Validation of the Keen Eye computer-based method for diagnosing visual neglect using a dual-task paradigm

Abstract

Visual neglect is a common and disabling consequence of right hemisphere damage. Standard paper-and-pencil assessments may fail to detect subtle or well-compensated cases of neglect, especially under low attentional demands. This study presents and validates the Keen Eye computer-based method for diagnosing neglect using a dual-task paradigm. The method involves the simultaneous detection of lateralized visual targets and identification of centrally presented digits, increasing attentional load and sensitivity to spatial biases. We tested 102 patients with right hemisphere damage (38 with neglect, 64 without) using a predefined set of target positions that systematically varied across the visual field. Classification models based on error patterns and asymmetry coefficients demonstrated high sensitivity and specificity in distinguishing patients with neglect. The method also revealed quadrant-specific and vertically biased omissions. The findings support the utility of attention-demanding computer-based tasks for improving diagnostic precision in visual neglect and suggest potential for identifying subclinical or hidden neglect profiles.