用于鉴别良恶性实体肾肿块的影像-影像组学机器学习模型:与专家级放射科医师的比较
本回顾性研究纳入254例肾细胞癌、26例乏脂肪血管平滑肌脂肪瘤和10例嗜酸细胞瘤的术前CT,比较四名专家级放射科医师与影像-影像组学机器学习模型鉴别良恶性实体肾肿块的表现。专家间性能差异较大,机器学习模型在多项分类任务中敏感度和特异度较高,可能有助于提高阅片者间一致性和性能。
要点
- 专家级放射科医师鉴别良恶性实体肾肿块的敏感度和特异度变异较大。
- 影像-影像组学机器学习模型在多项分类任务中敏感度90.0%、86.3%、73.4%,特异度89.1%、83.3%、91.7%。
- 机器学习可能是提高阅片者间一致性和性能的潜在途径。
核心结论
- 专家级放射科医师鉴别良恶性实体肾肿块的表现差异明显较大。
- 影像-影像组学机器学习可能是提高阅片者间一致性和性能的潜在途径。
研究设计与样本
- 回顾性研究,纳入254例肾细胞癌(190例透明细胞型、38例嫌色细胞型、26例乳头状型)、26例乏脂肪血管平滑肌脂肪瘤和10例嗜酸细胞瘤的术前CT检查。
- 由四名专家级放射科医师(阅读超过3000例泌尿生殖系统CT和MRI检查)识别病灶并手动分割用于影像-影像组学分析。
- 采用10折交叉验证训练和验证疾病特异性支持向量机影像-影像组学机器学习模型,医师与模型性能比较采用McNemar检验。
诊断标准与关键数值
- 四名放射科医师鉴别透明细胞型肾细胞癌与乳头状型及嫌色细胞型肾细胞癌:敏感度73.7-96.8%(中位数84.5%,方差122.7%),特异度48.4-71.9%(中位数61.8%,方差161.6%)。
- 四名放射科医师鉴别透明细胞型肾细胞癌与乏脂肪血管平滑肌脂肪瘤及嗜酸细胞瘤:敏感度73.7-96.8%(中位数84.5%,方差122.7%),特异度52.8-88.9%(中位数80.6%,方差269.1%)。
- 四名放射科医师鉴别乳头状型及嫌色细胞型肾细胞癌与乏脂肪血管平滑肌脂肪瘤及嗜酸细胞瘤:敏感度28.1-60.9%(中位数84.5%,方差122.7%),特异度75.0-88.9%(中位数50.0%,方差191.1%)。
- 10折交叉验证后,影像-影像组学机器学习模型鉴别透明细胞型肾细胞癌与乳头状型及嫌色细胞型肾细胞癌、透明细胞型肾细胞癌与乏脂肪血管平滑肌脂肪瘤及嗜酸细胞瘤、乳头状型及嫌色细胞型肾细胞癌与乏脂肪血管平滑肌脂肪瘤及嗜酸细胞瘤的敏感度分别为90.0%、86.3%、73.4%,特异度分别为89.1%、83.3%、91.7%。
临床应用与阅片要点
- 原文结论未据此给出推荐,仅指出影像-影像组学机器学习可能是提高阅片者间一致性和性能的潜在途径。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
OBJECTIVE. The objective of our study was to compare the performance of radiologicradiomic machine learning (ML) models and expert-level radiologists for differentiation of benign and malignant solid renal masses using contrast-enhanced CT examinations. MATERIALS AND METHODS. This retrospective study included a cohort of 254 renal cell carcinomas (RCCs) (190 clear cell RCCs [ccRCCs], 38 chromophobe RCCs [chrRCCs], and 26 papillary RCCs [pRCCs]), 26 fat-poor angioleiomyolipomas, and 10 oncocytomas with preoperative CT examinations. Lesions identified by four expert-level radiologists (> 3000 genitourinary CT and MRI studies) were manually segmented for radiologicradiomic analysis. Disease-specific support vector machine radiologic-radiomic ML models for classification of renal masses were trained and validated using a 10-fold cross-validation. Performance values for the expert-level radiologists and radiologic-radiomic ML models were compared using the McNemar test. RESULTS. The performance values for the four radiologists were as follows: sensitivity of 73.7-96.8% (median, 84.5%; variance, 122.7%) and specificity of 48.4-71.9% (median, 61.8%; variance, 161.6%) for differentiating ccRCCs from pRCCs and chrRCCs; sensitivity of 73.7-96.8% (median, 84.5%; variance, 122.7%) and specificity of 52.8-88.9% for differentiating ccRCCs from fat-poor angioleiomyolipomas and oncocytomas (median, 80.6%; variance, 269.1%); and sensitivity of 28.1-60.9% (median, 84.5%; variance, 122.7%) and specificity of 75.0-88.9% for differentiating pRCCs and chrRCCs from fat-poor angioleiomyolipomas and oncocytomas (median, 50.0%; variance, 191.1%). After a 10-fold cross-validation, the radiologic-radiomic ML model yielded the following performance values for differentiating ccRCCs from pRCCs and chrRCCs, ccRCCs from fat-poor angioleiomyolipomas and oncocytomas, and pRCCs and chrRCCs from fat-poor angioleiomyolipomas and oncocytomas: a sensitivity of 90.0%, 86.3%, and 73.4% and a specificity of 89.1%, 83.3%, and 91.7%, respectively. CONCLUSION. Expert-level radiologists had obviously large variances in performance for differentiating benign from malignant solid renal masses. Radiologic-radiomic ML can be a potential way to improve interreader concordance and performance.
引用
- Sun XY, Feng QX, Xu X, et al. Radiologic-Radiomic Machine Learning Models for Differentiation of Benign and Malignant Solid Renal Masses: Comparison With Expert-Level Radiologists. AJR Am J Roentgenol. 2019. PMID: 31553660. doi:10.2214/AJR.19.21617
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。