To address issues such as loss of detailed information, blurred target boundaries, and unclear structural hierarchy in medical image fusion, this paper proposes an adaptive feature medical image fusion network based on a full-scale diffusion model. First, a region-level feature map is generated using a kernel-based saliency map to enhance local features and boundary details. Then, a full-scale diffusion feature extraction network is employed for global feature extraction, alongside a multi-scale denoising U-shaped network designed to fully capture cross-layer information. A multi-scale feature integration module is introduced to reinforce texture details and structural information extracted by the encoder. Finally, an adaptive fusion scheme is applied to progressively fuse region-level features, global features, and source images layer by layer, enhancing the preservation of detail information. To validate the effectiveness of the proposed method, this paper validates the proposed model on the publicly available Harvard dataset and an abdominal dataset. By comparing with nine other representative image fusion methods, the proposed approach achieved improvements across seven evaluation metrics. The results demonstrate that the proposed method effectively extracts both global and local features of medical images, enhances texture details and target boundary clarity, and generates fusion image with high contrast and rich information, providing more reliable support for subsequent clinical diagnosis.
For pulmonary nodules in computed tomography (CT) images, which exhibit complex morphology and blurred boundaries, existing segmentation methods still fall short in modelling cross-level dependencies of multi-scale features, thereby limiting their performance in pulmonary nodule segmentation tasks. To address these challenges, this paper proposes a semantic segmentation method for pulmonary nodules based on multiscale feature interaction and cross-level coordinate attention (MFI-CLCA). This U-shaped network incorporated three architectures: a convolutional neural network (CNN), a Transformer, and Mamba. During the encoding phase, combining CNN and Mamba learning paradigms capured both global and local information in the input data. The convolutional component extracted complex boundary features of the target by combining multi-scale convolutional operations with adaptive fusion operations. Global and local multi-head attention mechanisms were introduced in the bottleneck layer and decoding phase respectively to model these hierarchical feature dependencies. The skip-connection section incorporated a multi-level coordinate attention module to adaptively focus on the information being passed through. Experimental results on the Lung Image Database Consortium (LIDC) dataset demonstrated that this approach achieved Dice scores of 90.52% and sensitivity of 91.93%, which outperforms existing state-of-the-art methods and validates its effectiveness for lung nodule segmentation tasks.
