Medical image 3D visualization is one of the fundamental processes in medical diagnostics. Using the acquired 3D images it is possible to find the volume of pathology zone, which is the evidence of a specific disease. As a consequence, two problems emerge: visualization of 3D medical images and pathology zone extraction from acquired images. Available imaging software in some cases provides the con-struction of 3D images based upon various medical data ob-tained by computer tomography, magnetic resonance imaging, scintigraphy, etc. But further processing of these images (image segmentation, pathology zone extraction) can result in loss of information during initial image reconstruction. Furthermore, existing medical imaging software does not provide the automatic extraction of regions of interest. As a consequence, often pathology zone volumes are calculated manually and that results in lack of precision. In this work methods for solving these problems are pro-posed. This work describes an approach of 3D model recon-struction from medical images by using detailed initial infor-mation obtained for forming DICOM files. For extraction of pathology zones and volume estimation an automatic proce-dure of region calculation is proposed. The described methods can provide practical improvements to the reliability of medical diagnostics.