Related Experiment Video
Updated: Apr 23, 2026
10:25
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
Published on: September 25, 2019
48.8K
Resting state fMRI feature-based cerebral glioma grading by support vector machine
Jiangfen Wu1, Zhiyu Qian, Ling Tao
1Department of Biomedical Engineering, College of Automation, Nanjing University of Aeronautics and Astronautics, No. 29, Yudao St., Qinhuai District, Nanjing, 210016, Jiangsu Province, China, wjfyunzhu@163.com.
International Journal of Computer Assisted Radiology and Surgery
|September 18, 2014
Summary
Resting-state functional MRI (RS-fMRI) parameters effectively differentiate glioma grades. Signal intensity correlation (SIC), fractional amplitude of low-frequency fluctuation (fALFF), and regional homogeneity (ReHo) show promise for noninvasive tumor grading.
Area of Science:
- Neuroimaging
- Oncology
- Medical Physics
Background:
- Accurate glioma grading is crucial for treatment selection.
- Noninvasive methods for tumor grading are highly desirable in clinical settings.
- Conventional MRI has limitations in fully characterizing tumor heterogeneity.
Purpose of the Study:
- To extract and evaluate RS-fMRI parameters for glioma grading.
- To assess the diagnostic performance of these parameters in differentiating low-grade glioma (LGG) from high-grade glioma (HGG).
- To explore the potential of RS-fMRI as a complementary diagnostic tool.
Main Methods:
- Tumor segmentation using conventional MRI and RS-fMRI.
- Analysis of four RS-fMRI parameters: signal intensity difference ratio, signal intensity correlation (SIC), fractional amplitude of low-frequency fluctuation (fALFF), and regional homogeneity (ReHo).
- Statistical comparison (Mann-Whitney test) and classification (Support Vector Machine - SVM) of parameters between LGG and HGG groups.
Main Results:
- High-grade glioma (HGG) exhibited more complex anatomical and RS-fMRI features compared to low-grade glioma (LGG).
- SIC, fALFF, and ReHo were significant features for classification.
- SVM classification achieved accuracies, sensitivities, and specificities exceeding 80%, with SIC showing the highest accuracy (89%).
Conclusions:
- RS-fMRI parameters demonstrate efficacy in classifying glioma tumor grades.
- The findings suggest RS-fMRI has clinical potential as an adjunct for glioma diagnosis.
- This noninvasive technique can aid in optimizing patient treatment strategies.
