基于多模态组学的乳腺癌腋窝淋巴结精准处理模型建立与验证

乳腺癌治疗中腋窝淋巴结的处理对控制肿瘤转移和改善预后至关重要，单纯依靠传统影像学评估腋窝淋巴结状态的准确率无法满足临床需求，而依靠外科手术评估腋窝淋巴结状态常会导致严重的术后并发症，极大降低患者的生活质量。申请人研究团队拥有多学科交叉型人才，前期在影像、人工智能算法、基因、病理等多个层面对腋窝淋巴结状态预测积累了大量研究基础，并收集了近万例乳腺癌患者的临床数据及随访信息，拟通过深度学习神经网络在影像组学基础上结合基因组学、病理组学以及临床信息建立大数据、高通量、多维度预测模型，用以全方位评估腋窝淋巴结转移状态，从而指导外科医生选择最恰当的腋窝淋巴结处理方式，进一步提高治疗疗效并提高患者的生存质量，同时完善现有诊疗指南中有关腋窝淋巴结手术的流程，为乳腺癌精准手术治疗提供更多依据。

The treatment of axillary lymph nodes is very important to control tumor metastasis and improve patients’ prognosis in the management of breast cancer. Assessing the status of axillary lymph nodes by using traditional imaging alone cannot meet the clinical needs. However, assessing by surgery would often lead to severe complications, greatly reducing patients' quality of life. Composed of multidisciplinary and innovative talents, our team have done numerous researches about axillary lymph node metastasis in the fields of artificial intelligence algorithm, radiomics, genomics, pathomics and clinical information and collected clinical data and follow-up information of nearly 10000 breast cancer patients. Now, we are trying to evaluate the status of axillary lymph node metastasis comprehensively by establishing a prediction model based on big data, high-throughput and multi-dimension. The model would be possible to guide surgeons dealing with axillary lymph nodes in a more reasonable way and improve both the therapeutic effect and the quality of patients’ life. Meanwhile, it would also improve the procedure of axillary lymph node surgery in the present guidelines, thus providing more evidence for breast cancer precision surgery.