A.I. Pipeline for Accurate Retinal Layer Segmentation Using OCT 3D Images

Image data set from a multi-spectral animal imaging system is used to address two issues: (a) registering the oscillation in optical coherence tomography (OCT) images due mouse eye movement and (b) suppressing shadow region under thick vessels/structures. Several classical AI-based algorithms combination are tested for each task see their compatibility with combined system. Hybridization of AI flow followed by Homography transformation shown be working (correlation value>0.7) registration. Resnet50 backbone better than famous U-net model detection loss value 0.9. A simple-to-implement analytical equation brightness manipulation 1% increment mean pixel values 77% decrease number zeros. The proposed allows formulating constraint optimization problem using controlling factor {\alpha} minimization zeros, standard deviation maximizing value. For Layer segmentation, used. AI-Pipeline consists CNN, Optical flow, RCNN, model, models sequence. thickness estimation process has 6% error as compared manual annotated data.