Eur Arch Psychiatry Clin Neurosci (1995) 246:53-59 9 Springer-Vertag 1995 Helmut Hildebrandt 9 Andreas Zieger Unconscious activation of motor responses in a hemiplegic, patient with anosognosia and neglect Received: 9 November 1994 / Accepted: 22 June 1995 Abstract Unconscious processing of environmental stimuli has been convincingly demonstrated for a large number of neuropsychological syndromes. However, only few studies have successfully recorded on-line the activation of the autonomous or the motoric nervous systems. The activation of the motor system by unconscious stimuli would be a strong argument that information processing in the brain is not organized into two serial steps (stimulus identification first, response selection second). This would argue for parallel processes of stimuli identification routines and response selection mechanisms on the basis of only provisionally worked out stimulus features: Investigating an anosognostic patient we found electrodermal activity (EDA) and electromyographic responses (EMG) to bimanual tasks, mental imagery, and to the request to execute specific actions. But overt behavior remained hemiplegic and the patient was unable to feel any kind of control of her paralyzed left arm and hand. Generally, the data can be interpreted along the lines of a theory of two different routes to action, leaving the unconscious selection of action patterns intact, whereas the intentional triggering of overt behavior with the left hand and arm was impossible for the patient. Key words Anosognosia 9 Continuous information transition 9 Neglect response selection Introduction Unconscious processing of environmental stimuli has been convincingly demonstrated for a large number of Helmut Hildebrandt ([~N) University of Oldenburg, Program in Neuropsychology, Post Box 2503, D-26111 Oldenburg, Germany email: hildebrandt at psychologie.uni-oldenburg.de Andreas Zieger North-West Hospital Sanderbusch, Department of Neurosurgery, HauptstraBe, D-26452 Sande, Germany neuropsychological syndromes [1]. Most of these studies relied on forcing the patients to choose between different alternatives and calculating the probability of right and wrong answers. However, there are only few studies that have attempted to ascertain unconscious reactions of patients directly [2, 3]. For neglect patients Vallar et al. [4, 5] documented the existence of visual evoked potentials and electrodermal reactions for unconscious stimuli, but we are not aware of any studies that have measured muscle responses in patients with neglect. A priori, there is no definite prediction of what kind of results such an experiment would uncover. On the one hand, it has been shown that neglect patients process unconscious information up to a categorical analysis of stimuli [6]. On the other hand, the overt behavior of such patients seems to be limited to reactions to stimuli located in the non-neglected hemi-space. Processing of stimuli therefore might gain no access to motor neurons and might be unable to release reactions [7]. Such findings would endorse a view of consciousness which argues that access to consciousness for perceptual stimuli depends on the action selection of the organism [8]. The alternative possibility would be that the innervations remain subliminal and do not result in overt behavior. This would parallel the situation of information uptake when the information is processed but does not reach consciousness. In this case the information would be processed through the whole perception action cycle and without direct reference to consciousness. Which of the triggered action patterns becomes dominant depends on multiple and interactive inhibitions and conscious selection. The model of Shallice [9] on supervisory attentional control where the contention scheduling works in such a manner that "winner takes all" (after the activation of different patterns of possible actions) would be in line with such an observation. Though the supervisory attentional control system is a central system for response selection, the contention scheduling is done automatically and in parallel. Therefore, the preactivation of action patterns by contention scheduling is independent of the conscious experience and decision to act. Given such a model, in 54 p r i n c i p l e it s h o u l d b e p o s s i b l e to d e m o n s t r a t e a disconnection o f p r o c e s s e s o f contention scheduling f r o m attentional s u p e r v i s o r y control after brain d a m a g e s . In order to a n a l y z e u n c o n s c i o u s p r o c e s s i n g in neglect, w e investigated a w o m a n w h o was h e m i p l e g i c and had a dense a n o s o g n o s i a for her left b o d y side e s p e c i a l l y for her left hand. H e r a n o s o g n o s i a c o m p r i s e d also a visual n e g l e c t and an anesthesia for her left b o d y half. M o r e o v e r , she did not a l w a y s a c k n o w l e d g e her left h a n d as her own. She t a l k e d about her left h a n d f r o m a t h i r d - p e r s o n p e r s p e c t i v e and she d e v e l o p e d several e x p l a n a t i o n s for w h y this foreign h a n d was attached to h e r b o d y and w h y it was i m p o s s i b l e to get rid o f it [10]. O u r h y p o t h e s i s was that unc o n s c i o u s and s u b l i m i n a l activation o f the left h a n d was possible, but out o f c o n s c i o u s control and w i l l e d intention. This h y p o t h e s i s was s u p p o r t e d b y the p a t i e n t ' s reports about r e f l e x - l i k e actions o f the left hand, w h i c h w e h a v e not y e t h a d the o p p o r t u n i t y to observe. A l l we can say in this r e g a r d is that during our n e u r o l o g i c a l , neurop s y c h o l o g i c a l , and b e h a v i o r a l investigations her left h a n d Fig. 1 Patient HMU: computertomographic angiographic findings. First row, left picture: angiography of the right-sided carotis artery revealed a big saccular aneurysma of the right media artery; picture in the middle: lateral schematic overview on the extension of the right intracerebral mass bleeding; right picture: lateral view of the computertomographic scan slice levels of the total investigation. The slice thickness was 6.0 cm. Second row a n d third row: axial computertomographic scan slices from the basal level (1A) up to the high parietal level (3B): the mass bleeding extended from the basal parts of the right sylvian flssura up to the (fronto-)temporo-parietal areas of the right cerebral hemisphere, disconnecting intraand intercortical association fibers. The computer-assisted planimetfic volume evaluation revealed a bleeding size from about 424 ecm. Note the compression of the fight side-ventricle, the temporal part of the right side-ventricle, and the small shift of mid-line structures such as the cerebral pedicle from right to left. Shift and compression were caused by perifocal brain edema, and by mass effect of the bleeding and arm always remained hemiplegic. To test our hypothesis w e d e c i d e d to m e a s u r e the e l e c t r o d e r m a l r e s p o n s e by skin c o n d u c t a n c e r e s p o n s e (SCR) and the m u s c l e tonus pattern b y surface e l e c t r o m y o g r a p h y ( E M G ) . Case report The patient, 59 years old, was admitted to the neurosurgical department because of an intracerebral huge hemorrhage at the right fissura sylvii (see Fig. 1). The mass bleeding was caused by an aneurysmal malformation of the right cerebral media artery, detected earlier by cerebral angiography. After neurosurgical evacuation of the hemon-hage and aneurysmal clipping, the patient remained in a deep coma for approximately 10 days. A follow-up CT, carried out 57 days after the operation, showed a low density area in the right cerebral hemisphere extending from the fronto temporal to the parietal lobe. The neurological investigation documented left-sided hemiplegia, abnormal reflexes of the left leg, and a dense anesthesia of the left body side. Sometimes the patient also reported alloaesthesia for tactile stimulation. In these situations the touching of the left side of the body was localized at the right side. Stimuli which caused serious pain sensations (i.e., pinching) were 55 commonly perceived correctly in the left side of the body, but not exactly at the place to which they were applied. A neuropsychological examination with the WAIS (Wechsler Adult Intelligence Scale), carried out 5 or 6 weeks after the evacuation of the hematoma, showed a normal verbal IQ, but a percentage of zero in the performance tasks (see Table 1). The Benton test (for identifying visually perceived patterns) yielded only random results, because the patient always had chosen the items given at the right side of the answer sheet. Drawings and writing probes showed that she had no constructional apraxia because line orientation and composition of the figures were not disturbed. On the other hand, the drawings mad writing probes documented the neglect of the left part of the egocentric space (see Fig. 2). Neuropsychological assessment of neglect Because of the clinical status of the patient, the testing was restricted to bedside diagnostics. Our behavioral observations included the typical features of patients with unremediated neglect. The patient's head was always oriented toward the right side, and generally she did not react to voices from the left side of her bed. When talking to people whom she knew were standing on the left T a b l e 1 Patient's results in the WAIS side of the bed, she raised her voice and pronounced her words carefully as if they were standing far away. Line bisection We asked the patient to bisect lines of different length and repeated each length four times. In half of the cases the patient bisected the line towards her intact hemifield, in 33% toward the neglected hemifield, and in 17% the line bisection was within the normal range. The deviation toward the intact field was more pronounced (mean 1.25 cm) than the deviation toward the neglected hemifield (mean 0.3 cm). Clock drawing We asked the patient to write down the dials of a clock in their proper positions along a prescribed circle (diameter 13.6 cm). Then we asked her to draw the positions of the sweephands of the clock for different times. The arrangement of the hours and the notation of the different times documented a profound neglect of the left part of the circle (see Fig. 2). Imagining the positions of letters in words Subtest Scaled score Information Comprehension Digit span Arithmetic Similarities Vocabulary Digit symbol Picture arrangement Picture completion Block design Object assembly 7 10 8 9 10 11 0 2 0 2 0 4L Fig. 2 Writings of the hours of a clock around a circle by the patient. After noticing that something was incorrect, the patient wrote the numbers at the outer circuit (1, 15, 14). The first sweep hand pointing to n u m b e r "12" and the one pointing to number "6" were drawn as three o'clock. The first sweep hand and the one pointing to "9" were drawn as nine o'clock. The first sweep hand and the one between "12" and "1" as eleven o'clock, the one pointing to ~ as four o'clock W e used the imagination of words to analyze neglect in mental imagery. We used words rather than visual scenes to control for the extention of the visual scene and because the writing probes had shown that the patient normally made no spelling errors. The task was to indicate the letters at different positions in words that varied in length ("Which is the fourth letter in 'sleep'?"). Extending the length of the words to more than four letters and the position up to the third letter showed that the patient had enormous difficulties with such kinds of tasks: She was right only in 50% of the cases. The performance declined even more if the chosen positions were greater than three. Generally, the patient answered to such questions by naming the last letter. Assessment of directional hyperkinesia The investigation of the mental imagery of words documented a typical behavioral pattern: Movements toward the intact hemifield seemed to stop only if a prominent anchor was reached (the last letter of the imagined word). This can be interpreted as a motor hyperkinesia [11] toward the intact field. To analyze the role of movements for the neglect we assigned the patient two different tasks: The first task was to copy the length of an arrow on a separate sheet of paper. The position of the arrows with regard to the patient was always directly in front of her. The starting points of the arrows were marked on the separate sheets and the patient had to start drawing at these points toward the tip and to stop when she reached a distance similar to the length of the arrow. The second task was to reproduce the distance between two points, marked on a 61-cm wide strip of paper (7 cm high). Reproductions had to be done either toward the right or towards the left. The two points were marked either in the middle of the strip or at the left or right end of the strip. (In this case the reproductions had to be done only toward the opposite end of the strips.) We expected that a directional hyperkinesia should be accompanied by a pronounced overestimation toward the intact hemifield and an underestimation toward the neglected hemifield. The results supported our expectations. Firstly, it was possible that the patient shifted her attention (by cuing) into the neglected field and therefore could perform the reproductions. Secondly, the reproduction did not differ in the position of the parts in the hemifields, but in the direction in which the reproductions had to be done. Reproductions toward the neglected hemifield were systematically underrated, whereas reproductions toward the intact hemifield were overestimated. Moreover, in several of the cases where the 56 direction in question was toward the neglected hemifield, the patient solved the tasks by acting toward the intact hemifield. On the other hand, in none of the trials toward the intact field did the patient err in the correct direction. The behavior of the patient in these tasks resembled the performance of patients with allochiria. Halligan et al. [ 12] reported a patient who copied parts of a figure located in the neglected field to her reproduction of the figure, but into the opposite half, i.e., the intact hemifield. Moreover, the localization of the hemorrhage of Halligan's patient and that of ours was quite similar (frontoparietal, right hemisphere) and so were the results of the neuropsychological investigations. Halligan et al. argued that motor hyperkinesia could not explain their results because of the global drawing performance and because they observed allochiria only under conditions of dual stimulation by the original picture. On the other hand, motor hyperkinesia was not tested directly. The reported findings contradict the interpretation of Halligan et al. The existence of allochiria was coupled to the specific direction in which the patient had to move her arm (we observed allochiria in 55% of all trials toward the neglected hemifield against 0% of the trials toward the opposite direction). Perceptual features can be ruled out as an explanation, because we did not use double stimulation and all stimuli were equivalent, except for their location in the patient's egocentric space. EDA and EMG measurement Procedure We used electrodermal activity and muscle tonus as indicators for unconscious activation of the left arm and left hand. Electrodermal activity was registered by a skin resistance measuring device (Bruno ZAK VDI), and registration of muscle tonus was conducted by an EMG-feedback device of the same manufacturer. The EDA was registered at the right thenm"and the hypothenar eminences of the left and right hand. Artifact-free amplitudes of skin conductance responses (SCR) of at least 0.02 ps (microsiemens) with a latency of 1-3 s were scored as stimulus-evoked changes. In order to find a baseline for both hands, the measurement started with a general relaxation period. Thereafter, we proceeded with different mental imagery conditions ranging from "having a warm bath" to "grasping a bottle" or "potato peeling". We requested the patient also to grasp a bottle we were holding next to her left hand. The EMG was registered at the forearm (above the musculus flexor carpi radialis). The analogous output of the measuring device was normalized to a range between -1 and + 1 voltage effective and registered each 13 ms by an Apple lie (Apple Computer, Cupertino). We used a single case experimental design for measuring the EMG. We started with a baseline ("A') and introduced different interventions ("B", "C", ...) always followed by a baseline period. Each intervention was repeated several times. The following interventions were carried out repeatedly: unspecific mental imagery of a movement sequence embracing the left hand, imagery of movements that had been executed at an earlier time in her life (clapping her hands after an exciting theater performance, cleaning the windows of her home) and requests to grasp a hand or a bottle we were holding next to her left hand. We also put a bottle near her left hand and then asked the patient to unscrew the stopper with her right hand. Results For the E D A m e a s u r e m e n t we f o u n d a considerable difference of S C L for the right and the left hand (6.7 ~s for the right vs. 2.0 ~s for the left hand). We then noticed different responses for the mental imagery of either having a w a r m bath or of potato peeling (see Table 2). Tactile stimulation of the left arm yielded no significant response, and the same holds for the request to grasp a bottle. O n the other hand, painful stimulation (pinching of the left arm) was followed by a significant electrodermal response. The m e a s u r e m e n t of E M G included the baseline periods and several experimental situations. We tested the different experimental situations against all baselines we had m e a s u r e d (we m e a s u r e d 17 baselines followed b y the experimental situations, cf. Fig. 3 and Table 3). As we had registered deflections of effective voltage, our E M G data varied around a m e a n 0 voltage. For calculating significant alterations of the experimental conditions we conducted two different tests. Firstly, we used the F-statistic to compare directly the variances of the E M G values during baseline and experimental conditions. Most of these separate tests (17 baseline against 16 experimental conditions) were significant, partly due to the high degrees of freedom. O n l y the condition of unspecific m e n t a l imagery of m o v e m e n t s d o c u m e n t e d a considerable portion of insignificant results (41% for riding a bicycle and 24% for clapping o n e ' s hands; see Table 3). For a second and more conservative test we averaged the variance of the E M G values over the 17 baseline conditions and c o m p u t e d the m e a n variance of the resting potential and its standard deviation. A m e a n of variance during the intervention periods outside the range of the m e a n variance of the resting potential plus one standard deviation was taken as a significant modification. All statistical evaluations were done by using the software package S T A T G R A P H I C S (Statistical Graphics Cooperation). Testing for significant experimental conditions led partly to significant increases in muscle i n n e r v a t i o n d e p e n d i n g on the kind of interventions (see Table 3). Requesting the patient to form unspecific mental images of m o v e m e n t s showed the lowest increase in E M G , followed by the request to grasp with the left h a n d for a bottle or for a h a n d without touching them. These two Table 2 EDA responses for the patient Condition Relaxation Right hand Baseline of 6.7 las Left hand Baseline of 2.0 ps 0.02 <=<0.04 gs ~ 0.02 <=< 0.04 ~s aAmplitude of skin conductance response (SCR) bSCR was measured at the same body side to which we applied tactile stimulation. Application was done to upper Mental imagery of having a warm bath Mental imagery of grasping a bottle Tactile stimulationb Pinchingb Mental imagery of potato peeling Request to grasp a bottle 0.02 ~s 0.02-0.04 ps > 0.04 ]as Not done Not done No response No response 0.02 <=< 0.04 ps 0.02 <=< 0.04 gs No response arm 57 EMG Measurement at the musculus flexor carpi radialis EMG 20.0 - Intervention: memory of clapping hands 11.2 A B A B A C A ... Interventions Fig. 3 Single case experimental design for electromyographic measurement ' i .9 -6.4 Table 3 Interventions and deflections of the EMG tested against standard deviation of mean variance of 17 baselines Interventions Unspecific mental imagery Riding a bicyle Clapping one's hands Specific mental imagery Clapping one's hands at a theater visit Cleaning the windows Grasping trials Shaking one's left hand Request to take a bottle Putting the bottle into the plegic hand Bimanual tasks Grasping the bottle with both hands Unscrewing the bottle Closing the bottle -15.2 Variance Right hand left hand ls