B – 10 Correlation Between a Computerized Cognitive Assessment and Traditional Neuropsychological Assessment Measures with Epilepsy Patients

Abstract Objective This study examined the convergent validity of Cognivue® (COGV), a brief computerized cognitive assessment, by evaluating domain-specific correlations with traditional neuropsychological measures in patients with epilepsy. A secondary objective explored associations between COGV domain performance and seizure type/location. Method Fifty-two adults (27 male; M age = 49.65, SD = 18.91; M education = 14.94 years, SD = 2.38) completed neuropsychological assessment in an outpatient neurology clinic. Seizure type and location varied, including frontotemporal, temporal, frontoparietal, frontal, and occipital focus. Participants completed COGV subtests (e.g., Memory, Visual Salience, Executive Functioning), Hopkins Verbal Learning Test-Revised (HVLT-R), Brief Visuospatial Memory Test-Revised (BVMT-R), Boston Naming Test (BNT), Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV), and Delis-Kaplan Executive Function System (DKEFS) subtests. Spearman’s correlations assessed relationships between COGV subtests and traditional measures. One-way ANOVA evaluated cognitive performance by seizure focus. Results Significant correlations were found between COGV Memory and HVLT Total (r = .553, p = .021) and between multiple subtests on COGV, WAIS-IV, and DKEFS (p < .001), supporting convergent validity. One-way ANOVA revealed significant differences in COGV performance based on seizure location for COGV Visual Salience (p = .033) and COGV Executive Functioning (p = .033), with lower scores in individuals with seizures originating in the occipital region. Additional near-significant differences were evidenced. Conclusion Findings support the clinical potential of COGV as a brief, valid, and sensitive tool for assessing cognitive function in epilepsy populations. The results also underscore potential relevance of seizure focus on cognitive profiling with COGV and highlight the need for replication in larger samples.