Local Minimum-Maximum Normalized Masking: Python tool for post-processing and quantification of en face OCT angiograms of choriocapillaris

2019-09-23T15:49:42Z (GMT) by Brennan Marsh-Armstrong
Image processing pipeline designed to reduce noise, accent vasculature, and quantify anatomical properties of flattened en-face image of the choriocapillaris taken from ultrahigh-speed FDML-OCT volumes. It uses a custom designed Local Minimum Maximum Normalized Masking (LMNM) algorithm as well as a graph-data-type representation of vasculature to accomplish this. The anatomical properties of vasculature it currently quantifies and visualizes are flow-void radius, flow-area, vessel radius, vessel length, and number of branches per branch point. File contains python 2.7 script, configuration text file, and example image to demonstrate processing. See accompanying publication or internal code comments for more details.