Nowadays, there is a tradeoff between the deep-learning module-compression ratio and the module accuracy. In this paper, a strategy for refining the pruning quantification and weights based on neural network filters is proposed. Firstly, filters in the neural network were refined into strip-like filter strips. Then, the evaluation of the filter strips was used to refine the partial importance of the filter, cut off the unimportant filter strips and reorganize the remaining filter strips. Finally, the training of the neural network after recombination was quantified to further compress the computational amount of the neural network. The results show that the method can significantly reduce the computational effort of the neural network and compress the number of parameters in the model. Based on experimental results on ResNet56, this method can reduce the number of parameters to 1/4 and the amount of calculation to 1/5, and the loss of model accuracy is only 0.01. On VGG16, the number of parameters is reduced to 1/14, the amount of calculation is reduced to 1/3, and the accuracy loss is 0.5%.

The task of nucleus segmentation plays an important role in medical image analysis. However, due to the challenge of detecting small targets and complex boundaries in datasets, traditional methods often fail to achieve satisfactory results. Therefore, a novel nucleus segmentation method based on the U-Net architecture is proposed to overcome this issue. Firstly, we introduce a Weighted Feature Enhancement Unit (WFEU) in the encoder decoder fusion stage of U-Net. By assigning learnable weights to different feature maps, the network can adaptively enhance key features and suppress irrelevant or secondary features, thus maintaining high-precision segmentation performance in complex backgrounds. In addition, to further improve the performance of the network under different resolution features, we designed a Double-Stage Channel Optimization Module (DSCOM) in the first two layers of the model. This DSCOM effectively preserves high-resolution information and improves the segmentation accuracy of small targets and boundary regions through multi-level convolution operations and channel optimization. Finally, we proposed an Adaptive Fusion Loss Module (AFLM) that effectively balances different lossy targets by dynamically adjusting weights, thereby further improving the model’s performance in segmentation region consistency and boundary accuracy while maintaining classification accuracy. The experimental results on 2018 Data Science Bowl demonstrate that, compared to state-of-the-art segmentation models, our method shows significant advantages in multiple key metrics. Specifically, our model achieved an IOU score of 0.8660 and a Dice score of 0.9216, with a model parameter size of only 7.81 M. These results illustrate that the method proposed in this paper not only excels in the segmentation of complex shapes and small targets but also significantly enhances overall performance at lower computational costs. This research offers new insights and references for model design in future medical image segmentation tasks.

In recent years, deep learning has achieved significant advancements in medical image segmentation, primarily focusing on CNN structures and Transformer-based architectures. However, these network architectures often suffer from high computational complexity and a large number of parameters. To address these challenges, this paper proposes a lightweight, novel structured network called LUT-SLS, based on U-Net and Transformer. Firstly, the overall structure integrates U-Net and Transformer, which effectively captures long-range dependencies in image relationships and contextual information, thereby improving segmentation accuracy. Secondly, a novel PLTS module is designed, which replaces the traditional self-attention mechanism with average pooling operations to extract global features and local details. Additionally, a novel MMLP structure is introduced, incorporating residual depth-separable operations into the traditional fully-connected framework. This enhances the processing of pooled features and further improves feature expression capability. Finally, the encoder and decoder parts are connected by the MSBN module, which facilitates the extraction of deep features while fusing encoder features. Experimental results demonstrate that the proposed model achieves competitive advantages in balancing the number of parameters, computational complexity, and performance compared to current leading models on multiple public datasets. This solution enables model deployment on IoT terminals, assisting doctors in making more accurate clinical decisions.