Share Email Print
cover

Proceedings Paper

Artificial Kohonen's neural networks for computer capillarometry
Author(s): Sergey Doncow; Leonid Orbachevskyi; Valentin Birukow; Nina V. Stepanova
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Analysis of disorders in microhemocirculation appearing during development of various genesis pathological processes allows to manifest their start-up and pathogenic mechanisms. Numerical estimation of capillary and arteriolar-veinular bed parameters gives a possibility to predict development of these states. Automatically controlled estimation of even simple parameters of microhemocirculation bed, such as vessel length in the eye conjuctiva, their amount, size, winding, etc., allows to speak about a new method in the express diagnostics of computer capillarometry. The present work is devoted to the methods of automated determination of microvessel length. The problem is solved by application of one-dimension Kohonen's networks for adaptive uniform piecewise approximation of capillary images. Kohonen's units are the approximation points, and their position is set during the self-organization of the neural network.

Paper Details

Date Published: 1 April 1998
PDF: 3 pages
Proc. SPIE 3402, Optical Information Science and Technology (OIST97): Optical Memory and Neural Networks, (1 April 1998); doi: 10.1117/12.304962
Show Author Affiliations
Sergey Doncow, Academy of Medical Technical Sciences/Joint Stock Co. MACDEL (Russia)
Leonid Orbachevskyi, Academy of Medical Technical Sciences/Joint Stock Co. MACDEL (Russia)
Valentin Birukow, Academy of Medical Technical Sciences/Joint Stock Co. MACDEL (Russia)
Nina V. Stepanova, Academy of Medical Technical Sciences/Joint Stock Co. MACDEL (Russia)


Published in SPIE Proceedings Vol. 3402:
Optical Information Science and Technology (OIST97): Optical Memory and Neural Networks
Andrei L. Mikaelian, Editor(s)

© SPIE. Terms of Use
Back to Top