Prostate accidental cancer prediction method and system based on multi-task learning

The invention discloses a prostate accidental cancer prediction method and system based on multi-task learning, and relates to the technical field of image processing, and the method comprises the following steps: S1, preprocessing mpMRI images of a prostate accidental cancer patient and a benign prostatic hyperplasia patient, and obtaining overlapped patch blocks through fine-grained segmentation; s2, inputting the patch blocks overlapped in the step S1 into a hierarchical Transform encoder to obtain multi-level feature maps with different resolutions; and S3, performing up-sampling on the multi-level semantic feature map obtained in the S2 to obtain features of the same dimension, and then performing splicing to form a new feature map. The method is convenient to realize prediction of the prostate accidental cancer. 本发明公开基于多任务学习的前列腺偶发癌预测方法及系统，涉及图像处理技术领域，包括以下步骤：S1：对前列腺偶发癌患者和良性前列腺增生患者的mpMRI图像进行预处理，通过细粒度的分割得到重叠的补丁块；S2：将所述S1中重叠的补丁块输入到层级Transformer编码器中来获取不同分辨率的多层次特征图；S3：将所述S2中得到的多层次语义特征图，进行上采样得到相同维度特征，然后进行拼接，形成新