BackgroundEarly and accurate identification of lymphatic node metastasis (LNM) and lymphatic vascular space invasion (LVSI) for endometrial cancer (EC) patients is important for treatment design, but difficult on multi‐parametric MRI (mpMRI) images.PurposeTo develop a deep learning (DL) model to simultaneously identify of LNM and LVSI of EC from mpMRI images.Study TypeRetrospective.PopulationSix hundred twenty‐one patients with histologically proven EC from two institutions, including 111 LNM‐positive and 168 LVSI‐positive, divided into training, internal, and external test cohorts of 398, 169, and 54 patients, respectively.Field Strength/SequenceT2‐weighted imaging (T2WI), contrast‐enhanced T1WI (CE‐T1WI), and diffusion‐weighted imaging (DWI) were scanned with turbo spin‐echo, gradient‐echo, and two‐dimensional echo‐planar sequences, using either a 1.5 T or 3 T system.AssessmentEC lesions were manually delineated on T2WI by two radiologists and used to train an nnU‐Net model for automatic segmentation. A multi‐task DL model was developed to simultaneously identify LNM and LVSI positive status using the segmented EC lesion regions and T2WI, CE‐T1WI, and DWI images as inputs. The performance of the model for LNM‐positive diagnosis was compared with those of three radiologists in the external test cohort.Statistical TestsDice similarity coefficient (DSC) was used to evaluate segmentation results. Receiver Operating Characteristic (ROC) analysis was used to assess the performance of LNM and LVSI status identification. P value <0.05 was considered significant.ResultsEC lesion segmentation model achieved mean DSC values of 0.700 ± 0.25 and 0.693 ± 0.21 in the internal and external test cohorts, respectively. For LNM positive/LVSI positive identification, the proposed model achieved AUC values of 0.895/0.848, 0.806/0.795, and 0.804/0.728 in the training, internal, and external test cohorts, respectively, and better than those of three radiologists (AUC = 0.770/0.648/0.674).Data ConclusionThe proposed model has potential to help clinicians to identify LNM and LVSI status of EC patients and improve treatment planning.Evidence Level3Technical EfficacyStage 2

中文翻译：

---

通过深度学习从多参数 MRI 全自动识别子宫内膜癌淋巴结转移和淋巴管侵犯

背景早期准确识别子宫内膜癌（EC）患者的淋巴结转移（LNM）和淋巴管间隙侵犯（LVSI）对于治疗设计很重要，但在多参数磁共振成像（mpMRI）图像上存在困难。 DL）模型从 mpMRI 图像中同时识别 EC 的 LNM 和 LVSI。研究类型回顾性。人群来自两个机构的 621 例经组织学证明的 EC 患者，其中 111 例 LNM 阳性和 168 例 LVSI 阳性，分为训练组、内部组、分别包含 398、169 和 54 名患者的外部测试队列。场强/序列 T2 加权成像 (T2WI)、对比增强 T1WI (CE-T1WI) 和扩散加权成像 (DWI) 采用涡轮旋转扫描使用 1.5 T 或 3 T 系统进行回波、梯度回波和二维回波平面序列。AssessmentEC 病变由两名放射科医生在 T2WI 上手动描绘，并用于训练 nnU-Net 模型以进行自动分割。我们开发了一个多任务深度学习模型，使用分段的 EC 病变区域和 T2WI、CE-T1WI 和 DWI 图像作为输入，同时识别 LNM 和 LVSI 阳性状态。将 LNM 阳性诊断模型的性能与外部测试队列中三名放射科医生的性能进行比较。统计测试 Dice 相似系数 (DSC) 用于评估分割结果。接受者操作特征 (ROC) 分析用于评估 LNM 和 LVSI 状态识别的性能。磷值<0.05被认为是显着的。结果EC病变分割模型在内部和外部测试队列中分别实现了平均DSC值0.700±0.25和0.693±0.21。对于 LNM 阳性/LVSI 阳性识别，所提出的模型在训练组、内部测试组和外部测试组中分别实现了 0.895/0.848、0.806/0.795 和 0.804/0.728 的 AUC 值，并且优于三位放射科医生的 AUC 值（AUC = 0.770/0.648/0.674)。数据结论所提出的模型有潜力帮助临床医生识别 EC 患者的 LNM 和 LVSI 状态并改善治疗计划。证据级别 3 技术功效阶段 2