Algorithmic Software Package for Automated Diagnostic Analysis of Human Fundus Images Obtained by Optical Coherence Tomography Angiography: Version 2.0

The article is devoted to automated analysis of ophthalmological images obtained by optical coherence tomography angiography. Intravital examination of the retinal and choroidal vascular beds is widely used by the medical community to diagnose vascular and nonvascular pathologies. Examination of the retinal vasculature can detect a wide range of diseases, such as hypertension, diabetes, atherosclerosis, cardiovascular disease, and stroke. A new means of automating analysis of ophthalmological images is described: an algorithmic and software package that allows for highly accurate, autonomous differential diagnostics to separate the normal and pathological state of blood vessels using images obtained by optical coherence tomography angiography, version 2.0. In particular, the new interface of this algorithmic and software package is examined in detail. The main problems are formulated and solved: (1) redesign of the interface and visual component of the entire complex using the Solara framework for the Python programming language; (2) setting up support for images obtained using various microscopes; (3) establishing a cloud architecture of the algorithmic software package. The article provides examples of image analysis results demonstrating the potential of the system and directions for further research: it is planned to expand the list of supported image types, add new image analysis functionality, and study new features for more accurate diagnostics.