 (U.S.A.).
bility of transiently blocking certain specific functions in the central nervous system using non-invasive techniques. The present study investigates the effects of magnetic stimulation on behavioral detection of somatosensory stimuli to the skin (see Libet (1987) for a review of the concepts of signal detection versus conscious awareness).
Subjects and methods We studied the effects of transcranial magnetic stimulation on detection of somatosensory stimuli in 7 normal volunteers (5 men and 2 women) and 1 patient with epilepsy who had a grid of subdural electrodes implanted over the left sensorimotor areas of the brain. All subjects were naive to the purposes of the experiments, and all gave their informed consent for participation in the study. Their ages ranged from 28 to 60 years (mean, 38 years). The volunteers were tested in a quiet room while lying on a comfortable bed and the patient in his hospital room bed, Somatosensory stimuli (S1), as percutaneous electrical stimuli, were delivered through bipolar ring electrodes to digital nerves. Stimulus intensities ranged from t to 6 times sensory threshold (ST). Magnetic stimulation (MS) was delivered with a Cadwell MES-10 magnetic stimulator and 2 different magnetic coils: (1) a circular coil, 9 cm in diameter,
MAGNETIC STIMULATION AND SOMATOSENSORY DETECTION which evokes a relatively large electric field, and (2) an 8-shaped coil, each component measuring 4.5 cm in diameter. The latter coil evokes a more focal electric field under its center, the intersection of both components of the coil (Cohen et al. 1990). Stimulus intensity, location of stimulation, and position of the magnetic coils varied for each experiment.
Experiment 1 MS was delivered at different time intervals in relation to S1. Seven subjects participated in this study. S1 was delivered to the right index finger with an intensity of 1.5 ST. A circular coil was used to deliver MS to the scalp overlying contralateral sensorimotor areas (with the border of the coil over C3) at different times before and after $1 (Fig. 1A). The intensities of MS were those necessary to evoke an electromyographic ( E M G ) response smaller than 500 /xV from right abductor pollicis brevis (APB) when the coil was positioned at the optimal scalp location for motor stimulation in each subject. Up to 300 scalp stimuli were delivered to each subject for this experiment. At each time interval, 3 different conditions were tested: finger stimulation only (condition a), scalp stimulation only (condition b), and simultaneous finger and scalp stimulation (condition c) (Fig. 1B). Ten trials were randomly performed for each condition at each time interval. For each trial, the subjects were instructed to answer 'yes' if they felt
367 the finger stimulus and ' n o ' if they did not feel it or were unsure. Condition a gave the subject continuing feedback about the target-finger stimulus and the experimenter control for possible decrease in the finger stimulus intensity during the study. Condition b controlled for the production of false positive responses ('yes' when there was no finger stimulation). In another experiment, performed in only 1 subject, he was allowed to answer 'different' if he felt the finger stimulus but it was different from that delivered in condition a. Two subjects, who answered ' n o ' in 1 or more of 10 trials in condition a and 'yes' in 1 or more of 10 trials in condition b were considered unreliable and were not included in the study. R e p e a t e d measures analysis of variance ( A N O V A ) and the Scheffe test were used to test attenuation in detection of somatosensory stimuli at different time intervals of - 450, - 400, - 300, - 200, - 100, 0, 20, 100, 200, and 300 msec.
Experiment 2 Both index fingers were stimulated simultaneously. Seven subjects participated in this experiment. S1 was delivered to both index fingers simultaneously (condition a), to the left index finger only (condition b), and to the right index finger only (condition c) with an intensity of 1.5 ST, or no stimulation was delivered (condition d). A circular coil was used to deliver MS to the scalp overlying contralateral sensorimotor areas
B
A
a
I
Finger stimulation
I
Magnetic stimulation
I
Finger stimulation
l
=._ Magnetic " stimulation
Finger stimulation
! b Magnetic stimulation
-400
0 I TIME (msec)
+300
C ~
Fig. 1. A: paradigm used for experiment 1. Magnetic stimulation was' randomly delivered at different times (msec) before and after electrical stimulation of the finger (S1). B: at each time interval, 3 different conditions were tested: finger stimulation only (a), magnetic stimulation of the scalp only (b), and, the test condition, simultaneous finger and scalp stimulation (c).
I..G. COHEN ET AL.
368 (with the border of the coil over C3) 20 msec after S1 in each of the 4 randomly presented conditions (10 trials each). The intensity of MS in this experiment was the same as in experiment 1. For each trial, the subjects were asked to indicate whether they felt the finger stimulus and to identify which finger was stimulated. One-factor A N O V A was used to compare detection of somatosensory stimuli from the fingers in each condition.
CORONAL
SAGITTAL
Experiment 3 This experiment was designed to study the effects of MS on detection of a pair of electrical stimuli delivered to the contralateral index finger. Five subjects participated in this study. A pair of electrical stimuli (S1-$2), separated by different time intervals, was delivered to the right index finger. S1 and $2 stimulus intensities were 1.5 ST. A circular coil was used to deliver MS to the scalp over the left sensorimotor areas (with the border of the coil over C3) at different times after $2. Similar to the paradigm in experiment 1, 3 different conditions were randomly presented: finger stimulation only (condition a), scalp stimulation only (condition b), and simultaneous finger and scalp stimulation (condition c). For each trial, the subjects were asked to indicate whether they felt 1 stimulus, a pair of stimuli, or no stimulus to the right index finger.
Fig. 2. Butterfly-shaped coil used in experiment 5. Magnetic stimutalion was delivered along coronal and sagittal axes at scalp positions situated l em apart.
feel it or were unsure. The scalp positions at which stimulation evoked the largest responses from right APB were compared with those at which stimulation was more effective in attenuating detection of somatosensory stimuli.
Experiment 4 The effects of stimulating different scalp positions were studied in 4 subjects participating in this study. S1 was delivered to the right index finger with an intensity of 1.5 ST. An 8-shaped coil was used to deliver MS to the scalp over C3 and Cz (with the coil centered over the target locations) 20 msec after S1. The 3 conditions described in experiment 1 were also tested in this experiment. Ten trials were randomly performed for conditions a, b, and c. For each trial, subjects had to answer "yes' if they felt the finger stimulus and 'no' if they did not feel it or were unsure.
Experiment 5 This experiment involved the study of motor evoked potentials to MS and detection of somatosensory stimuli. Four subjects participated in this experiment. S1 was delivered to the right index finger with an intensity of 1.5 ST. An 8-shaped coil was used to deliver MS to different scalp positions situated 1 cm apart along a coronal and sagittal axis (with the coil centered over the target locations) 20 msec after S1 (Fig. 2). The cross point of both axes was C3. E M G responses were recorded from right APB. The 3 conditions described in experiment 1 were also tested in this experiment. Ten trials were randomly performed for conditions a, b, and c. For each trial, the subjects had to answer 'yes' if they felt the finger stimulus and 'no' if they did not
Experiment 6 The effects of varying the intensity of the electrical (finger) stimulus were studied in experiment 6. Four subjects participated in this study. S1 was delivered randomly to the right index finger with an intensity ranging from 1 to 6 times ST (10 trials each). A circular coil was used to deliver MS to the scalp over C3 2(I msec after SI. The 3 conditions described in experiment I were also tested in this experiment. Ten trials were randomly performed for each condition a, b, and c. For each trial, the subjects had to answer ,yes' if they felt the finger stimulus and 'no' if they did not feel it or were unsure.
Experiment 7 The effects of varying the intensity of the magnetic (scalp) stimulus were studied. Four subjects =participated in this experiment. S1 was delivered to the right index finger with an intensity of 1.5 ST. MS was delivered randomly to the scalp over C3 at different stimulus intensities expressed as a percentage of maximal output of the magnetic stimulator and also in relation to the intensity capable of evoking the smallest observable E M G response (motor threshold) from the contralateral APB, For each trial, the subjects had to answer 'yes' if they felt the finger stimulus and 'no' if they did not feel it or were unsure.
MAGNETIC STIMULATION AND SOMATOSENSORY DETECTION
369
Experiment 8 To determine if detection of $1 can be blocked by a single electric shock in the absence of motor responses from the stimulated finger, we studied 1 patient with intractable seizures who underwent brain surgery and had a grid of subdural electrodes implanted over the sensorimotor areas. S1 was delivered to the right fifth finger with an intensity of 1.5 ST. Cortical stimulation consisted of single biphasic pulses with different intensities and durations. The interelectrode distance in the grid was 0.5 cm. The intensity of stimulation ranged from 3 to 30 mA. For the purpose of this experiment, single biphasic pulses were delivered through 2 subdural electrodes (bipolar montage) overlying the fifth finger sensory representation area (stimulation through these electrodes evoked paresthesias but not motor responses in the fifth finger). S1 was delivered 20 msec before the cortical stimulus. The 3 conditions described in experiment 1 were tested. We also delivered S1 while one of the authors rubbed the patient's palm and fingers with his hand. The purpose of this condition was to determine if detection of S1 was preserved with simultaneous somatosensory tactile interference (condition d). Ten trials each were randomly performed for conditions a, b, c, and d. For each trial, the subject had to answer 'yes' if he felt the finger stimulus and ' n o ' if he did not feel it or was unsure.
Results
Paresthesias induced by magnetic stimulation Three subjects consistently reported tingling sensations following the magnetic stimulus. Two of them could not differentiate satisfactorily sensations produced by MS from those produced by cutaneous finger stimulation. These subjects were eliminated since they reported excessive false positive responses.
Delivery of MS at different time intervals in relation to $1 (experiment 1) Detection of S1 was partially attenuated when MS was delivered between 300 and 100 msec before S1, was completely blocked when MS was delivered simultaneously with and 20 msec after S1 in every subject (Fig. 3), and was partially attenuated when MS was delivered between 20 and 200 msec after S1. Every subject reported finger stimulation when S1 was delivered alone, and did not feel any stimulation when MS was delivered alone, indicating that the intensity of somatosensory stimulation was always sufficient to allow detection and that there were no false positive responses. One subject had the option of also answering 'different' when stimulation was felt, and experiment 1 was repeated on 2 different days (Fig. 4). When the
SIMULTIANEOUS
'!I
10
8 iI d iii Ii
6
i i
')iii!
7- 4 O9
~s 6
2
z
:~nger stim. before
Scalp stlrn, before -600
-400
-200
0
200
400
TIME ( m s e c )
Fig. 3. Attenuation in detection of somatosensory stimuli as a function of time intervals between finger and scalp stimulation in 7 subjects. Each curve represents the responses to the presentation of condition c from 1 subject. (Responses to condition a were always 'yes,' that is, stimuli to the finger were clearly felt, and to condition b were always 'no,' that is, stimuli to the finger were never felt.) Each point in the curve has been constructed on the basis of 30 trials under the 3 conditions (a, b and c) described. The vertical dotted line shows 0 msec, the time when finger and scalp were stimulated simultaneously. To the left are time intervals where magnetic stimulation preceded finger stimulation. To the right are time intervals where finger stimulation preceded MS. Note that detection of somatosensory stimuli was intact when MS was delivered 400 msec before finger stimulation. It progressivelydeteriorated when MS was delivered between 300 and 100 msec before finger stimulation, and it was blocked when MS was delivered simultaneously and up to 20 msec after finger stimulation. It progressively recovered in the following 200 msec. * P < 0.01, * * P < 0.05.
answer 'different' was counted as 'yes,' it indicated detection block (dashed lines in Fig. 4). When it was counted as 'no,' it indicated detection distortion, as well as detection block (solid lines in Fig. 4). The area between both curves shows the time of 'detection distortion' for the same subject in 2 different sessions.
Simultaneous stimulation of both index fingers (experiment 2) When S1 was presented to both index fingers simultaneously (condition a) and to the left index finger only (condition b, ipsilateral to the scalp position stimulated), the 7 subjects felt S1 in the finger ipsilateral to the scalp position stimulated ( P ~
Eh'ctroencephak~graphy and climcul Neurophysiology, 81 t 1991 ) 366-376 :~: 1991 Elsevier Scientific Publishers Ireland, Ltd. ()924-980X/91/$03.50 ADONIS 09249811X9100097F
ELMOCO 89192
A t t e n ~ t i o n in detection of somatosensory stimufi by transcranial magnetic stimulation Leonardo G. Cohen, Stefania BandineHi, Susumu Sato, Conrad Kufta and Mark Hallett Human Cortical Physiology Unit, Human Motor Control Section, Medical NeurologT Branch, National Institute ~ff"Neurological Disorders and Stroke, National Institutes o f Health, Bethesda, MD 20892 (U.SIA. 1 (Accepted for publication: 4 March 1991 )
Summary
Effects of magnetic stimulation (MS) of the scalp and direct cortical electrical stimulation on detection of an electrical stimulus to the index finger (SI) were studied in 7 normal volunteers and a patient with epilepsy. Detection of somatosensory stimuli was attenuated when MS was delivered 200 msec before S1, was blocked when MS was delivered simultaneously to and 20 msec after SI. and was fully recovered when MS was delivered 200 msec after S1. This effect showed topographic specificity, being produced by scalp stimulation of restricted scalp positions eontralateral to the finger stimulated, was maximal with low intensities of finger stimulation and high intensities of MS (usually over that required for motor threshold), and could also be produced in the absence of motor evoked responses in a peripheral hand muscle. These restdts show that a focal cortical stimulus can briefly attenuate detection of somatosenso~' stimuli before, during, and after cortical arrival of a somatosensory afferent volley. Several different mechanisms probably contribute to this phenomenon.
Key words: Somatosensory stimuli; Transcranial magnetic stimulation
Direct cortical electrical stimulation, as studied during brain surgery, can produce feelings of deafness with delivery over the auditory cortex (Penfield and Rasmussen 1950), vision of lights or blindness with delivery over the calcarine cortex (Penfield 1958), and speech arrest with delivery over Broca's area. Such stimulation can also block detection of a peripheral stimulus with delivery over the somatosensory areas (Libet et al. 1972). Faradic stimulation delivered over the postcentral gyrus was reported to induce paresthesias (Cushing 1909). These studies had the limitations that they could only be done during surgery in patients with brain disease. More recently, it has been shown that a brief electrical or magnetic stimulus applied over the scalp can activate areas of the motor cortex and evoke brief muscle twitches (Merton and Morton 1980; Barker et al. 1985). Such stimuli provide a non-invasive means of studying the speed of conduction in human central motor pathways (Mills et al. 1987). Negative phenomena can also result from this type of stimulation. An appropriately timed magnetic stimulus delivered to the scalp can delay an expected motor response (Day et al. 1989). Similarly, an appropriately timed stimulus delivered over occipital areas can disrupt visual processing (Amassian et al. 1989a). These findings show the possi-
Correspondence to: Leonardo G. Cohen, M.D., Building 10. Room 5N226, NINDS, NIH, Bethesda, MD 20892 (U.S.A.).
bility of transiently blocking certain specific functions in the central nervous system using non-invasive techniques. The present study investigates the effects of magnetic stimulation on behavioral detection of somatosensory stimuli to the skin (see Libet (1987) for a review of the concepts of signal detection versus conscious awareness).
Subjects and methods We studied the effects of transcranial magnetic stimulation on detection of somatosensory stimuli in 7 normal volunteers (5 men and 2 women) and 1 patient with epilepsy who had a grid of subdural electrodes implanted over the left sensorimotor areas of the brain. All subjects were naive to the purposes of the experiments, and all gave their informed consent for participation in the study. Their ages ranged from 28 to 60 years (mean, 38 years). The volunteers were tested in a quiet room while lying on a comfortable bed and the patient in his hospital room bed, Somatosensory stimuli (S1), as percutaneous electrical stimuli, were delivered through bipolar ring electrodes to digital nerves. Stimulus intensities ranged from t to 6 times sensory threshold (ST). Magnetic stimulation (MS) was delivered with a Cadwell MES-10 magnetic stimulator and 2 different magnetic coils: (1) a circular coil, 9 cm in diameter,
MAGNETIC STIMULATION AND SOMATOSENSORY DETECTION which evokes a relatively large electric field, and (2) an 8-shaped coil, each component measuring 4.5 cm in diameter. The latter coil evokes a more focal electric field under its center, the intersection of both components of the coil (Cohen et al. 1990). Stimulus intensity, location of stimulation, and position of the magnetic coils varied for each experiment.
Experiment 1 MS was delivered at different time intervals in relation to S1. Seven subjects participated in this study. S1 was delivered to the right index finger with an intensity of 1.5 ST. A circular coil was used to deliver MS to the scalp overlying contralateral sensorimotor areas (with the border of the coil over C3) at different times before and after $1 (Fig. 1A). The intensities of MS were those necessary to evoke an electromyographic ( E M G ) response smaller than 500 /xV from right abductor pollicis brevis (APB) when the coil was positioned at the optimal scalp location for motor stimulation in each subject. Up to 300 scalp stimuli were delivered to each subject for this experiment. At each time interval, 3 different conditions were tested: finger stimulation only (condition a), scalp stimulation only (condition b), and simultaneous finger and scalp stimulation (condition c) (Fig. 1B). Ten trials were randomly performed for each condition at each time interval. For each trial, the subjects were instructed to answer 'yes' if they felt
367 the finger stimulus and ' n o ' if they did not feel it or were unsure. Condition a gave the subject continuing feedback about the target-finger stimulus and the experimenter control for possible decrease in the finger stimulus intensity during the study. Condition b controlled for the production of false positive responses ('yes' when there was no finger stimulation). In another experiment, performed in only 1 subject, he was allowed to answer 'different' if he felt the finger stimulus but it was different from that delivered in condition a. Two subjects, who answered ' n o ' in 1 or more of 10 trials in condition a and 'yes' in 1 or more of 10 trials in condition b were considered unreliable and were not included in the study. R e p e a t e d measures analysis of variance ( A N O V A ) and the Scheffe test were used to test attenuation in detection of somatosensory stimuli at different time intervals of - 450, - 400, - 300, - 200, - 100, 0, 20, 100, 200, and 300 msec.
Experiment 2 Both index fingers were stimulated simultaneously. Seven subjects participated in this experiment. S1 was delivered to both index fingers simultaneously (condition a), to the left index finger only (condition b), and to the right index finger only (condition c) with an intensity of 1.5 ST, or no stimulation was delivered (condition d). A circular coil was used to deliver MS to the scalp overlying contralateral sensorimotor areas
B
A
a
I
Finger stimulation
I
Magnetic stimulation
I
Finger stimulation
l
=._ Magnetic " stimulation
Finger stimulation
! b Magnetic stimulation
-400
0 I TIME (msec)
+300
C ~
Fig. 1. A: paradigm used for experiment 1. Magnetic stimulation was' randomly delivered at different times (msec) before and after electrical stimulation of the finger (S1). B: at each time interval, 3 different conditions were tested: finger stimulation only (a), magnetic stimulation of the scalp only (b), and, the test condition, simultaneous finger and scalp stimulation (c).
I..G. COHEN ET AL.
368 (with the border of the coil over C3) 20 msec after S1 in each of the 4 randomly presented conditions (10 trials each). The intensity of MS in this experiment was the same as in experiment 1. For each trial, the subjects were asked to indicate whether they felt the finger stimulus and to identify which finger was stimulated. One-factor A N O V A was used to compare detection of somatosensory stimuli from the fingers in each condition.
CORONAL
SAGITTAL
Experiment 3 This experiment was designed to study the effects of MS on detection of a pair of electrical stimuli delivered to the contralateral index finger. Five subjects participated in this study. A pair of electrical stimuli (S1-$2), separated by different time intervals, was delivered to the right index finger. S1 and $2 stimulus intensities were 1.5 ST. A circular coil was used to deliver MS to the scalp over the left sensorimotor areas (with the border of the coil over C3) at different times after $2. Similar to the paradigm in experiment 1, 3 different conditions were randomly presented: finger stimulation only (condition a), scalp stimulation only (condition b), and simultaneous finger and scalp stimulation (condition c). For each trial, the subjects were asked to indicate whether they felt 1 stimulus, a pair of stimuli, or no stimulus to the right index finger.
Fig. 2. Butterfly-shaped coil used in experiment 5. Magnetic stimutalion was delivered along coronal and sagittal axes at scalp positions situated l em apart.
feel it or were unsure. The scalp positions at which stimulation evoked the largest responses from right APB were compared with those at which stimulation was more effective in attenuating detection of somatosensory stimuli.
Experiment 4 The effects of stimulating different scalp positions were studied in 4 subjects participating in this study. S1 was delivered to the right index finger with an intensity of 1.5 ST. An 8-shaped coil was used to deliver MS to the scalp over C3 and Cz (with the coil centered over the target locations) 20 msec after S1. The 3 conditions described in experiment 1 were also tested in this experiment. Ten trials were randomly performed for conditions a, b, and c. For each trial, subjects had to answer "yes' if they felt the finger stimulus and 'no' if they did not feel it or were unsure.
Experiment 5 This experiment involved the study of motor evoked potentials to MS and detection of somatosensory stimuli. Four subjects participated in this experiment. S1 was delivered to the right index finger with an intensity of 1.5 ST. An 8-shaped coil was used to deliver MS to different scalp positions situated 1 cm apart along a coronal and sagittal axis (with the coil centered over the target locations) 20 msec after S1 (Fig. 2). The cross point of both axes was C3. E M G responses were recorded from right APB. The 3 conditions described in experiment 1 were also tested in this experiment. Ten trials were randomly performed for conditions a, b, and c. For each trial, the subjects had to answer 'yes' if they felt the finger stimulus and 'no' if they did not
Experiment 6 The effects of varying the intensity of the electrical (finger) stimulus were studied in experiment 6. Four subjects participated in this study. S1 was delivered randomly to the right index finger with an intensity ranging from 1 to 6 times ST (10 trials each). A circular coil was used to deliver MS to the scalp over C3 2(I msec after SI. The 3 conditions described in experiment I were also tested in this experiment. Ten trials were randomly performed for each condition a, b, and c. For each trial, the subjects had to answer ,yes' if they felt the finger stimulus and 'no' if they did not feel it or were unsure.
Experiment 7 The effects of varying the intensity of the magnetic (scalp) stimulus were studied. Four subjects =participated in this experiment. S1 was delivered to the right index finger with an intensity of 1.5 ST. MS was delivered randomly to the scalp over C3 at different stimulus intensities expressed as a percentage of maximal output of the magnetic stimulator and also in relation to the intensity capable of evoking the smallest observable E M G response (motor threshold) from the contralateral APB, For each trial, the subjects had to answer 'yes' if they felt the finger stimulus and 'no' if they did not feel it or were unsure.
MAGNETIC STIMULATION AND SOMATOSENSORY DETECTION
369
Experiment 8 To determine if detection of $1 can be blocked by a single electric shock in the absence of motor responses from the stimulated finger, we studied 1 patient with intractable seizures who underwent brain surgery and had a grid of subdural electrodes implanted over the sensorimotor areas. S1 was delivered to the right fifth finger with an intensity of 1.5 ST. Cortical stimulation consisted of single biphasic pulses with different intensities and durations. The interelectrode distance in the grid was 0.5 cm. The intensity of stimulation ranged from 3 to 30 mA. For the purpose of this experiment, single biphasic pulses were delivered through 2 subdural electrodes (bipolar montage) overlying the fifth finger sensory representation area (stimulation through these electrodes evoked paresthesias but not motor responses in the fifth finger). S1 was delivered 20 msec before the cortical stimulus. The 3 conditions described in experiment 1 were tested. We also delivered S1 while one of the authors rubbed the patient's palm and fingers with his hand. The purpose of this condition was to determine if detection of S1 was preserved with simultaneous somatosensory tactile interference (condition d). Ten trials each were randomly performed for conditions a, b, c, and d. For each trial, the subject had to answer 'yes' if he felt the finger stimulus and ' n o ' if he did not feel it or was unsure.
Results
Paresthesias induced by magnetic stimulation Three subjects consistently reported tingling sensations following the magnetic stimulus. Two of them could not differentiate satisfactorily sensations produced by MS from those produced by cutaneous finger stimulation. These subjects were eliminated since they reported excessive false positive responses.
Delivery of MS at different time intervals in relation to $1 (experiment 1) Detection of S1 was partially attenuated when MS was delivered between 300 and 100 msec before S1, was completely blocked when MS was delivered simultaneously with and 20 msec after S1 in every subject (Fig. 3), and was partially attenuated when MS was delivered between 20 and 200 msec after S1. Every subject reported finger stimulation when S1 was delivered alone, and did not feel any stimulation when MS was delivered alone, indicating that the intensity of somatosensory stimulation was always sufficient to allow detection and that there were no false positive responses. One subject had the option of also answering 'different' when stimulation was felt, and experiment 1 was repeated on 2 different days (Fig. 4). When the
SIMULTIANEOUS
'!I
10
8 iI d iii Ii
6
i i
')iii!
7- 4 O9
~s 6
2
z
:~nger stim. before
Scalp stlrn, before -600
-400
-200
0
200
400
TIME ( m s e c )
Fig. 3. Attenuation in detection of somatosensory stimuli as a function of time intervals between finger and scalp stimulation in 7 subjects. Each curve represents the responses to the presentation of condition c from 1 subject. (Responses to condition a were always 'yes,' that is, stimuli to the finger were clearly felt, and to condition b were always 'no,' that is, stimuli to the finger were never felt.) Each point in the curve has been constructed on the basis of 30 trials under the 3 conditions (a, b and c) described. The vertical dotted line shows 0 msec, the time when finger and scalp were stimulated simultaneously. To the left are time intervals where magnetic stimulation preceded finger stimulation. To the right are time intervals where finger stimulation preceded MS. Note that detection of somatosensory stimuli was intact when MS was delivered 400 msec before finger stimulation. It progressivelydeteriorated when MS was delivered between 300 and 100 msec before finger stimulation, and it was blocked when MS was delivered simultaneously and up to 20 msec after finger stimulation. It progressively recovered in the following 200 msec. * P < 0.01, * * P < 0.05.
answer 'different' was counted as 'yes,' it indicated detection block (dashed lines in Fig. 4). When it was counted as 'no,' it indicated detection distortion, as well as detection block (solid lines in Fig. 4). The area between both curves shows the time of 'detection distortion' for the same subject in 2 different sessions.
Simultaneous stimulation of both index fingers (experiment 2) When S1 was presented to both index fingers simultaneously (condition a) and to the left index finger only (condition b, ipsilateral to the scalp position stimulated), the 7 subjects felt S1 in the finger ipsilateral to the scalp position stimulated ( P ~
