Xiangliang Tan1, Kan Deng2, Yingjie Mei2, Tianjing Zhang2, Yang Song3, Qiaoli Yao1, and Yikai Xu1
1Medical Imaging Center, Nanfang Hospital, Southern Medical University, Guangzhou, China, 2Philips Healthcare, Guangzhou, China, 3Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China
1Medical Imaging Center, Nanfang Hospital, Southern Medical University, Guangzhou, China, 2Philips Healthcare, Guangzhou, China, 3Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China
Our study suggests that the grey matter
volume parameter is an effective classification feature for the radiomics
models to identify non-NPSLE patients from HC subjects.
Figure 2(A) ROC
curves of the non-NPSLE classification on different classifiers (B) Histograms
of AUC on different classifiers. SVM, support vector machine, LDA, linear
discriminant analysis, LR, logistic regression.
Figure 1. Brain
structures with grey matter volume decreased in non-NPSLE compared to healthy
controls