Background: To evaluate the association between single nucleotide polymorphisms (SNPs) at the 194 and 399 codons of XRCC1, and the risk of severe acute skin and oral mucosa reactions in nasopharyngeal carcinoma patients in China. Methods: 114 patients with nasopharyngeal carcinoma were sequentially recruited in this study. Heparinized peripheral blood samples were taken for SNPs analysis before the start of radiation treatment. SNPs in XRCC1 (194Arg/Trp and 399Arg/Gln) gene were analyzed by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). Dermatitis at upper neck and oral mucositis were clinically recorded according to the Common Terminology Criteria for Adverse Events v.3.0. Results: The variant allele frequencies were 0.289 for XRCC1 194Trp and 0.263 for XRCC1 399Gln. Of the 114 patients, 24 experienced grade 3 acute dermatitis and 48 had grade 3 acute mucositis. The XRCC1 399Arg/Gln was significantly associated with the development of grade 3 dermatitis (Odds Ratio, 2.65; 95% CI, 1.04–6.73; p = 0.037, χ2 = 4.357). In addition, it was also associated with higher incidence of grade 3 mucositis with a borderline statistical significance (Odds Ratio, 2.11; 95% CI, 0.951–4.66; p = 0.065, χ2 = 3.411). The relationship between XRCC1 194Arg/Trp and acute dermatitis, and mucositis was not found. Conclusions: Our investigation shows, for the first time, that patients with the XRCC1 399Arg/Gln genotype were more likely to experience severe acute dermatitis and oral mucositis. With further validation, the information can be used to determine personalized radiotherapy strategy.

The rapid advancement of deep learning has established Convolutional Neural Networks (CNNs) as mainstream for medical image segmentation, yet their limited receptive field hinders long-range dependency capture. While Transformers excel at modeling global features via self-attention, their high computational complexity burdens high-resolution image processing. To leverage the complementary strengths of both architectures and integrate local and global features under a lightweight framework for enhanced accuracy and efficiency, this work proposes a novel encoder based on parallel CNN and Swin Transformer. Its effective integration is the Semantics and Detail Infusion (SDI) module, which fuses multi-scale features and employs attention to prioritize critical details, enriching features for decoder resolution recovery. Evaluations were conducted on two publicly available datasets, namely the Synapse Multi-Organ Segmentation dataset and the Aortic Vessel Tree dataset. The proposed model achieved Dice coefficients of 84.19% and 87.91%, respectively, and corresponding Hausdorff Distances of 12.64 mm and 7.06 mm. These results represent significant improvements over the UNet benchmark, with Dice score gains of 7.34% and 5.02%, respectively. The results further underscore the model’s robustness, efficiency, and clinical relevance in accurately delineating complex anatomical structures, particularly in abdominal segmentation tasks. By effectively fusing CNN and Transformer advantages, our approach meets high-performance standards for medical image segmentation while offering practical benefits for real-world clinical deployment in resource-constrained environments. The code is publicly available on https://github.com/Palpitate-v/HybridNet .