A local context fusion based segmentation model which uses a local context attention mechanism to filter out irrelevant feature information and enhance the attention to important regions is presented for accurate polyp segmentation. The features at different scales are captured by multi-kernel dilated convolution for improving the accuracy of polyp boundary segmentation. Pyramid context selection module utilizes shallow encoder features to compensate for the low-level information lost by the deeper encoder,enabling the model to adapt to polyps of various sizes. The proposed model achieves accuracies of 97.67％,97.19％ and 99.23％ on Kvasir-SEG,EndoScene and CVC-ClinicDB datasets,respectively,with mIoU of 91.20％,88.31％ and 94.75％,respectively,exhibiting higher accuracy and generalizability than the existing classical methods and validating its superior performance in polyp segmentation. The proposed model can improve polyp segmentation accuracy and provide a more accurate aid for polyp segmentation.

A local context fusion based segmentation model which uses a local context attention mechanism to filter out irrelevant feature information and enhance the attention to important regions is presented for accurate polyp segmentation. The features at different scales are captured by multi-kernel dilated convolution for improving the accuracy of polyp boundary segmentation. Pyramid context selection module utilizes shallow encoder features to compensate for the low-level information lost by the deeper encoder,enabling the model to adapt to polyps of various sizes. The proposed model achieves accuracies of 97.67％,97.19％ and 99.23％ on Kvasir-SEG,EndoScene and CVC-ClinicDB datasets,respectively,with mIoU of 91.20％,88.31％ and 94.75％,respectively,exhibiting higher accuracy and generalizability than the existing classical methods and validating its superior performance in polyp segmentation. The proposed model can improve polyp segmentation accuracy and provide a more accurate aid for polyp segmentation.

Given that the imbalance of semantic feature transfer caused by the skip connection of the brain tumor image segmentation network and the insufficient correlation of multi-scale features lead to the loss of details,resulting in poor segmentation accuracy for small tumors,an improved segmentation model of the Res-Unet framework is proposed.The model introduces a multi-scale attention fusion module which makes the model better adaptable to tumors of different sizes by mixing multi-scale features,and adds a spatial attention module to the skip connection to enhance feature expression while avoiding the interference of useless information,preserving the spatial details of feature maps.Through the auxiliary classifier module,the decoder performs feature prediction on feature maps of different scales.The BraTS2020 dataset is used for experiments and evaluations,and the model segmentation performance is evaluated with Dice score.The results show that the improved network achieved average Dice scores of 0.887 7,0.822 9,and 0.802 7 for whole tumor,tumor core,and enhancing tumor,respectively.Compared with the channel attention model,the improved model increases the scores of enhancing tumor and tumor core by 2.6%and 0.14%,respectively,which proves its effectiveness and accuracy for brain tumor MR image segmentation.