IgA肾病MEST-C分级的深度学习自动化
本研究训练并评估神经网络自动化MEST-C分级的工具。活检分训练组(n=42)、测试组(n=66)和应用组(n=88)。测试组中网络正确识别>73%像素为M、E、S或C;应用组预测M1、E1、S1、T1、T2、C1和C2的AUC分别为0.88、0.91、0.88、0.94、0.96、0.96和0.92;网络S和T评分与复合生存终点相关,风险比分别为9.67(P=.006)和7.67(P<.001)。
要点
- 测试组中网络正确识别>73%像素为M、E、S或C。
- 应用组预测M1、E1、S1、T1、T2、C1和C2的AUC分别为0.88、0.91、0.88、0.94、0.96、0.96和0.92。
- 网络S和T评分与复合生存终点相关,风险比分别为9.67(P=.006)和7.67(P<.001)。
核心结论
- 本研究凸显了使用深度学习方法自动识别和量化MEST-C分类各要素的可能性。
研究设计与样本
- 活检分为三个独立组:训练组(n=42)用于训练网络,测试组(n=66)用于比较网络像素分割与病理学家分割,应用组(n=88)用于比较网络与病理学家计算的MEST-C评分。
主要发现
- 测试组中,网络正确识别>73%的像素为M、E、S或C。
- 应用组中,神经网络预测M1、E1、S1、T1、T2、C1和C2的受试者工作特征曲线下面积分别为0.88、0.91、0.88、0.94、0.96、0.96和0.92。
- 病理学家与网络评估之间的kappa系数对于E、S、T和C评分分别为0.68、0.79、0.73和0.70,对于M评分为0.52。
- 网络S和T评分与复合生存终点(死亡、透析、移植或血清肌酐加倍)的发生相关,风险比分别为9.67(P=.006)和7.67(P<.001)。
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)。
原文
BACKGROUND: Although the MEST-C classification is among the best prognostic tools in immunoglobulin A nephropathy (IgAN), it has a wide interobserver variability between specialized pathologists and others. Therefore we trained and evaluated a tool using a neural network to automate the MEST-C grading. METHODS: Biopsies of patients with IgAN were divided into three independent groups: the Training cohort (n = 42) to train the network, the Test cohort (n = 66) to compare its pixel segmentation to that made by pathologists and the Application cohort (n = 88) to compare the MEST-C scores computed by the network or by pathologists. RESULTS: In the Test cohort, >73% of pixels were correctly identified by the network as M, E, S or C. In the Application cohort, the neural network area under the receiver operating characteristics curves were 0.88, 0.91, 0.88, 0.94, 0.96, 0.96 and 0.92 to predict M1, E1, S1, T1, T2, C1 and C2, respectively. The kappa coefficients between pathologists and the network assessments were substantial for E, S, T and C scores (kappa scores of 0.68, 0.79, 0.73 and 0.70, respectively) and moderate for M score (kappa score of 0.52). Network S and T scores were associated with the occurrence of the composite survival endpoint (death, dialysis, transplantation or doubling of serum creatinine) [hazard ratios 9.67 (P = .006) and 7.67 (P < .001), respectively]. CONCLUSIONS: This work highlights the possibility of automated recognition and quantification of each element of the MEST-C classification using deep learning methods.
引用
- Jaugey A, Maréchal E, Tarris G, et al. Deep learning automation of MEST-C classification in IgA nephropathy. Nephrol Dial Transplant. 2023. PMID: 36792061. doi:10.1093/ndt/gfad039
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。