Kashyap Mehr, Wang Xi, Panjwani Neil, Hasan Mohammad, Zhang Qin, Huang Charles, Bush Karl, Chin Alexander, Vitzthum Lucas K, Dong Peng, Zaky Sandra, Loo Billy W, Diehn Maximilian, Xing Lei, Li Ruijiang, Gensheimer Michael F
Stanford University School of Medicine, Department of Medicine, Stanford, CA, US.
Stanford University School of Medicine, Department of Radiation Oncology, Stanford, CA, US.
Radiology. 2025 Jan;314(1):e233029. doi: 10.1148/radiol.233029.
Background Detection and segmentation of lung tumors on CT scans are critical for monitoring cancer progression, evaluating treatment responses, and planning radiation therapy; however, manual delineation is labor-intensive and subject to physician variability. Purpose To develop and evaluate an ensemble deep learning model for automating identification and segmentation of lung tumors on CT scans. Materials and Methods A retrospective study was conducted between July 2019 and November 2024 using a large dataset of CT simulation scans and clinical lung tumor segmentations from radiotherapy plans. This dataset was used to train a 3D U-Net-based, image-multiresolution ensemble model to detect and segment lung tumors on CT scans. Model performance was evaluated on internal and external test sets composed of CT simulation scans and lung tumor segmentations from two affiliated medical centers, including single primary and metastatic lung tumors. Performance metrics included sensitivity, specificity, false positive rate, and Dice similarity coefficient (DSC). Model-predicted tumor volumes were compared with physician-delineated volumes. Group comparisons were made with Wilcoxon signed-rank test or one-way ANOVA. P < 0.05 indicated statistical significance. Results The model, trained on 1,504 CT scans with clinical lung tumor segmentations, achieved 92% sensitivity (92/100) and 82% specificity (41/50) in detecting lung tumors on the combined 150-CT scan test set. For a subset of 100 CT scans with a single lung tumor each, the model achieved a median model-physician DSC of 0.77 (IQR: 0.65-0.83) and an interphysician DSC of 0.80 (IQR: 0.72-0.86). Segmentation time was shorter for the model than for physicians (mean 76.6 vs. 166.1-187.7 seconds; p<0.001). Conclusion Routinely collected radiotherapy data were useful for model training. The key strengths of the model include a 3D U-Net ensemble approach for balancing volumetric context with resolution, robust tumor detection and segmentation performance, and the ability to generalize to an external site.
在CT扫描上检测和分割肺部肿瘤对于监测癌症进展、评估治疗反应以及规划放射治疗至关重要；然而，手动勾勒轮廓劳动强度大且受医生差异影响。目的：开发并评估一种用于在CT扫描上自动识别和分割肺部肿瘤的集成深度学习模型。材料与方法：于2019年7月至2024年11月进行了一项回顾性研究，使用了来自放射治疗计划的CT模拟扫描和临床肺部肿瘤分割的大型数据集。该数据集用于训练基于3D U-Net的图像多分辨率集成模型，以在CT扫描上检测和分割肺部肿瘤。在由两个附属医院的CT模拟扫描和肺部肿瘤分割组成的内部和外部测试集上评估模型性能，包括单发原发性和转移性肺部肿瘤。性能指标包括敏感性、特异性、假阳性率和Dice相似系数（DSC）。将模型预测的肿瘤体积与医生勾勒的体积进行比较。采用Wilcoxon符号秩检验或单因素方差分析进行组间比较。P<0.05表示具有统计学意义。结果：该模型在1504例带有临床肺部肿瘤分割的CT扫描上进行训练，在150例CT扫描的联合测试集上检测肺部肿瘤时，敏感性达到92%（92/100），特异性达到82%（41/50）。对于每个包含单个肺部肿瘤 的100例CT扫描子集，模型与医生之间的DSC中位数为0.77（IQR：0.65 - 0.83），医生之间的DSC为0.80（IQR：0.72 - 0.86）。模型的分割时间比医生短（平均76.6秒对166.1 - 187.7秒；p<0.001）。结论：常规收集的放射治疗数据对模型训练有用。该模型的关键优势包括用于平衡体积上下文与分辨率 的3D U-Net集成方法、强大的肿瘤检测和分割性能以及推广到外部机构的能力。
