Segmentation and ЗD Reconstruction of MRI Images for Improving Vein Diseases Diagnostic

Magnetic resonance imaging (MRI) is one of the most accurate and precise methods for diagnosing many diseases. Using MRI makes it possible to obtain detailed information about the structure and the state of organs, blood vessels, and tissues. The process of analyzing MRI images can take a lot of time and effort, as it requires manual analysis of dozens of slices made in different planes. Preliminary automatic segmentation and 3D-reconstruction can simplify this stage. Starting from the mid-2010s, when the widespread of the neural networks reached medical tasks, it became clear that artificial intelligence can compete with humans in some computer vision tasks. This lead to a drastic improvement of the previous results and the emergence of new perspectives. This paper presents information concerning the approach to kidney segmentation using convolutional neural networks and statistical analysis. The resulting model makes fast predictions, shows promising results, and can be used either for visualization or for creating a training set for more complex architectures to improve accuracy.