RESUMO
BACKGROUND: We previously proposed a novel virtual reality-based method to assess human field of perception (FOP) and field of regard (FOR), termed the FOPR test. This study assessed the diagnostic validity of the FOPR test for hemispatial neglect (HSN). METHODS: We included 19 stroke patients with a lesion in the right hemisphere and with HSN (HSN+SS), 22 stroke patients with a lesion in the right hemisphere and without HSN (HSN-SS), and 22 healthy controls aged 19-65 years. The success rate (SR) and response time (RT) in the FOPR test for both FOP and FOR were assessed (FOP-SR, FOR-SR, FOP-RT, and FOR-RT, respectively). Using a Bland-Altman plot, agreements between the FOPR test and conventional tests were confirmed, and the FOPR test accuracy was verified using the support vector machine (SVM). Measured values were analysed using ANOVA and Kruskall-Wallis tests for group comparison. RESULTS: The Bland-Altman plot showed good agreement between FOPR test and conventional tests; individuals within 95% agreement limits were within the range of 94.8-100.0%. The SVM classification accuracy, using FOP and FOR variables from the left hemispace, ranged from 83.3 to 100.0% in a binary classification (HSN vs non-HSN). The FOPR test demonstrated differences in SR and RT for both FOP and FOR across the groups. CONCLUSION: The FOPR test was valid for the HSN diagnosis and provided quantitative and intuitive information regarding visuospatial function. Furthermore, it might enhance our understanding of visuospatial function including HSN by applying the time relative component and concepts of perception and exploration, FOP and FOR. TRIAL REGISTRATION: NCT03463122. Registered 13 March 2018, retrospectively registered.