将囊性肾脏肿块分层为良性与潜在恶性:应用机器学习于Bosniak分类
本研究基于2011年1月至2018年6月共257个囊性肾脏肿块,由三位放射科医生按Bosniak 2019版分类为185个良性(I或II类)和72个潜在恶性(IIF、III或IV类),提取6个CT纹理特征并用随机森林、逻辑回归和支持向量机建模,三种算法AUC分别为0.88、0.90和0.89,在区分良性与潜在恶性囊性肾脏肿块方面表现出较高判别能力。
要点
- 基于CT纹理的机器学习模型对Bosniak 2019版定义的良性与潜在恶性囊性肾脏肿块有较高判别能力,随机森林、逻辑回归、支持向量机AUC分别为0.88、0.90、0.89。
- 纹理特征均值、标准差、正像素均值、熵、偏度升高与潜在恶性组统计学相关(每项P≤0.0015)。
- 作者认为若经验证,CT纹理机器学习算法可能有助于减少应用Bosniak分类时的阅片者间变异。
核心结论
- 三种基于CT纹理的机器学习算法在区分Bosniak 2019版定义的良性与潜在恶性囊性肾脏肿块方面表现出较高判别能力。
- 作者认为若经验证,基于CT纹理的机器学习算法可能有助于减少应用Bosniak分类时的阅片者间变异。
研究设计与样本
- 本研究为经IRB批准、符合HIPAA的研究,2011年1月至2018年6月共4,454例成人患者接受肾脏肿块协议CT或CT尿路造影。
- 最终研究队列纳入257个囊性肾脏肿块,由三位放射科医生独立按Bosniak 2019版分类,结果为185个良性(Bosniak I或II类)和72个潜在恶性(Bosniak IIF、III或IV类)肿块。
- 使用商用软件TexRAD(Feedback PLC,Cambridge,UK)提取均值、标准差、正像素均值、熵、偏度、峰度6个纹理特征。
- 采用随机森林、逻辑回归和支持向量机三种机器学习算法将囊性肾脏肿块分为两组,并以十折交叉验证进行测试。
诊断标准与关键数值
- 随机森林模型敏感度0.67、特异度0.91、阳性预测值0.75、阴性预测值0.88、AUC 0.88。
- 逻辑回归模型敏感度0.63、特异度0.93、阳性预测值0.78、阴性预测值0.86、AUC 0.90。
- 支持向量机模型敏感度0.56、特异度0.91、阳性预测值0.71、阴性预测值0.84、AUC 0.89。
- 纹理特征均值、标准差、正像素均值、熵、偏度升高与潜在恶性组统计学相关(每项P≤0.0015)。
临床应用与阅片要点
- 作者认为若经验证,基于CT纹理的机器学习算法可能有助于减少应用Bosniak分类时的阅片者间变异。
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
PURPOSE: To create a CT texture-based machine learning algorithm that distinguishes benign from potentially malignant cystic renal masses as defined by the Bosniak Classification version 2019.
METHODS: In this IRB-approved, HIPAA-compliant study, 4,454 adult patients underwent renal mass protocol CT or CT urography from January 2011 to June 2018. Of these, 257 cystic renal masses were included in the final study cohort. Each mass was independently classified using Bosniak version 2019 by three radiologists, resulting in 185 benign (Bosniak I or II) and 72 potentially malignant (Bosniak IIF, III or IV) masses. Six texture features: mean, standard deviation, mean of positive pixels, entropy, skewness, kurtosis were extracted using commercial software TexRAD (Feedback PLC, Cambridge, UK). Random forest (RF), logistic regression (LR), and support vector machine (SVM) machine learning algorithms were implemented to classify cystic renal masses into the two groups and tested with tenfold cross validations.
RESULTS: Higher mean, standard deviation, mean of positive pixels, entropy, skewness were statistically associated with the potentially malignant group (P ≤ 0.0015 each). Sensitivity, specificity, positive predictive value, negative predictive value, and area under curve of RF model was 0.67, 0.91, 0.75, 0.88, 0.88; of LR model was 0.63, 0.93, 0.78, 0.86, 0.90, and of SVM model was 0.56, 0.91, 0.71, 0.84, 0.89, respectively.
CONCLUSION: Three CT texture-based machine learning algorithms demonstrated high discriminatory capability in distinguishing benign from potentially malignant cystic renal masses as defined by the Bosniak Classification version 2019. If validated, CT texture-based machine learning algorithms may help reduce interreader variability when applying the Bosniak classification.
引用
- Miskin N, Qin L, Matalon SA, et al. Stratification of cystic renal masses into benign and potentially malignant: applying machine learning to the bosniak classification. Abdom Radiol (NY). 2020. PMID: 32613401. doi:10.1007/s00261-020-02629-w
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。