Prediction of MGMT Promoter Methylation in Glioma Using Diffusion MRI-Based Habitat Subregion Analysis
10.3969/j.issn.1005-5185.2025.09.007
- VernacularTitle:基于扩散MRI的胶质瘤生境亚区分析预测MGMT启动子甲基化
- Author:
Huinan XIAO
1
;
Kaiji DENG
;
Wanyi ZHENG
;
Zhenxing WU
;
Yuting SHI
;
Yingying HE
;
Xue XU
;
Yunjing XUE
;
Rifeng JIANG
Author Information
1. 福建医科大学附属协和医院放疗科,福建 福州 350001
- Publication Type:Journal Article
- Keywords:
Glioma;
Magnetic resonance imaging;
Diffusion weighted imaging;
O(6)-methylguanine-DNA methyltransferase;
Methylation;
Molecular typing;
Habitat imaging;
Pathology,surgical;
Forecasting
- From:
Chinese Journal of Medical Imaging
2025;33(9):936-947
- CountryChina
- Language:Chinese
-
Abstract:
Purpose To evaluate the predictive performance of mean apparent propagator-magnetic resonance imaging(MAP-MRI)combined with habitat analysis for determining O6-methylguanine-DNA methyltransferase(MGMT)promoter methylation status in glioma.Materials and Methods This retrospective study analyzed MRI and clinical data from 55 patients with surgically confirmed glioma at Fujian Medical University Union Hospital from January 2019 to December 2023.All patients underwent structural and diffusion-weighted imaging.Three-dimensional volumes of interest were delineated in the tumor solid region using ImageJ software.The nn-FAE tool was used to segment the tumor solid region into two habitat subregions based on mean diffusivity(MD)maps:high-MD and low-MD habitats.Average diffusion parameter values were extracted from the entire tumor solid region and each habitat subregion.Differences in parameters between methylated and unmethylated groups were compared,and the area under the curve was calculated.Results Among 55 patients,significant differences were observed in all MAP-MRI parameters and MD in the tumor solid region and low-MD habitat,as well as all parameters in the high-MD habitat between methylated and unmethylated groups(t/Z=-3.780-3.153,all P<0.05).The return-to-origin probability(RTOP)in the low-MD habitat demonstrated the highest diagnostic performance,with the area under the curve improving from 0.771 before habitat analysis to 0.827 after habitat analysis.In the high-grade subgroup,significant differences were observed in return-to-axis probability(RTAP)and RTOP in the tumor solid region;RTOP,non-Gaussianity,non-Gaussianity axial,and RTAP in the low-MD habitat;and non-Gaussianity in the high-MD habitat(t/Z=-2.820--1.976,all P<0.05).RTOP in the low-MD habitat again showed optimal diagnostic efficacy(the area under the curve 0.725 before habitat analysis,0.798 after).Multivariate analysis identified RTAP and RTOP in the tumor solid region and low-MD habitat as independent predictors of MGMT methylation.Conclusion MAP-MRI diffusion parameters demonstrate the ability to predict MGMT promoter methylation status in glioma,with superior performance compared with diffusion tensor imaging.Habitat imaging further enhances the predictive efficacy of MAP-MRI parameters for MGMT promoter methylation.
