Objective: "Visual Restitution Therapies" (VRT) claim to ameliorate visual field defects of neurological patients by repeated visual light stimulation, leading to training-related neuroplasticity and resulting in reconnection of lesioned neurons in early cortical areas. Because existing systems are stationary, uncomfortable, and unreliable, we developed a training instrument based on virtual reality goggles. The goal of the "Salzburg Visual Field Trainer" (SVFT) is twofold: (1) The device facilitates the clinical evaluation of established neuropsychological rehabilitation approaches, such as VRT. (2) The device enables patients to independently perform VRT based (or other) neuropsychological training methodologies flexibly and comfortably.
Methods and analysis: The SVFT was developed on the principles of VRT. Individual configuration of the SVFT is based on perimetric data of the respective patient's visual field. To validate the utmost important aspect of neuropsychological rehabilitation methodologies-that is displaying stimuli precisely in desired locations in the user's visual field-two steps were conducted in this proof-of-concept study: First, we assessed the individual "blind spots" location and extent of 40 healthy, normal sighted participants. This was done with the help of our recently developed perimetric methodology "Eye Tracking Based Visual Field Analysis" (EFA). Second, depending on the individual characteristics of every participant's blind spots, we displayed-by means of the SVFT-15 stimuli in the respective locations of every participants' blind spots and 85 stimuli in the surrounding, intact visual area. The ratio between visible and non-visible stimuli, which is reflected in the behavioral responses (clicks on a remote control) of the 40 participants, provides insight into the accuracy of the SVFT to display training stimuli in areas desired by the investigator. As the blind spot is a naturally occurring, absolute scotoma, we utilized this blind area as an objective criterion and a "simulated" visual field defect to evaluate the theoretical applicability of the SVFT.
Results: Outcomes indicate that the SVFT is highly accurate in displaying training stimuli in the desired areas of the user's visual field with an accuracy of 99.0%. Data analysis further showed a sensitivity of .98, specificity of .99, a positive predictive value of .96, a negative predictive value of .996, a hit rate of .99, a random hit rate of .74 and a RATZ-Index of .98. This translates to 14.7% correct non-reactions, 0.7% false non-reactions, 0.3% false reactions and 84.3% correct reactions to displayed test stimuli during the evaluation study. Reports from participants further indicate that the SVFT is comfortable to wear and intuitive to use.
Conclusions: The SVFT can help to investigate the true effects of VRT based methodologies (or other neuropsychological approaches) and the underlying mechanisms of training-related neuroplasticity in the visual cortex in neurological patients suffering from visual field defects.