Methods of the Straightening Transformation and Analysis of Outer Contours of Objects in Medical Image Recognitio

Abstract

One of the most important problems appearing in computer systems for the recognition and analysis of medical images, supporting diagnostics and medical therapy, is the processing of obtained images or their series in such a way that it is possible to unambiguously determine the disease symptoms occurring in them. Based on the detected symptoms, it is then possible to identify various disease entities, which in turn is the starting point for planning further therapy. Characterizing the process of automatic medical diagnostics, it can be stated that information in computer diagnostic systems can come from data on the occurrence of disease symptoms obtained during the analysis of medical images thanks to the use of various methods of artificial intelligence. Where subjected to recognition there are images on which any morphological changes of any of the organs can be observed, i.e. in the recognition of diseases visible on images obtained by various computed tomography techniques or radiographic images (e.g. ERCP, angiograms), the purpose of analyzing such images is to identify and unequivocal recognition of such morphological changes occurring in the examined organ. To this end, syntactic image recognition methods are often used. However, before proceeding to the proper extraction of the sought-after features that indicate the occurrence of disease symptoms, it is usually necessary to pre-process the analyzed photos so as to improve their quality, remove unnecessary (from the point of view of the diagnosis made) details while separating the recognized object