MRI-based habitat imaging predicts high-risk molecular subtypes and early risk assessment of lower-grade gliomas
- Xiangli Yang 1,2,3, Wenju Niu 3, Kai Wu 4, Guoqiang Yang 5,6,7, Hui Zhang 8,9,10,11
- Xiangli Yang 1,2,3, Wenju Niu 3, Kai Wu 4
- 1Department of Radiology, First Hospital of Shanxi Medical University, Taiyuan, 030001, China.
- 2Shanxi Bethune Hospital, Third Hospital of Shanxi Medical University, Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Taiyuan, 030032, China.
- 3College of Medical Imaging, Shanxi Medical University, Taiyuan, 030001, China.
- 4Department of Information Management, First Hospital of Shanxi Medical University, Taiyuan, 030001, China.
- 5Department of Radiology, First Hospital of Shanxi Medical University, Taiyuan, 030001, China. yangguoqiang@sxmu.edu.cn.
- 6College of Medical Imaging, Shanxi Medical University, Taiyuan, 030001, China. yangguoqiang@sxmu.edu.cn.
- 7Shanxi Key Laboratory of Intelligent Imaging and Nanomedicine, First Hospital of Shanxi Medical University, Taiyuan, 030001, China. yangguoqiang@sxmu.edu.cn.
- 8Department of Radiology, First Hospital of Shanxi Medical University, Taiyuan, 030001, China. zhanghui_mr@163.com.
- 9College of Medical Imaging, Shanxi Medical University, Taiyuan, 030001, China. zhanghui_mr@163.com.
- 10Shanxi Key Laboratory of Intelligent Imaging and Nanomedicine, First Hospital of Shanxi Medical University, Taiyuan, 030001, China. zhanghui_mr@163.com.
- 11Intelligent Imaging Big Data and Functional Nano-imaging Engineering Research Center of Shanxi Province, First Hospital of Shanxi Medical University, Taiyuan, 030001, China. zhanghui_mr@163.com.
- 0Department of Radiology, First Hospital of Shanxi Medical University, Taiyuan, 030001, China.
|
March 29, 2025
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.This study developed an MRI-based habitat model to predict high-risk molecular subtypes in lower-grade gliomas (LrGGs). The model accurately identifies aggressive tumors and assesses survival prognosis, aiding in early diagnosis and risk-stratified management.
Area Of Science
- Neuro-oncology
- Medical Imaging
- Machine Learning
Background
- Lower-grade gliomas (LrGGs) include high-risk molecular subtypes with malignant transformation potential and poor prognosis.
- Early identification of these subtypes is crucial for effective clinical management and treatment strategies.
Purpose Of The Study
- To develop and validate a non-invasive MRI-based model for predicting high-risk molecular subtypes in LrGGs.
- To assess the prognostic value of the developed model for patient survival.
Main Methods
- Retrospective analysis of 345 LrGG patients with comprehensive molecular marker screening.
- Construction and evaluation of seven predictive models (habitat, radiomics, combined) using preoperative MRI sequences.
- Utilized Extra Trees classifier for habitat-based prediction and radiomics score for prognostic assessment.
- Kaplan-Meier analysis, log-rank test, concordance index, and calibration curves for model validation.
Main Results
- The Extra Trees habitat model demonstrated strong performance in predicting high-risk LrGG subtypes (AUCs 0.802-0.768 across datasets).
- The combined prognostic model showed the highest predictive accuracy (C-indices 0.781-0.743).
- Calibration curves confirmed the combined model's reliability in forecasting 1-3 year survival probabilities.
Conclusions
- An MRI-based habitat model effectively predicts high-risk LrGG molecular subtypes and patient survival prognosis non-invasively.
- This approach offers significant value for early detection of malignant transformation and personalized risk stratification in LrGGs.
Related Experiment Videos
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.