In this paper, a neural hybrid image classification for intelligent diagnosis systems from signal to image conversion (image representation) is suggested. Such hybrid approach (multiple models) mainly aims to ensure a satisfactory reliability for faults diagnosis systems, and particularly for medical diagnosis. Thus, an overview is given on how neural global and local approximators are interesting for image classification and why image classification (from signals to images conversion) is efficient than signal classification for fault diagnosis systems. Then, a neural hybrid image classification approach is suggested for intelligent medical diagnosis help, from biomedical signals, using MLP and RBF networks, under supervised learning. In this approach, each image is divided in several sub-images (local indicators) which are classified by global approxirnators (MLP) and by local approxirnators (RBF). Afterwards, a fuzzy decision-making system is suggested to give the final diagnosis with a Confidence Index (CI) parameter.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.