Mammary gland nuclear magnetic resonance focus segmentation system based on DINOv2

The invention relates to the technical field of focus segmentation, in particular to a DINOv2-based breast nuclear magnetic resonance focus segmentation system, which comprises an input module, a preprocessing module, a feature extraction module, a region evaluation module, a segmentation module, an output module and a storage module, and is characterized in that the input module is used for receiving breast nuclear magnetic resonance image data; the preprocessing module is used for denoising and standardizing the image data transmitted by the input module and generating preprocessed image data; the feature extraction module is used for performing feature extraction on the preprocessed image by adopting a DINOv2 algorithm to generate a feature map; the region evaluation module is used for evaluating the density of the mammary tissue by using a statistical model and outputting an evaluation result of whether the region is a high-risk region or not; and the segmentation module is used for segmenting a lesion ar