Studies on automatic analysis of polarization images of tissues for early diagnosis of neoplastic transformation

The interrelations between statistics of the 1st-4th orders of the totality of Mueller-matrix images and geometric structure of birefringent architectonic nets of different morphological structure have been analyzed. It has been shown that the third and the fourth statistical moments of matrix elements two-dimensional distributions are the most sensitive to changing orientation structure of optically anisotropic protein fibrils of architectonics of physiologically normal and pathologically changed biological tissues.