BRAIN AND COGNITION 25, 1-23 (1994) Apperceptive Visual Agnosia: A Case Study PauL A. SHELTON, Dawn Bowers, RANJAN Duara,* AND KENNETH M. HEILMAN Neurology Service, Department of Veterans Affairs Medical Center, and the Center for Neuropsychological Studies, Department of Neurology, University of Florida College of Medicine, Gainesville, Florida 32610-0236; and *Mt. Sinai Hospital and the Department of Neurology, University of Miami College of Medicine, Miami, Florida A man with an infarction of his inferior temporal and occipital association cortex bilaterally, which spared primary visual cortex, had impaired visual recog- nition of objects, faces, colors, words, and gestures. Analysis of visual function indicated that the recognition failures resulted from an agnosia, rather than ele- mental visual impairment. Whereas his impairment of gesture recognition ap- peared to be related to an associative agnosia, his inability to recognize objects was related to an apperceptive agnosia. There may be four subtypes of appercep- tive agnosia: one where the internal object representations or structural descrip- tions are impaired, another where an adequate percept cannot be derived, a third where the internal referent and percept are dissociated, and a fourth where both levels are impaired. Our patient demonstrated a failure to relate individual ele- ments to the whole, a failure to integrate multiple elements, and a reliance on global perception. He had normal object imagery. These results suggest that, whereas internal representations were intact, he was unable to form adequate perceptual representations. © 1994 Academic Press. Inc. INTRODUCTION Visual agnosia is the inability to recognize visual stimuli despite ade- quate elementary visual function, including visual acuity, visual fields, visual scanning, and attention, as well as preserved general mental abil- ity. Individuals with visual agnosia fail to identify the stimulus overtly by naming, demonstration of use, or selection of the named item from an array. They may lack subjective familiarity with the stimulus, and automatic or emotional responses to the stimulus may be inadequate. This work was supported in part by the Medical Research Service of the Department of Veterans Affairs and the State of Florida, Department of Health and Rehabilitation Ser- vices, Division of Aging and Adult Services. Address correspondence and reprint requests to Kenneth M. Heilman, M.D., P.O. Box 100236, JHMHC, University of Florida College of Medicine, Gainesville. FL 32610-0236. 1 0278-2626/94 $6.00 Copyright © 1994 by Academic Press, Inc. All rights of reproduction in any form reserved. 2 SHELTON ET AL. Visual agnosia may be restricted to a specific category of stimuli such as faces, colors, objects, words, or subcategories. Recognition of class membership or, alternatively, of individual identity may be impaired pref- erentially. According to Lissauer’s (1889) two-stage model of recognition, visual agnosias have been classified as (1) apperceptive, where the indi- vidual never achieves a complete conscious visual impression or percept of the stimulus, and (2) associative, where such an adequate visual im- pression or percept is divested of its meaning, that is, its semantic associ- ations (Lissauer, 1889). In associative visual agnosia, detailed visual percepts are formed, pro- viding sufficient information to support the process of identification. These individuals can describe real objects in detail, can copy and draw items normally, and can match items to samples (Rubens & Benson, 1971; Taylor & Warrington, 1971). Recognition performance is worse with photographs and line drawings, however. Tactile and auditory recog- nition may be preserved, but short-term verbal memory is typically im- paired, and modality-specific memory deficits restricted to visual memory may be present (Ross, 1980). In contrast, apperceptive visual agnosia is often associated with deficits in elementary visual processing. Some have argued that the behavioral manifestations of visual agnosia result from impairment of elementary visual function alone (Bay, 1953; Tyler, 1967). Although many such individuals are initially cortically blind or are consid- ered to be blind, sufficient visual field and visual acuity can be docu- mented on formal testing. Substantially preserved stimulus threshold, spatial resolution, adaptation time, sensation time, and visual field have been documented in several patients (Benson & Greenberg, 1969; Ett- linger, 1956; Levine & Calvanio, 1978). They fail to identify items sponta- neously and cannot accurately draw or match items to sample, although in some patients the ability to trace outlines accurately has been demon- strated (Adler, 1944; Landis, Graves, Benson, & Hebben, 1982). The psychological mechanisms underlying the visual agnosias have not been fully elucidated. Current models of the processes underlying visual object recognition propose that visual information is processed in four stages (Davidoff & Wilson, 1985; Ratcliff & Newcombe, 1982; War- rington & Taylor, 1978). According to these four-stage models, recogni- tion failure could result because (1) visual perceptual information may be inadequate, (2) a sufficiently detailed percept may be disconnected from internal structural representations or stored structural descriptions, or (3) the internal representations themselves may be degraded, and finally (4) an intact and activated representation may be dissociated from seman- tic knowledge. The first three stages correspond to the process of percep- tual categorization of Warrington and Taylor (1978) and a failure at any of these first three stages may appear as an apperceptive agnosia. Although the localization of causal pathology affiliated with associative APPERCEPTIVE VISUAL AGNOSIA 3 agnosia is well known, there is little information regarding the localization of lesions in apperceptive agnosia. Most reported cases have had diffuse cerebral injury or bilateral posterior cerebral lesions with the implication of involvement of the primary visual cortex or the geniculocalcarine path- ways bilaterally (Adler, 1944; Benson & Greenberg, 1969; Kertesz, 1979; Levine, 1978; Tyler, 1967). Recently we had the opportunity to examine and test a patient who conforms to the classical description of apperceptive visual agnosia. The lesion responsible for the disorder was restricted, as demonstrated by CT and PET scan. To characterize the nature of his perceptual deficit and the role of associative processes in apperceptive visual agnosia, a variety of tests of elementary visual function and of more complex perceptual processes were performed. Case F.W.T., a right-handed white man, was 66 years old when he experi- enced the sudden onset of headache, visual loss, and right-sided weak- ness. The following morning he was found to be totally blind, with mild left-sided weakness and decreased arousal initially. These improved grad- ually over a period of days. Although initially he stated that he was unable to see, he could feed himself. He accurately reached out to shake hands when greeting visitors and when reminded of these inconsistencies he stated that he could indeed see but not clearly. He stated that everything seemed to ‘trun together’’ and explained that he had lost his glasses. He failed to recognize familiar visitors including his wife until they spoke and when asked to describe the view outside his window he would describe in detail a completely different scene. For example, instead of a hospital rooftop he described a playground where children played on swings and slides on grass under the trees. According to his wife he was describing their backyard. When seated in a conference room filled with people sitting around a table and asked to describe what he saw, he instead described a cityscape of tall buildings. He was thought to have had a stroke secondary to intermittent atrial fibrillation with sinus pauses of 2-3 sec and was treated with digoxin, quinidine, coumadin, and a pacemaker. Echocardiogram subsequently revealed a calcified mitral annulus but normal left ventricular function and no mural thrombus. In the past he had arrhythmias treated with verapamil, but he had no other risk factors for stroke. He had an eighth- grade education and worked as a housepainter and in a maintenance department after his military service. He was alert and cooperative but distractible, with motor impersistence for eye closure and arm extension. There was a paucity of spontaneous speech and movement. His facial expression was blank unless he was 4 SHELTON ET AL. spoken to directly, and in conversation he displayed reduced range of affect and emotional expression. His gaze was generally fixed in the dis- tance with little visual tracking of movement in the room and he appeared unaware of objects placed before him until his attention was directed toward them specifically. He was disoriented for time and place and he had difficulty recalling the names of his children and occasionally his second wife. He demonstrated severe anterograde amnesia for auditory verbal material, for example recalling one of three concrete nouns at 2 min. Recall of remote past events of personal relevance was better but he recalled no recent public figures or important dates. Digit span was 5 forwards but 0 backwards. He acknowledged that his memory was poor but was unconcerned. His speech was normal, his language was fluent with good repetition, and his comprehension of conversational speech was normal. He cor- rectly named 30 objects, animals, and body parts when given semantic cues or when body parts were touched by the examiner. Verbal fluency was one item in 60 sec. He was able to write only his name. Praxis, simple calculations, right-left orientation, and finger recognition were normal. Insight, general knowledge, vocabulary, abstract reasoning, and problem-solving skills were compatible with his occupational and educa- tional level. He was unable to draw or copy simple figures and, when asked to complete a clock face, clustered the numbers in the right lower quadrant. On a line cancellation test the upper half and both extreme right and left peripheral stimuli were neglected. Similarly, he demon- strated visual extinction for upper and peripheral left spatial stimuli. He was unable to find his way on the ward, unable to describe the route home or to familiar locations, and unable to place cities or states on a map. He had a mild proximal weakness of both upper extremities. His deep tendon reflexes were mildly increased in the left upper and lower extremi- ties, and his plantar responses were extensor bilaterally. His snout reflex was positive, but his grasp and glabellar reflexes were negative. He had mild dysmetria in the left arm only, his stance was unsteady, but his gait was narrow-based. Memory. To assess his memory more fully, we performed additional auditory verbal and nonverbal as well as tactile memory tests. His audi- tory verbal memory was tested with a word list of 10 items and 10 paired associates. His auditory nonverbal memory was tested with 20 familiar sounds, and his visual recognition memory was tested with a set of 20 recurring nonsense figures. We tested his tactile object recognition by having the patient select a previously felt item from a sample of four other objects and in all modalities memory was tested after a distraction-filled interval of 5 min (Table 1). He displayed severe anterograde memory APPERCEPTIVE VISUAL AGNOSIA 5 TABLE | Recognition Memory Performance Recognition memory (%) at chance (%) Tactile 2/4 (50) (25) Auditory nonverbal 10/20 (50) (50) Visual 12/24 (50) (50) Verbal 0/20 (0) (0) impairment regardless of stimulus modality, performing at chance for all but tactile stimulation. Visual Function General observations. The patient was able to see and point to objects presented visually but was unable to identify them by name or use. He was better at selecting named items from an array. He never felt confident in his decision or had a sense of familiarity. He recognized keys, a comb, and a clock by their characteristic sounds and easily identified family and familiar hospital staff by their voices despite his complete failure to recognize their faces. He recognized the objects when presented tactually and named all objects whose use was described in spoken words, indicat- ing that his failure to identify objects was not due to anomia or general memory deficit and that the recognition failure was restricted to the visual modality. His recognition of photographs of objects was worse than that of real objects, and he was unable to recognize or copy line drawings or match objects or drawings to samples. When presented with a complex picture, he correctly localized all the elements on the page by pointing but was unable to identify them or describe their relationship. His pupillary responses were normal, and his voluntary extraocular movements were full with apparently normal saccadic latency, but his movements were initiated by head-turning. The subject’s pursuit move- ments were full but broken into small saccades in all directions. Visual acuity. Standard eye charts could not be used to document visual acuity due to the patient’s recognition failure. However, he was able to indicate that all the 14-point numbers on the Snellen chart were different, suggesting visual acuity at the 20/100 level. He could count fingers and consistently indicated the orientation of vertical or horizontal lines of 1-mm width and 10-mm length at 35 cm (8-min arc) and localized 2-mm dots on a white background at 50 cm (12-min arc), Finally, optoki- netic nystagmus was consistently elicited with a black and white grading with a period of 2.5 mm at 35 cm (18-min arc). Visual fields. F.W.T. had bilateral superior quadrantanopias for mov- 6 SHELTON ET AL. Fic. 1. The patient's visual fields on confrontation testing using a moving 4-mm white target. Hatched marks denote areas where patient failed to detect stimulus. ing and stationary solid targets, but consistently identified the onset and offset of a light flash in all visual fields, including the upper hemifield to 60° above the horizontal meridian. Formal perimetry could not be accomplished but the kinetic perimeter for a 4-mm white target on bed- side testing is shown in Fig. 1. Visual localization, He accurately localized all light stimuli in the lower visual fields but made a systematic error toward the lower right for targets in the upper visual fields. He erred by 30-40 cm for targets in the upper left and by 30 cm for midline and above eye level (see Shelton er al., 1990 for a complete discussion of the deficits in visual attention and visual mislocatization). Size, distance, and line orientation. He consistently discriminated be- tween black dots differing by 2-mm diameter. He was unable to discrimi- nate relative orientations of line pairs of less than 90° initially, but at a follow-up evaluation scored 75% correct on a modified line orientation test requiring a same—different judgment for separations of 18-90°. He correctly judged which of two objects was more distant for differences as small as 0.5 cm. He performed normally when estimating the distance between 10 pairs of lines. Finally, on a test of distance judgment including an immediate spatial memory component, he made a mean error of 3.7 cm over 20 trials (control group X = 1.2 cm, SD = 0.8). Color. {n bedside testing the patient failed to name, point to, or match items of a specific color. When presented with cards in hues of varying saturation, he correctly named two of eight highly saturated colors and selected three named colors from an array of eight, but none with lower saturation. Form. When presented with drawings of geometric figures, letters, and numbers with a size (5 x 3 cm) exceeding the limits of his visual acuity and of high contrast, he identified no geometric forms, 8% of numbers, APPERCEPTIVE VISUAL AGNOSIA 7 TABLE 2 Visual Form Recognition Recognition (%) Copying Matching Writing on Naming Pointing (%) to sample (%) | command (%) Letters (1) 4/17 (24) 20/30 (67) _ _ 12/17 (71) Numbers (1) — 2/34 (8) 16/18 (89) _ _ 15/15 (100) Figures am 0/9 (0) 0/9 (0) 0/9 (0) _ V/9 (EL) (2) 1/9 (11) _ 3/9 (33) 10/12 (83) 2/9 (22) Note. Performance at the initial (1) and 9-month (2) evaluation. and 24% of letters (Table 2). He accurately traced 50% of stimuli, and in the majority of the remaining figures his tracing was deficient only in the neglected field such that his tracing became complete when the stimulus was inverted. His approach to tracing differed from normal in a consistent manner. Rather than using the fewest lines possible, consistent with the Gestalt principle of good continuation, his segments tended to radiate from or toward junctions in a piecemeal fashion (Fig. 2). When stimulus contrast was reduced, the connectivity of his tracings became further reduced, and he began to confuse the margins of the stimulus card with the figures themselves. He was able to accurately trace 75% of letters that he could identify visually by name but only 38% of those he failed to name. He performed significantly better in selecting a named stimulus from an array of 5, and he failed to recognize any Setter that he could not write from memory (Fig. 3). He was unable to identify by name any of the integers but para- doxically correctly read 16 and 46 and read 376 as ‘*673.”" Movement. Kinetic perimetry was described above. Recognition of symbolic movements was tested in a standardized fashion, as used in testing for ideomotor apraxia. The patient was first asked to make a gesture or to pantomime the use of an object upon verbal command. The examiner then demonstrated the correct movement in random order with two movement foils, one being a correct movement for the wrong gesture or object and the other foil being a novel or nonsense movement. He was then required to select by pointing or by number the correct movement. Finally, he was asked to imitate the correct movement for each gesture or tool demonstrated by the examiner. He recognized only 8% of panto- mimes and gestures despite normal pantomime and gesture production upon command (Table 3). In no case did his correct imitation of a gesture facilitate recognition of the movement. Object recognition. He was allowed to see 15 common objects in a variety of different views, identifying them by name, by describing or demonstrating use, by matching to an array of five items, and finally by 8 SHELTON ET AL. + oe R K 4 Z Fic. 2. Demonstration of the movements made by the patient when tracing symbols and figures. Numbers indicate the sequence of line segments traced by the patient. pointing to a named item in an array (Table 4). In only one instance did he indicate a sense of familiarity, and he was unable to copy line draw- ings. He failed to recognize or describe photographs of famous faces and was unable to match them to sample. Auditory and tactile object recognition. Tactile recognition was as- sessed by asking the patient to name 20 common objects felt with both hands. Auditory nonverbal recognition was formally tested with taped APPERCEPTIVE VISUAL AGNOSIA 9