“Head and neck tumors are common malignant tumors in China, and lymph node metastasis is a key factor affecting their prognosis. Recently, technology media reported that in response to the problems of lesion information loss, low contrast, and blurred boundaries in magnetic resonance imaging, an expert team has proposed a lymph node segmentation network for metastatic head and neck tumors to assist doctors in diagnosis. This network achieves accurate detection and segmentation of lymph node metastasis lesions through cross layer and cross field attention modules, multi-scale feature fusion modules, and enhanced attention prediction head modules, which has positive significance for assisting lymph node diagnosis. The experimental results show that the network has achieved significant results in the segmentation of lymph node metastasis lesions.”