Abstract:Objective To explore the clinical feasibility of detecting the penumbra of white matter hyperintensity (WMHp) by
radiomics nomogram based on T2WI and FLAIR MR images. Methods we performed a retrospective analysis of 57 patients with
white matter high signal progression and 57 health examinees from 2012 to 2017 in our hospital. All of the subjects were then
divided into training set (n=80) and validation set (n=34). Texture features were extracted from T2WI and FLAIR images in the
WMHp of each patient and healthy subject, respectively. Radiomics signature was established by dimensionally reducing features
in the training set by the least absolute shrinkage and selection operator, and the prediction model was constructed with clinical risk
factors as potential predictors of Logistic regression analysis, then the nomogram was made. Performance of nomogram and clinical
application value was evaluated by data of validation set. Results The accuracy of nomogram to predict WMHp was 0.881 and 0.862
in the training and validation sets, respectively. Hosmer-lemeshow test showed no statistically significant difference between training
and validation sets (P>0.05). Receiver opercting characteristic curves were used to evaluate the accuracy of the nomogram, radiomics
signature and diabetes factors,with the AUC of 0.829, 0.819 and 0.64, respectively. Conclusion The radiomics monogram based on
multi-parameter MR images can be used as a quantitative tool to help clinicians identify WMHp for early intervention treatment.
舒震宇,徐玉芸,宋侨伟. 基于多参数磁共振成像的影像组学列线图在识别脑白质高信号半暗带中的应用研究[J]. 中国医疗设备, 2020, 35(5): 96-100.
SHU Zhenyu, XU Yuyun, SONG Qiaowei. Application of Radiomics Nomogram Based on Multi-Parameter of Magnetic Resonance
Imaging for Detect the Penumbra of White Matter Hyperintensity. China Medical Devices, 2020, 35(5): 96-100.