There are several illuminating examples of specific deficits in the perception of some aspect of motion perception following small cortical lesions in individual patients (see Introduction). Such single case studies remain particularly important in neuropsychology but they are bound to be rare. There is always the possibility that an equally small lesion elsewhere in visual cortex might have a similar effect, making it difficult to attribute an impairment solely to a particular functionally or anatomically defined visual area. The current study was designed to clarify this problem by studying a large number of patients with damage restricted to one of four different regions that are known to be involved in some aspect of motion processing. Such an investigation complements single case studies.
The present functional imaging results strongly support the notion that occipito-temporal cortex (lesion Group 1) and frontal–prefrontal cortex (lesion Group 4) are not essential for most or any of the discriminations involving the different types of motion tasks used here. Because the neuronal substrate of the psychophysical tasks embodied in Expts 1–5 suggests early visual processing, we first compared the performance of the four groups of patients among themselves and against the control subjects for stimuli presented in the contralateral visual field. The current view of the functional architecture of the human visual motion system suggests that there should be differences between the four patients groups, which were defined on anatomical criteria. The task in Expt 6, however, is higher-level and assumed to be mediated by neurons whose receptive fields are very large and encompass a large portion, if not the whole, visual field.
The fMRI results and psychophysical data from the four patients’ groups support the proposal that both occipito-parietal (Group 2) and rostro-dorsal parietal areas (Group 3) are important for efficient global motion perception, although not in an identical manner. For instance, although patients with occipito-parietal lesions (Group 2) were impaired on all tasks except 2D-FFM, patients with rostro-dorsal parietal lesions (Group 3) were not impaired on direction and SDT. The latter were the two tasks that produced least and only unilateral activation in the dorsal parietal cortex of the normal subjects.
One of the most influential notions about the gross organization of the cortical visual system is that it is divisible into dorsal (chiefly parietal) and ventral (chiefly occipito-temporal) functional systems (Goodale & Milner, 1992
; Ungerleider & Mishkin, 1982
). The results of the present study point unequivocally to the involvement of the dorsal pathway in the motion discrimination tasks. However, patients with occipito-parietal lesions (Group 2) was more impaired than the patients with rostro-dorsal parietal lesions (Group 3) indicating that the crucial damage is probably to the areas in the intraparietal sulcus and the inferior parietal lobule and corresponding to areas VIP, AIP, LIP, and MIP, as defined anatomically and physiologically in macaque monkeys and functionally in human neuroimaging studies. It also suggests that motion areas such as PO, which lie more medially in the parietal lobe, are less important with respect to these tasks. The finding that the group with occipito-temporal lesions was not impaired on any of the tasks is entirely consistent with the notion the ventral pathway is much more concerned with the perception of colour and form than with motion, although motion can be used to create form. Nevertheless, there were functional activations in this region (see Figures and ) correlated with the tasks of motion discrimination thresholds and 2D-FFM.
The absence of any impairment following prefrontal lesions is also interesting because it indicates that although the frontal eye-fields and the supplementary eye-fields contain abundant motion selective neurons in macaque monkeys, and are regions often functionally activated in human subjects by moving displays, their activation is not required for visual motion discrimination per se. Nor were they functionally activated in the present study with the possible exception of frontal area six in the task involving the discrimination of 2D-FFM. Visual motion is evidently processed in a variety of cortical brain regions for different purposes: as a means of perceiving motion itself, in order to create form from motion, to segment a complex moving scene, and to provide the information for appropriate motor responses like eye movements and reaching and grasping. Brain lesions can inform us about where these occur in a manner still difficult with single cell recording in monkeys or functional neuroimaging in humans.