Related Experiment Video
Updated: Jul 21, 2025
Translational Orthotopic Models of Glioblastoma Multiforme
Published on: February 17, 2023
Development and External Validation of an MRI-based Radiomics Nomogram to Distinguish Circumscribed Astrocytic
Shuang Li1, Xiaorui Su2, Juan Peng3
1Huaxi MR Research Center (HMRRC), Department of Radiology, West China Hospital of Sichuan University, Chengdu, Sichuan, China (S.L., X.S., S.Z., H.S., Q.T., Q.G.); Research Unit of Psychoradiology, Chinese Academy of Medical Sciences, Chengdu, Sichuan, China (S.L.); Functional and Molecular Imaging Key Laboratory of Sichuan Province, West China Hospital of Sichuan University, Chengdu, Sichuan, China (S.L.).
New models can distinguish circumscribed astrocytic gliomas (CAGs) from diffuse gliomas (DGs) using radiomics. The interface radiomics model shows high accuracy, aiding pre-surgical evaluation and improving glioma patient management.
Area of Science:
- Neuro-oncology
- Medical Imaging
- Radiomics
Background:
- The 5th edition of the World Health Organization (WHO) classification of Central Nervous System (CNS) tumors introduced "diffuse" and "circumscribed" gliomas.
- Distinguishing between these glioma types is crucial for accurate diagnosis and treatment planning.
Purpose of the Study:
- To develop and validate radiomics models for differentiating circumscribed astrocytic gliomas (CAGs) from diffuse gliomas (DGs).
- To evaluate the performance of models based on different volumes of interest (VOIs) and combined radiomics and clinical features.
Main Methods:
- Retrospective analysis of MRI data from 475 patients with CAGs and DGs across three institutions.
- Tumor segmentation to define VOIs: tumor and peritumor, whole tumor, and interface.
- Development of radiomics, clinical, and combined models, using Synthetic Minority Oversampling Technique for dataset balancing.
Main Results:
- The VOI interface radiomics model achieved an AUC of 0.806 in cross-validation and 0.897 in external validation.
- A combined model incorporating interface radiomics and clinical features demonstrated superior performance with an AUC of 0.94 in external validation.
- Seven features selected via ANOVA and SVM were key to the model's predictive power.
Conclusions:
- Radiomics models focusing on the peritumoral area are effective in distinguishing CAGs from DGs.
- These models show potential for pre-surgical assessment of tumor nature and improved clinical management of glioma patients.
