Pattern recognition algorithms for the diagnosis of skin mycoses and onychomycosis
Andrey M. Alekseev1, Julia A. Brodskaya1; 1Gagarin's State Technical University of Saratov, Saratov, Russia
Abstract
The paper discusses the construction and implementation of algorithms for the diagnosis (recognition) of skin mycoses and onychomycoses based on clinical signs and symptoms, followed by the development of decision support software. A prototype of an expert system has been developed to automate the primary diagnosis of skin mycoses and onychomycosis through the logical output of swi-prolog. The study also generates a specialized data set for subsequent evaluation of the effectiveness of recognition algorithms, which should include images of affected tissues, as well as the results of blood tests for the presence of antibodies and antigens. Since accurate assessment of the size of the skin lesion, especially with irregular foci, is a difficult task and results in frequent diagnostic errors (in 50-60% of cases), the use of an automated diagnostic system in this class of diseases will reduce the number of diagnostic errors. The results of the work are discussed with experts in the subject area.
Speaker
Andrey M. Alekseev
Gagarin's State Technical University of Saratov
Russia
Discussion
Ask question
