基于对比增强多期 CT 的小肾肿块深度学习评估
Deep Learning Assessment of Small Renal Masses at Contrast-enhanced Multiphase CT.
机构信息
From the Departments of Radiology (C.D., P.Z., Z.S., M.Z., J.Z.), Urology (Y.X., J.G.), and Pathology (J.H.), Zhongshan Hospital, Fudan University, Shanghai, China; Shanghai Institute of Medical Imaging, Shanghai, China (C.D., P.Z., Z.S., M.Z.); Department of Radiology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China (L.Y., F.C.); Departments of Urology (J.L.) and Radiology (J.Z.), Xiamen Branch, Zhongshan Hospital, Fudan University, 668 Jinhu Road, Huli District, Xiamen 361015, China; Department of Urology, Zhangye People's Hospital affiliated to Hexi University, Zhangye, China (J.Y.); Department of Radiology, the First People's Hospital of Lianyungang, Lianyungang, China (X.Z.); Department of Radiology, Quanzhou First Hospital, Fujian Medical University, Quanzhou, China (R.H.); Department of Pathology, Sir Run Run Shaw Hospital, Hangzhou, China (R.W.); Digital Medical Research Center, School of Basic Medical Sciences, Fudan University, Shanghai, China (K.W., S.W.); Shanghai Key Laboratory of MICCAI, Shanghai, China (K.W., S.W.); Xiamen Municipal Clinical Research Center for Medical Imaging, Xiamen, China (J.Z.); and Xiamen Key Clinical Specialty, Xiamen, China (J.Z.).
出版信息
Radiology. 2024 May;311(2):e232178. doi: 10.1148/radiol.232178.
Background Accurate characterization of suspicious small renal masses is crucial for optimized management. Deep learning (DL) algorithms may assist with this effort. Purpose To develop and validate a DL algorithm for identifying benign small renal masses at contrast-enhanced multiphase CT. Materials and Methods Surgically resected renal masses measuring 3 cm or less in diameter at contrast-enhanced CT were included. The DL algorithm was developed by using retrospective data from one hospital between 2009 and 2021, with patients randomly allocated in a training and internal test set ratio of 8:2. Between 2013 and 2021, external testing was performed on data from five independent hospitals. A prospective test set was obtained between 2021 and 2022 from one hospital. Algorithm performance was evaluated by using the area under the receiver operating characteristic curve (AUC) and compared with the results of seven clinicians using the DeLong test. Results A total of 1703 patients (mean age, 56 years ± 12 [SD]; 619 female) with a single renal mass per patient were evaluated. The retrospective data set included 1063 lesions (874 in training set, 189 internal test set); the multicenter external test set included 537 lesions (12.3%, 66 benign) with 89 subcentimeter (≤1 cm) lesions (16.6%); and the prospective test set included 103 lesions (13.6%, 14 benign) with 20 (19.4%) subcentimeter lesions. The DL algorithm performance was comparable with that of urological radiologists: for the external test set, AUC was 0.80 (95% CI: 0.75, 0.85) versus 0.84 (95% CI: 0.78, 0.88) ( = .61); for the prospective test set, AUC was 0.87 (95% CI: 0.79, 0.93) versus 0.92 (95% CI: 0.86, 0.96) ( = .70). For subcentimeter lesions in the external test set, the algorithm and urological radiologists had similar AUC of 0.74 (95% CI: 0.63, 0.83) and 0.81 (95% CI: 0.68, 0.92) ( = .78), respectively. Conclusion The multiphase CT-based DL algorithm showed comparable performance with that of radiologists for identifying benign small renal masses, including lesions of 1 cm or less. Published under a CC BY 4.0 license.
背景 准确描述可疑的小肾肿块对于优化管理至关重要。深度学习（DL）算法可能有助于这一努力。目的 开发和验证一种用于在对比增强多期 CT 中识别良性小肾肿块的 DL 算法。材料与方法 本研究纳入了在对比增强 CT 上直径为 3cm 或更小的手术切除的肾肿块。该 DL 算法是通过使用一家医院在 2009 年至 2021 年期间的回顾性数据开发的，患者被随机分配到训练集和内部测试集的比例为 8:2。2013 年至 2021 年期间，对来自五家独立医院的数据进行了外部测试。2021 年至 2022 年期间，从一家医院获得了前瞻性测试集。使用受试者工作特征曲线下的面积（AUC）评估算法性能，并使用 DeLong 检验与七位临床医生的结果进行比较。结果 共评估了 1703 名患者（平均年龄 56 岁±12[标准差]；619 名女性），每位患者均有单个肾肿块。回顾性数据集包括 1063 个病变（874 个在训练集中，189 个在内部测试集中）；多中心外部测试集包括 537 个病变（12.3%，66 个良性），其中 89 个亚厘米（≤1cm）病变（16.6%）；前瞻性测试集包括 103 个病变（13.6%，14 个良性），其中 20 个（19.4%）亚厘米病变。DL 算法的性能与泌尿科放射科医生相当：对于外部测试集，AUC 为 0.80（95%CI：0.75，0.85）与 0.84（95%CI：0.78，0.88）（ =.61）；对于前瞻性测试集，AUC 为 0.87（95%CI：0.79，0.93）与 0.92（95%CI：0.86，0.96）（ =.70）。对于外部测试集中的亚厘米病变，算法和泌尿科放射科医生的 AUC 分别为 0.74（95%CI：0.63，0.83）和 0.81（95%CI：0.68，0.92）（ =.78）。结论 基于多期 CT 的 DL 算法在识别良性小肾肿块方面与放射科医生的表现相当，包括 1cm 或更小的病变。根据 CC BY 4.0 许可发布。
Zotero 插件