Background NeuroEyeCoach (NEC) is a rehabilitation intervention programme aimed at those with visual field defects as a result of brain trauma, stroke, other insults or surgical procedures. It alleviates the disability associated with visual field defect by improving the patient’s visual scanning technique. Smaller studies using similar stimulus configurations have shown success of eye-movement therapy on visual search performance metrics, such as shorter search times and less detection errors (Zihl 1985; Meienberg et al., 1981; Zihl & Hebel 1997). To evaluate the effect of rehabilitation on patient’s activities of daily life, a questionnaire is often used where the patient self-reports their experience of disability on 10 aspects of daily life on a 5 point Likert-scale (Nelles et al 2001). NEC therefore includes visual search tasks and a disability questionnaire pre and post intervention to evaluate outcome. NEC is delivered through an online programme, adaptive to the patient’s visual disability and self-administered at home or in a clinical setting (Sahraie et al., 2016). NEC was developed by NovaVision Inc. (Boca Raton, Florida), is registered with the FDA as 510(k) exempt medical device in the US and CE registered as Type I Medical Device in the EU; and. Over 200 participants have completed NEC to-date. Research Aim The aim of the study is determine the change in visual search performance metrics and self-reported disability after a course of NEC through retrospective data analysis of already collected data from patients that have completed this therapy. The specific questions that we will explore in this analysis include: 1. Does the use of NEC in patients with partial blindness after brain injury result in improved visual search? This will be determined by comparing the median reaction time, number of errors and correct detection of presented stimuli before and after NEC. 2. Do patients who use NEC have less disability? We will analyse the pre and post treatment self-reported disability score. 3. Are there discernible predictors of good response to NEC? The improvements in visual search parameters and self-reported disability will be analysed in a multivariable model adjusting for age, sex, time from injury, side of visual field defect, baseline disability, site of NEC performance (home vs. clinic), duration of treatment, and stimulus parameters. Investigators Jose G. Romano, MD, is Professor of Clinical Neurology and Chief of the Stroke Division at the University of Miami, Miller School of Medicine. He will serve as the corresponding investigator for this proposal, and will contribute to data interpretation and final report. He has long interest and experience in vision recovery after stroke, serving as scientific advisor to NovaVision, the company that developed NEC. He also has experience managing multi-center stroke trials: he is PI of the MaRISS study (Genentech), MPI of the Regional Coordinating Center for the Stroke Trials Network (NIH-NINDS), MPI of the Transitions of Stroke Care Disparities Study (NIH-NIMHD), MPI of the MyRIAD study (NIH-NINDS), was MPI of the MoSIS Study (NIH-NINDS), and served as site PI of multiple NIH/NINDS funded clinical trials. He is a member of the StrokeNet Prevention Working Group and the Trial Recruitment and Retention Advisory Committee, and member of the AAN Axon Registry Committee. Arash Sahraie is Professor of Psychology in University of Aberdeen, UK, with over 25 years of research experience in visual impairment following brain injury. He has published extensively on capacities and limitations of brain mechanisms underpinning residual visual processing in injured brain. He has developed a restitution technique for alleviation of the sight loss after occipital brain injury and has worked collaboratively to develop a compensatory rehabilitation intervention based on improved eye-movement to compensate for the resultant sight loss. He is Chief Scientific Advisor of NovaVision Inc. Sigrid Kenkel, psychologist, is head of NovaVision Clinical Services, attending patients in computer-based vision rehabilitation for the last 20 years. During this time she has contributed to many NovaVision research studies. Study Procedures Data This study will include a retrospective analysis of existing data; no patient-related interventions will take place. A data review of all consecutive cases who have used NEC between November 2014 and September 2018 will be conducted. Sigrid Kenkel from NovaVision will audit the data and exclude non-patient related information (e.g., entries related to device demonstrations) or cases when the patients did not complete the therapy. The final dataset to be analysed will be stripped of any personal identifiers. The following variables will be included: sex; age; visual field defect side; time from insult to start of treatment; duration of treatment; stimulus parameters for visual search in set sizes of 4, 8, 16 and 25 items; number of errors and correct detections; average and median reaction times (RT) for stimuli presented on the left or right side of fixation for each set size; and self-reported disability scale before and after therapy. Data analysis will be conducted on all completed sets as below. NovaVision will transfer the de-identified and encrypted dataset to a secure server at the University of Aberdeen. The data on secure server will only be accessed by AS. Data Analysis 1. Does the use of NEC in patients with partial blindness after brain injury result in improved visual search? The median data for target presentation on the left and right will be averaged for each set size and for each patient. To obtain an aggregate effect of therapy on visual search this RT data will be averaged across all set sizes and the pre- versus post- data will be plotted. The percentage of data points falling below the line with a slope of 1 will represent the proportion of patient who have shown improved visual search. A paired-sample t-test on mean RT before and after therapy will indicate if there is significant effect of NEC. RT for visual search will also be analysed separately for blind and sighted hemifield to establish whether the effect of therapy is restricted to one or both hemifields. Again paired sample t-test will be conducted for each hemifield (blind/sighted) separately. Average number of errors for pre and post therapy will also be compared (paired sample t-test). 2. Do patients who use NEC have less disability? A plot of disability scores obtained pre- vs. post-training will be obtained. Again, percentage of data points falling below the line with slope of 1 will indicate the proportion of those improving following NEC. Again, paired sample t-test will indicate if the improvement is significant. We will also sub-divide the patient pool to three groups based on their level of disability; namely those with low self-reported disability score (<10), medium (10-20) and high disability score (>20) and report the percentage of cases which fall below the line with slope of 1 for each group. 3. Are there discernible predictors of good response to NEC? Having a large dataset allow us to examine the predictors of recovery. There are two objective and one subjective measure of improvements. These are changes in RT and Error Rates in visual search tasks; and subjective rating of disability. We will use multivariate regression analysis to find out if factors such as age, sex, time from injury and side of visual field defect can predict the level of improvements. Statistical significance corrected for multiple comparisons As only 5 t-tests are planned, all p values will be corrected and those with significance of p< .01 will be reported. References J. Zihl, “Visual scanning behavior in patients with homonymous hemianopia,” Neuropsychologia, vol. 33, no. 3, pp. 287–303, 1995. O. Meienberg, W. H. Zangemeister, M. Rosenberg, W. F. Hoyt, and L. Stark, “Saccadic eyemovement strategies in patients with homonymous hemianopia,” Annals of Neurology, vol. 9, no. 6, pp. 537–544, 1981. J. Zihl and N. Hebel, “Patterns of oculomotor scanning in patients with unilateral posterior parietal or frontal lobe damage,” Neuropsychologia, vol. 35, no. 6, pp. 893–906, 1997. G. Nelles, J. Esser, A. Eckstein, A. Tiede, H. Gerhard, and H. C. Diener, “Compensatory visual field training for patients with hemianopia after stroke,” Neuroscience Letters, vol. 306, no. 3, pp. 189–192, 2001. A. Sahraie, N. Smania, J. Zihl, “Use of NeuorEyeCoachTM to improve eye movement efficacy in patients with homonymous visual field loss” BioMed Research International, Article ID 5186461. 2016.