Search

A reciprocal relationship between perceptual learning and functional brain changes towards perceptual learning effectiveness has been demonstrated previously; however, the underlying neural correlates remain unclear. Further, visual perceptual learning (VPL) is implicated in visual field defect (VFD) recovery following chronic stroke. We investigated resting-state functional connectivity (RSFC) in the visual cortices associated with mean total deviation (MTD) scores for VPL-induced VFD recovery in chronic stroke. Patients with VFD due to chronic ischemic stroke in the visual cortex received 24 VPL training sessions over 2 months, which is a dual discrimination task of orientation and letters. At baseline and two months later, the RSFC in the ipsilesional, interhemispheric, and contralesional visual cortices and MTD scores in the affected hemi-field were assessed. Interhemispheric visual RSFC at baseline showed the strongest correlation with MTD scores post-2-month VPL training. Notably, only the subgroup with high baseline interhemispheric visual RSFC showed significant VFD improvement following the VPL training. The interactions between the interhemispheric visual RSFC at baseline and VPL led to improvement in MTD scores and largely influenced the degree of VFD recovery. The interhemispheric visual RSFC at baseline could be a promising brain biomarker for the effectiveness of VPL-induced VFD recovery. [ABSTRACT FROM AUTHOR]

Introduction: Visual field defects (VFDs) represent a debilitating poststroke complication, characterized by unseen parts of the visual field. Visual perceptual learning (VPL), involving repetitive visual training in blind visual fields, may effectively restore visual field sensitivity in cortical blindness. This current multicenter, double‐blind, randomized, controlled clinical trial investigated the efficacy and safety of VPL‐based digital therapeutics (Nunap Vision [NV]) for treating poststroke VFDs. Methods: Stroke outpatients with VFDs (>6 months after stroke onset) were randomized into NV (defective field training) or Nunap Vision‐Control (NV‐C, central field training) groups. Both interventions provided visual perceptual training, consisting of orientation, rotation, and depth discrimination, through a virtual reality head‐mounted display device 5 days a week for 12 weeks. The two groups received VFD assessments using Humphrey visual field (HVF) tests at baseline and 12‐week follow‐up. The final analysis included those completed the study (NV, n = 40; NV‐C, n = 35). Efficacy measures included improved visual area (sensitivity ≥6 dB) and changes in the HVF scores during the 12‐week period. Results: With a high compliance rate, NV and NV‐C training improved the visual areas in the defective hemifield (>72 degrees2) and the whole field (>108 degrees2), which are clinically meaningful improvements despite no significant between‐group differences. According to within‐group analyses, mean total deviation scores in the defective hemifield improved after NV training (p =.03) but not after NV‐C training (p =.12). Conclusions: The current trial suggests that VPL‐based digital therapeutics may induce clinically meaningful visual improvements in patients with poststroke VFDs. Yet, between‐group differences in therapeutic efficacy were not found as NV‐C training exhibited unexpected improvement comparable to NV training, possibly due to learning transfer effects. [ABSTRACT FROM AUTHOR]