Development of a Clinical Decision Support Tool for Early-Detection of Directed Energy Induced Injury to the Retina

To develop a support tool for detecting retina damage, first a clinical decision support algorithm for rapid analysis of fundus images of the retina before and after laser injury was developed. The retinal images then went through a rigorous image enhancement phase, image data augmentation, and pixel normalization. Following the enhancement and augmentation processes, algorithms were developed to identify retinal abnormalities associated with laser damage. The laser –induced injury detection technology in human fundus images was optimized to achieve greater than 95% accuracy. Defining the parameters including ease of use, sensitivity and specificity, a strategy to acquire FDA approval as a clinical decision support tool will be developed. Algorithms will be transitioned to develop a fully functional prototype for commercial and clinical partners in Phase III.