A fuzzy reasoning system for boundary detection in radiological images

Abstract

In this paper, a new method of edge detection in blurred images, like radiological ones, has been developed. It is based on a fuzzy system which incorporates expert knowledge about the images. Fuzzy logic is used to obtain the edge of a bone in the radiological image without adjusting the system. Classical algorithms, like Sobel filter, obtain similar results, but the final user needs to modify different parameters in order to adapt the accuracy of the resulting image. The robustness of the fuzzy system allows in a second step to process the resulting image in order to detect different shapes. This second process is used for automatic medical image analysis. An illustrative computational fuzzy edge detection system has been implemented and its performance evaluated.