Algorithm for Morphological Analysis of Brain Magnetic Resonance Imaging Images in the Task of Differential Diagnosis of Primary Tumors

This paper proposes an algorithm for morphological analysis of brain images in the task of differential diagnosis of primary extra-axial tumors based on the processing of magnetic resonance imaging images. The concept of radiomics and its stages in the task of differential diagnosis of primary extra-axial tumors are considered. The analysis of applicability of geometrical features in the task of differential diagnostics of extra-axial tumors on the basis of magnetic resonance imaging images has been carried out. The classification accuracy was 97.74%, and the area under the receiver operating characteristic curve value was 0.97 on the presented sample of features.