Fast in vivo assessment of the urinary bladder connective tissue state for evaluation of the radiation injury severity

In this work we proposed an algorithm for “operator-independent” quantitative assessment of the bladder images under various pathological conditions obtained by the cross-polarization optical coherence tomography (CP OCT). The aim of our study was to develop a method for operative diagnostics of the bladder connective tissue matrix state based on CP OCT data, which would allow calculating with certain accuracy the probability of development of the severe complications during radiation therapy of the pelvic organs diseases. "OKT-1300U" equipment was used. 4 groups of CP OCT sets of bladder images were analyzed for numerical analysis and for their automatic classification: contingent standards (n = 22), chronic cystitis with predominance of epithelial atrophy and fibrosis in the mucosa and submucosa structures (n = 122), I-II (n = 40) and III-IV (n = 34) degrees of radiation injury. An algorithm based on the analysis of principal component and a Random Forest Tree classification algorithm were applied for automatic identification of pathological changes in CP OCT images...