Image Processing and Syndrome Feature Analysis to Improve Image Interpretation

摘要

A series of methods have been developed, which are to support medical image analysis and interpretation in the case of data uncertainty and the lack of information for reliable and correct image interpretation. The methods allow the better representation of information features for expert's analysis; they use the expert's domain knowledge and help to recover new knowledge from the database of processes image descritions. Created formal decision rules for image analysis and interpretation imitate human argumentation, and present the solution in easily interpretable form. Developed methods have been applied for early peripheral lung cancer diagnosis. Thier use helped to enhance expert's diagnostic abilities and essentially imporved results of medical image analysis and decision-making for the experts of different qualifcations.