Mutimodal Mutual Attention Network

— Medical image segmentation is a crucial aspect of medical image processing, and has been widely used in the detection and clinical diagnosis for brain, lung, liver, heart and other diseases. In this paper, we propose a novel multimodal mutual attention network, called MMAUNet, for medical image segmentation. MMA-UNet is divided into two parts. The first part obtains more highdimensional features by skip connection and improved network structure. The second part incorporates a multimodal mutual attention mechanism, encompassing feature mutual attention, spatial mutual attention, and channel mutual attention. This mechanism facilitates the effective fusion of high-dimensional and low-dimensional features, leading to enhanced context information. Experimental results on Kagglelung dataset, Liver dataset, Cell dataset, Drive dataset and Kvasir-SEG dataset show that MMA-UNet has achieved better segmentation performance than that of other baseline methods, on lung, liver, cell contour, retinal vessel and polyps, etc.