NOTE COVERT PROCESSING OF INFORMATION IN HEMIANESTHESIA: A CASE REPORT T. BrochierI, M. Habib 2 and M. Brouchon 1 eU.R.A. C.N.R.S. 1166, Marseille; 2Service de Neurologie, H6pital de la Timone, Marseille) INTRODUCTION This paper deals with the problem of the modular and parallel processing of somesthetic information and its relationship to motor control. Several investigations have attempted to describe the role of somesthetic afferent information in motor control, both in experimentally deafferented animals (Bossom, 1974; Taub, Goldberg and Taub, 1975; Po1it and Bizzi, 1979) and in humans with pathological deafferentation (Rothwell, Traub, Day et aI., 1982; Jeannerod, Michel and Prablanc, 1984). Despite a diversity of approaches, it is generally assumed that large lesions extending into the parietal areas and deafferentations resulting from peripheral damage lead to severely clumsy, inappropriate movements and to incompetence in the use of the hand accompanied by complete loss of sensitivity to tactile and proprioceptive stimulations. In spite of her hemianesthesia, a patient (F.R.) who suffered a complete and sudden loss of somesthetic sensibility in her left arm resulting from a clearly delineated lesion of the primary sensory area (S.I), demonstrated a recovery of motor control which was very surprising in the light of previous data (Rothwell et aI., 1982; Jeannerod et aI., 1984). This observation led us to raise the question of the processes involved in this recovery, and more particularly of the nature of the somesthetic information responsible for the programming and execution of accurate movements. CASE REPORT F.R. was a 69-year-old right-handed woman, who in January, 1991, suddenly presented a complete sensory loss in her left arm and hand following a C.V.A. An initial examination was carried out during the first two weeks following the stroke when the patient was still hospitalized. A more complete check-up was given starting six weeks later. The experimental procedure was applied between the 4th and 5th months post-C.V.A. along with a new set of clinical and paraclinical examinations to test for stabilization of the somesthetic disorders. Somesthesia Six weeks after the stroke, a strong hypoesthesia of the left hand and forearm was demonstrated by clinical examination. The proximal part of the left arm was less affected. In every other part of the body, somesthetic sensitivity was spared. The patient's perception of the intrinsic qualities of a tactually explored object was assessed with several tasks, in which her ability to discriminate smooth versus rough, hard versus soft, warm versus cold, large versus small, and heavy versus light never differed from the chance level. Spontaneous responses were only evoked by painful stimulation, either mechanical or by heat. Tactual localization was evaluated by requiring the patient to point with her normal hand to tactile targets on her "deafferented hand" (even if she reported not to feel anything). The responses Cortex, (1994) 30, 135-144 136 T. Brachier, M. Habib and M. Brauchon were inaccurate but not randomly distributed. Palmar stimulation was distinguished from finger stimulation and the tactile stimulation of a given finger was differentiated from the stimulation of its adjacent fingers. However, 2-point discrimination was never achieved, as the patient remained unable to discriminate between two localizations when both stimuli were applied simultaneously. Complete extinction was observed when the hand or forearm was simultaneously stimulated on the left and right side. In the proprioceptive modality, the patient's position sense was severely impaired. When a finger was passively moved by the experimenter, she could not verbally indicate the final position of the displaced finger, not even roughly (up or down, left or right). In a complementary task, she could not adequately reproduce the movement of the passively displaced finger with her homologous healthy finger. As a consequence of these somesthetic disorders, object discrimination and tactile recognition were extremely affected. These observations were confirmed by recording somesthetic evoked potentials, which showed absence of cortical responses in the primary sensory area when the left medial wrist nerve was stimulated. Further somesthetic examinations at the time of the experimental testing (4 months after the vascular accident) did not show improvement in the patient's performance on these tasks, suggesting a stabilization of the somesthetic disorders. Motricity In spite of the integrity of her motor system, F.R. showed substantial disorders in movement execution with her anesthetized arm during the first few weeks post-C.Y.A. She never used this arm spontaneously in daily life, even for very simple tasks such as combing her hair, eating, or reaching for objects in her immediate surroundings. Six weeks later and during further sessions of examination, she showed substantial improvement in her motor abilities. Her left arm and hand recovered activity and were frequently involved either in bimanual tasks - she reported that she could again dress herself, knit, and wash - or in unimanual actions when she had to pick up, carry, or manipulate objects of various shapes and sizes (a needlepoint canvas, a key, a phone, a bag, etc.). To test these distal motor skills, the patient's performance was analyzed in two tasks in which the fingers had to be controlled independently. In the first task, the blindfolded patient was asked to touch her thumb with each of her other four fingers at least 10 times in succession. During the first few weeks post-c.y.A., F.R. had great difficulty achieving these digital sequences. The movement of a single finger led to movement of the other fingers, indicating the inability of this patient to make independent finger movements while coordinating the action of different parts of the hand. Furthermore, she needed 15 seconds to perform a complete sequence (from forefinger to little finger). Two months post-C.Y.A. a significant improvement was noted. Each finger could be moved independently and with accuracy. Some errors still occurred, e.g. simultaneous activation of the middle finger and the third finger or inaccurate touching of the little finger with the thumb. But even these were spontaneously corrected in the subsequent sequences. The mean duration of a complete sequence was reduced to 2.3 seconds, which was only slightly slower than with the right hand (mean duration = 1.7 seconds). Moreover, movement did not deteriorate when the patient was asked to repeat the sequence 20 times. Similar improvement was observed in a task proposed by Jeannerod et al. (1984) to test motor skillfulness in a hemianesthetic case. The blindfolded patient was asked to crumple a sheet of paper so as to obtain a compact ball in the hollow of the hand. During the first few weeks following her cerebral injury, our patient exhibited a stereotypic movement of opening and closing of the whole hand in performing this task, leading to a rapid loss of the sheet of paper. Six weeks later, F.R. demonstrated a surprising ability to perform the taks. The different fingers were moved in a coordinated manner and with ease. Consequently, the patient was able to manipulate the sheet in the hollow of the hand without dropping it (the sheet was lost in only 2 trials out of 10) and the final result was clearly better. The whole sheet was crumpled in the hand except for one corner due to less effective activity of the little finger. Covert processin[? in hemianesthesia 137 Fig. I - Feb/"lu/l)" /<)<)/. 7:2 weighted Magl/etic Resol/(/I/ce /lIla[?l's. LOl/e of inperilltellsl' sigl/(/I limited to the right post central gyrus and slightly extending into the adjacent white matter. Ri[?ht side of the brain on the left. Cognition F.R. did not show impairment of memory, language, gnosic and praxic abilities, and none of the specific deficits usually associated with parietal damage were observed (visuomotor ataxia, hemineglect, impairment of oculomotor control). Lesion M.R.I. examination, 15 days after the stroke, showed a lesion limited to areas I, 2, 3a and 3b of the ascending parietal gyrus of the right hemisphere. Body trunk representation in the superior part of the gyrus (Woolsey, Erickson and Gilson, 1979) was not affected by the cortical injury. No rostral damage was noted in the precentral gyrus (primary motor area) 138 T. Brochier, M. Habib and M. Brouchon Fig. 2 - Diagram showing the estimated extent of the cortical lesion of the right parietal-lobe; from M.R.1. February 1991. and no caudal damage was noted in the associative parietal areas. Furthermore, the secondary somesthetic area (S.II), located in the inferior part of the postcentral gyrus, was apparently spared by the lesion (Figures I and 2). Procedure The set of symptoms presented by this patient showed a dissociation between her severe deficits in somesthetic sensitivity and her surprising recovery of motor control. This dissociation could be interpreted in terms of relative independency of these two aspects of sensorimotor organization, or as a sign of subconscious processing of somesthetic information. Based on previous work (Paillard and Brouchon, 1968 and 1974, Paillard, BrouchonViton and Jordan, 1978), pointing to the role of infraconscious somesthetic information processing in the control of active movements, we attempted to test the latter hypothesis using the following tasks: detection of tactile stimulation; detection of movement; discrimination of direction; letter recognition. Covert processing in hemianesthesia 139 In each condition, a forced-choice procedure was applied in a series of 30 trials, each signalled by an auditory cue. Tactile and kinaesthetic stimulation was given. Tactile stimulation was applied with a small blunt pin on the palm or forearm, restrained on a support. Proprioceptive stimulation consisted in short passive movements of the forefinger or wrist while the rest of the hand or the forearm was restrained. The tasks were as follows. Detection of Tactile Stimulation Following the auditory cue, the patient's skin was or was not stimulated. The two options were "touch" or "no touch". Detection of Movement Either dynamic or static stimulation of equal duration (0.5 s) was applied. Dynamic tactile stimulation was done by moving the blunt pin along the skin, whereas static stimulation was applied in a single location in the palm. Dynamic proprioceptive stimulation was generated by segmental passive movement (hand or forefinger). In the static condition, the forefinger or hand was held in the position resulting from the last passive movement. The two options were "move" or "no move". Discrimination of Direction The tactile and proprioceptive movements were either directed toward the patient's body or toward the experimenter. Movement length was always the same (about 2 cm). The two options were "back" or "forward". Letter Recognition Two-centimeter-high letters were either drawn on the patient's palm (for tactile stimulation) or outlined in the air with the patient's passively moved forefinger or wrist (for proprioceptive stimulation). The 3 letters, 0, U and S, were chosen in order to prevent recognition based on stimulus length, angles, or number of segments. The only relevant cues for this task were whether the curve was closed or open and the number of direction changes. RESULTS For each test, the number of correct responses and corresponding percentages were calculated. A chi 2 test was used to detect whether the percentages of correct responses were significantly different from the chance level. Right Hand In the tactile detection test, movement detection test, and direction discrimination test, F.R. 's responses were always 100% correct. The percentage of correct responses was also 100% in the letter recognition task carried out in the proprioceptive modality, while in the tactile modality, F.R. recognized 96.6% of the letters written on the hand and 93.3% of those written on the forearm. Left Hand Detection of Tactile Stimulation The percentage of correct responses reached 73.6% for hand stimulation and 70.3% for forearm stimulation. Although the error rate was high, F.R. could detect stimulation occurrence significantly better than chance. T. Brochier, M. Habib and M. Brouchon 140 Detection of tactile stimulation c Detection of movement E " "E E OJ OJ '0 '0 .2. 13 ~ oo .2. 13 ~ oo ... ... H Tactile Discrimination of direction Kinaesthetic Letter recognition E E " OJ '0 .2. ~ oo ... H Tactile H Kinaesthetic Tactile Kinaesthetic Fig. 3 - Percentage of correct responses in the four experimental tasks, both for tactile and kinaesthetic modalities. H: Hand stimulation. F.A.: Forearm stimulation. F.F.: Forefinger stimulation. W. Wrist stimulation. Arrow indicates the percentages of correct responses that is significantly different from chance (p