V originále
Visualization of 3D medical images is usually accompanied by necessity of image data segmentation, in order to present interpretable resulting images. A method based on pattern recognition in multimodality images is proposed here. Results achieved by testing the proposed methods on images from Visible Human Project are presented.