The dedicated decision support system in recognition of some uncertain disease entities

This work presents the principles of image recognit io , where quality-based methods are applied. The n eural networks and additional software have been proposed. This goal w as achieved by using non-parametric recognition alg orithms. In this paper the two-state hybrid classification method has been pro posed, where artificial intelligence algorithm is i ncluded. In recognition process, the learning method, selection and optimization of diagnostic parameters have been introduced. The int egrated part of the classifier structure is voting mechanism, which indicates inco rrect states of the system – for example the unreco gnized images. Effectiveness of the system has been shown by means of examples, wh re ambiguous data have been incorporated – it is very often a practice of medical diagnostics.