Timely detection of human diseases is impossible without modern analysis methods, such as those used in medical digital image processing. This improves diagnostic efficiency and reduces health risks. Based on this, a new approach to edge detection in RGB medical images is proposed. This approach is based on the combined use of Canny and Roberts edge detection operators in individual color channels, taking into account the neutrosophic ground truth domain. The obtained results are compared using metrics for the qualitative perception of the resulting images, quantitative assessments of the effectiveness of edge detection, and input image segmentation. A series of images after edge detection using the proposed approach is also presented. Overall, the effectiveness and feasibility of the proposed approach for subsequent medical decision-making are emphasized. The significant potential of the considered approach is demonstrated in comparison with classical methods of using edge detection operators.

Terms—diagnostics, digital medical imaging, edge detection, health risk, human disease, human health