Related Experiment Video
Updated: Dec 3, 2025
07:37
Translational Orthotopic Models of Glioblastoma Multiforme
Published on: February 17, 2023
3.3K
Prediction of Glioma Grade Using Intratumoral and Peritumoral Radiomic Features From Multiparametric MRI Images
IEEE/ACM Transactions on Computational Biology and Bioinformatics
|October 26, 2020
Summary
This study introduces a new radiomics method using MRI scans to predict glioma grade noninvasively. Combining intratumoral and peritumoral features significantly improves accuracy, outperforming existing techniques.
Area of Science:
- Neuroimaging
- Radiomics
- Oncology
Background:
- Accurate glioma grading is crucial for treatment and prognosis.
- Current biopsy methods are invasive and costly.
- Noninvasive, accurate glioma grading methods are needed.
Purpose of the Study:
- To develop a novel radiomics pipeline for noninvasive glioma grading.
- To incorporate both intratumoral and peritumoral MRI features.
- To improve prediction accuracy compared to existing methods.
Main Methods:
- Utilized preoperative MRI scans from 285 patients.
- Extracted and refined 2153 radiomic features from intratumoral (ITV) and peritumoral (PTV) volumes.
- Developed radiomic signatures using LASSO and mRMR feature selection.
- Employed a novel algorithm to define the PTV.
- Validated performance using five-fold cross-validation and an external dataset.
Main Results:
- Radiomic signatures from ITV and PTV achieved high accuracy (AUC 0.968).
- Combined intratumoral and peritumoral (IPTV) signature reached an AUC of 0.975.
- The IPTV signature outperformed state-of-the-art methods.
- Demonstrated strong generalization performance on an external validation set.
Conclusions:
- The proposed radiomics pipeline accurately and noninvasively predicts glioma grade.
- Integrating intratumoral and peritumoral features enhances prediction performance.
- This method offers a promising alternative to invasive biopsy for glioma grading.
