Department of Urology, University Medical Center Goettingen, Goettingen, Germany.
Department of Clinical and Interventional Radiology, University Medical Center Goettingen, Goettingen, Germany.
Eur Radiol. 2024 Oct;34(10):6254-6263. doi: 10.1007/s00330-024-10731-6. Epub 2024 Apr 18.
To distinguish histological subtypes of renal tumors using radiomic features and machine learning (ML) based on multiphase computed tomography (CT).
Patients who underwent surgical treatment for renal tumors at two tertiary centers from 2012 to 2022 were included retrospectively. Preoperative arterial (corticomedullary) and venous (nephrogenic) phase CT scans from these centers, as well as from external imaging facilities, were manually segmented, and standardized radiomic features were extracted. Following preprocessing and addressing the class imbalance, a ML algorithm based on extreme gradient boosting trees (XGB) was employed to predict renal tumor subtypes using 10-fold cross-validation. The evaluation was conducted using the multiclass area under the receiver operating characteristic curve (AUC). Algorithms were trained on data from one center and independently tested on data from the other center.
The training cohort comprised n = 297 patients (64.3% clear cell renal cell cancer [RCC], 13.5% papillary renal cell carcinoma (pRCC), 7.4% chromophobe RCC, 9.4% oncocytomas, and 5.4% angiomyolipomas (AML)), and the testing cohort n = 121 patients (56.2%/16.5%/3.3%/21.5%/2.5%). The XGB algorithm demonstrated a diagnostic performance of AUC = 0.81/0.64/0.8 for venous/arterial/combined contrast phase CT in the training cohort, and AUC = 0.75/0.67/0.75 in the independent testing cohort. In pairwise comparisons, the lowest diagnostic accuracy was evident for the identification of oncocytomas (AUC = 0.57-0.69), and the highest for the identification of AMLs (AUC = 0.9-0.94) CONCLUSION: Radiomic feature analyses can distinguish renal tumor subtypes on routinely acquired CTs, with oncocytomas being the hardest subtype to identify.
Radiomic feature analyses yield robust results for renal tumor assessment on routine CTs. Although radiologists routinely rely on arterial phase CT for renal tumor assessment and operative planning, radiomic features derived from arterial phase did not improve the accuracy of renal tumor subtype identification in our cohort.
使用基于多期 CT 的放射组学特征和机器学习 (ML) 来区分肾肿瘤的组织学亚型。
回顾性纳入 2012 年至 2022 年在两家三级中心接受肾肿瘤手术治疗的患者。对这些中心以及外部影像设施的术前动脉（皮质髓质）期和静脉（肾性）期 CT 扫描进行手动分割，并提取标准化的放射组学特征。在预处理和解决类别不平衡问题后，使用基于极端梯度增强树 (XGB) 的 ML 算法，通过 10 倍交叉验证来预测肾肿瘤亚型。使用多类接收器工作特征曲线下面积 (AUC) 进行评估。算法在一个中心的数据上进行训练，并在另一个中心的数据上进行独立测试。
训练队列包括 n = 297 例患者（64.3%为透明细胞肾细胞癌 [RCC]，13.5%为乳头状肾细胞癌 [pRCC]，7.4%为嫌色细胞 RCC，9.4%为嗜酸细胞瘤，5.4%为血管平滑肌脂肪瘤 [AML]），而测试队列 n = 121 例患者（56.2%/16.5%/3.3%/21.5%/2.5%）。XGB 算法在训练队列中的静脉/动脉/联合对比期 CT 中表现出 AUC = 0.81/0.64/0.8 的诊断性能，在独立测试队列中 AUC = 0.75/0.67/0.75。在两两比较中，识别嗜酸细胞瘤的诊断准确性最低（AUC = 0.57-0.69），识别 AML 的准确性最高（AUC = 0.9-0.94）。
放射组学特征分析可在常规 CT 上区分肾肿瘤亚型，其中嗜酸细胞瘤最难识别。
放射组学特征分析为常规 CT 上的肾肿瘤评估提供了可靠的结果。尽管放射科医生通常依赖动脉期 CT 进行肾肿瘤评估和手术计划，但在我们的队列中，源自动脉期的放射组学特征并未提高肾肿瘤亚型识别的准确性。
