Results

For the line bisection task, neglect was defined as a mean error to the right or to the left of the true midpoint that exceeded the 95% confidence intervals (95% CIs) for the bisection errors of control subjects.7 Neglect on the cancellation task was defined by the failure to cross out one or more targets. Based on the definitions, we found 30 subjects with neglect. Ipsilesional neglect on the line bisection task was defined by a mean magnitude of error to the side opposite the brain lesion that exceeded the 95% CI of control subjects.7 Based on this definition, five of 30 patients (16%) had ipsilesional neglect. For the cancellation task, we defined ipsilesional neglect as a failure to cross out more targets on the side of the paper ipsilateral to the lesion than on the side contralateral to the lesion. No patient met this criterion on the cancellation task. Therefore, ipsilateral neglect was seen only in line bisection tasks. The ipsilesional neglect group comprised one man and four women (table1). In one case (patient 5), ipsilesional neglect clearly evolved during recovery from contralateral neglect. Patient 5 had a mean bisection error to the same side as the brain lesion that exceeded 95% CIs for control subject performance (table2).

All five patients with ipsilesional neglect had vascular lesions involving the frontal-subcortical circuits. Lesion distribution included three restricted to the basal ganglia, one restricted to the thalamus, and one with a large lesion that involved cortical and subcortical structures. No patients with ipsilesional neglect were seen with lesions limited to the posterior association areas.

By contrast with line bisection performance, patients who demonstrated ipsilesional neglect eventually located all the targets in the cancellation task. Left sided omissions were present at those times when patients showed either minimal leftward bisection bias (patient 1 and 2) or had not yet developed ipsilesional neglect and had rightward bisection bias (patient 5). However, unlike normal subjects, every patient’s search began on the right and proceeded from right to left, even after they were able to mark every target. Because the five patients with IN performed normally on the cancellation task in the direct condition, testing in the indirect condition would have little meaning and was therefore not undertaken.

On the line bisection task, two patients reversed the direction of bisection bias between the direct and indirect conditions, a performance consistent with a predominant influence of a left sided perceptual-attentional bias. The remaining three patients bisected to the left of true midpoint in both direct and indirect conditions, suggesting a predominant influence of a motor-intentional, left sided bias.

Discussion

The occurrence of ipsilesional neglect is not as rare as suggested by the scarcity of reports in the literature. There are several possible reasons for the limited number of reports. Firstly, ipsilesional neglect may be task dependent. In this study, ipsilesional neglect was seen only in line bisection tasks. We did not find it on the target cancellation task but this may have been related to a ceiling effect as others have found ipsilesional neglect on various neglect tests such as target cancellation, tactile exploration, figure copying, and representational drawing.4 Secondly, ipsilesional neglect may be a stage in the recovery process that develops in the postacute phase after the initial brain injury. Initially, there may be diaschisis of the uninjured portion of the right hemisphere, and ipsilesional neglect only becomes manifest when these areas recover. Three patients with ipsilesional neglect tended to acutely show contralesional neglect on cancellation tasks and minimal line bisection error, a pattern of performance associated with frontal brain lesions.8 However, over time, the cancellation bias resolved and line bisection error shifted to the left of midpoint. Similar observations were made in previous descriptions of ipsilesional neglect.2 8

Robertson et al 4 suggested that ipsilesional neglect is induced by a compensatory left sided scanning strategy. There are two findings that would not fully support the compensation hypothesis. Firstly, on the line bisection task, patients 3 and 4 never demonstrated contralateral neglect before they demonstrated ipsilesional neglect, and patient 4 was tested 3 days after his stroke. Hence, these patients had neither the impetus nor the time to acquire compensatory bias to the left. Secondly, if a left sided compensatory scanning strategy were responsible for ipsilesional neglect, we would have expected all our patients to show a predominant influence of motor-intentional factors in neglect behaviour. However, several subjects had predominantly perceptual-attentional neglect.

All of the patients with ipsilesional neglect had lesions involving frontal-subcortical circuits (frontal lobe, basal ganglia, and thalamus). Frontal lobe lesions may be associated with neglect.10 The frontal lobes are part of the systems that mediate both attention and intention. That attentional neglect can be in one direction and intentional neglect in the other direction provides evidence that, although the frontal lobes contribute to both systems, these systems, at least in part, are functionally independent.11 The prevailing bias may depend on task demands. Therefore, after a frontal lesion, a subject’s performance may reflect a combination of ipsilesional neglect on an intentional basis and contralesional neglect on an attentional basis. As the spatial attentional requirements are reduced, they will show ipsilesional neglect, and as the attentional requirements are increased, they will show contralesional neglect. In addition, the attentional and intentional components may have different recovery patterns, and ipsilesional intentional neglect may be one stage during the course of recovery from contralesional neglect. Our findings in patients 1, 2, and 5 support this postulate.

Cunelli et al 12 documented dissociation of neglect behaviour using a keyboard exploration task under visual guidance and while blindfolded. Unexpectedly, one patient showed normal performance under visual guidance but displayed a left sided bias while blindfolded. According to the authors, the anomalous performance “cannot be explained by any current theories of neglect.” The same report mentions a patient with left hemispheric damage with ipsilesional neglect for words but contralesional neglect on other measures of neglect such as target cancellation and figure search. According to our formulation, the first instance of dissociated behaviour would result if blindfolding somehow reduced the spatial attentional demands of the task, thereby “unmasking” an intentional bias. The second might reflect relatively increased attentional demands invoked by searching across multiple objects in an array versus the relatively reduced attentional demands of processing single words.

Villardita described another dissociation using tactile exploration of a maze after right or left hemispheric stroke.13 Patients without neglect after either right or left hemispheric injuries performed the task similarly, locating a greater proportion of items from space ipsilateral to the lesion. By contrast, right hemispheric damage with contralateral left visuospatial neglect, as determined through figure copying and cancellation tests, showed the opposite behaviour on this tactile task, locating a greater proportion of items from contralesional space. Villardita thought that in the tactile modality after elimination of vision, subjects were biased toward contralesional space in an attempt to compensate for defective visuospatial processing. Alternatively, perhaps blindfolding reduced the ipsilateral visuospatial attentional bias and the contralesional intentional bias became manifest, thereby inducing ipsilateral neglect.

Riddoch et al 14 provided another account for task dependent dissociations of spatial bias. After left hemispheric injury, their patient showed right sided neglect in figure copying and cancellation tasks but showed left sided neglect for reading, a dissociation reported by others.15 However, the patient also neglected rightward characters when spelling. After further experimental assessment, the authors concluded that the direction of the patient’s neglect depended on whether stimuli were represented as parts of one object or as independent parts of perceptual objects. Their account could apply to our findings if line stimuli are processed primarily as single perceptual entities, whereas target arrays or figures are processed as multiple separate objects. The hypothesis might be tested by devising “bisection” tasks using lines composed of individual elements.

The reason that frontal lesions induce ipsilesional neglect remains unclear. Denny-Brown and Chambers16 proposed that whereas normally the frontal lobe mediates avoidance behaviour, the parietal lobe mediates approach behaviour. Therefore, investigators have posited that frontal lesions cause a release of approach behaviours.9 The frontal lobe plays an important part in directed attention and intention, and these approach behaviours may be perceptual-attentional or action-intentional. In previous studies attempting to separate perceptual from action-motor factors in contralesional neglect, lesions of the frontal-subcortical structures were most often associated with motor-intentional neglect.17 In this respect, our findings using the video apparatus are somewhat anomalous. We found two out of five patients with ipsilesional neglect with performance indicative of primarily perceptual-attentional dysfunction. This finding is most compatible with an account of ipsilesional neglect, like contralesional neglect, resulting from either attentional or intentional approach behaviours toward stimuli in contralesional space.