A Novel Downsampling Strategy Based on Information Complementarity for Medical Image Segmentation

In convolutional neural networks (CNNs), downsampling operations are crucial to model performance. Although traditional downsampling methods (such as max pooling and strided convolution) perform well in feature aggregation, receptive field expansion, and computational reduction, they may lead to the loss of key spatial information in semantic segmentation tasks, thereby affecting the pixel-by-pixel prediction accuracy. To this end, this study proposes a downsampling method based on information complementarity - Hybrid Pooling Downsampling (HPD). The core is to replace the traditional method with MinMaxPooling, and effectively retain the light and dark contrast and detail features of the image by extracting the maximum value information of the local area. Experiment on various CNN and Transformer architectures using the ACDC and Synapse datasets demonstrate that HPD outperforms traditional segmentation methods. Specifically HPD improves the mean Dice similarity coefficient by 0.5%. The results show that the HPD module provides an efficient solution for semantic segmentation tasks. Code is available at https://github.com/apple1986/HPD.