Multi-perspective feature compensation enhanced network for medical image segmentation

Abstract

Medical image segmentation’s accuracy is crucial for clinical analysis and diagnosis. Despite progress with U-Net-inspired models, they often underuse multi-scale convolutional layers crucial for enhancing detailing visual features and overlooking the importance of merging multi-scale features within the channel dimension to enhance decoder complexity. To address these limitations, we introduce a Multi-perspective Feature Compensation Enhanced Network (MFCNet) for medical image segmentation. Our network design is characterized by the strategic employment of dual-scale convolutional kernels at each encoder level. This synergy enables the precise capture of both granular and broader context features throughout the encoding phase. We further enhance the model by integrating a Dual-scale Channel-wise Cross-fusion Transformer (DCCT) mechanism within the skip connections. This innovation effectively integrates dual-scale features. We subsequently implemented the spatial attention (SA) mechanism to amplify the saliency areas within the dual-scale features. These enhanced features were subsequently merged with the feature map of the same level in the decoder, thereby augmenting the overall feature representation. Our proposed MFCNet has been evaluated on three distinct medical image datasets, and the experimental results demonstrate that it achieves more accurate segmentation performance and adaptability to varying target segmentation, making it more competitive compared to existing methods. The code is available at: https://github.com/zrm-code/MFCNet.