Radiomics analysis based on T2-weighed imaging and T2 mapping for staging endometrial fibrosis

基于T2加权成像和T2映射的放射组学分析在子宫内膜纤维化分期中的应用

阅读:1

Abstract

Endometrial fibrosis can lead to uterine infertility. Accurate staging of endometrial fibrosis is crucial for developing treatment plans and performing dynamic follow-ups. This study aimed to evaluate the feasibility of radiomics models based on T2-weighed imaging (T2WI) and T2 mapping for staging endometrial fibrosis. This prospective study included 120 patients with severe endometrial fibrosis (SEF) and 50 patients with mild-moderate endometrial fibrosis (MMEF) confirmed by hysteroscopy, and 100 healthy controls (HC). Radiomic features were extracted from the volume of interest of endometrium on T2WI images and T2 maps to generate three models: T2WI, T2 mapping, and both T2WI and T2 mapping (merged). Feature importance selection was assessed with recursive feature elimination (RFE). Subsequently, logistic regression (LR), support vector machine (SVM), and decision tree (DT) were developed to determine the optimal radiomics models. Endometrial thickness (ET) and mean T2 value (Mean T2) were analyzed to construct ET+T2 model. The performance of the models was evaluated using receiver operating characteristic curve analysis and area under the curve (AUC). The merged radiomics model constructed by LR showed the highest performance with the macro and micro average AUC of 0.897 and 0.898, sensitivity of 0.744 and 0.873, specificity of 0.880 and 0.816, precision of 0.738 and 0.873, F1-score of 0.740 and 0.873, respectively. The LR-merged radiomics model had better classification performance [AUC (macro/micro), 0.897/0.898; overall accuracy, 0.765] than that of the ET+T2 model [AUC (macro/micro), 0.788/0.786; overall accuracy, 0.593]. Radiomics analysis based on T2WI and T2 mapping had the potential for the noninvasively staging endometrial fibrosis.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。