Analýza medicínskych snímok

Abstract

Detection of abnormalities in chest X-ray images is often crucial for diagnosing and determining the course of treatment for various diseases. This thesis focuses on the use of neural networks for the classification and localization of these abnormalities. The theoretical part explores existing methods of applying neural networks to this problem, preprocessing techniques for improving detection, the metrics used, and available datasets. The practical part describes the implemented solutions for classifying and localizing abnormalities, including data preparation and preprocessing methods. The results of the experiments are evaluated and compared with existing solutions.