Comparison of Multispectral Image-Processing Methods for Brain Tissue Classification in BrainWeb Synthetic Data and
Hsian-Min Chen1,2,3, Hung-Chieh Chen4,5, Clayton Chi-Chang Chen4
1Center for QUantitative Imaging in Medicine (CQUIM), Department of Medical Research, Taichung Veterans General Hospital, Taichung, Taiwan.
Biomed Research International
|March 22, 2021
Summary
A new algorithm, TRIO (TRIOA), accurately quantifies brain tissue volumes, outperforming existing methods like SPM and FAST, especially for cerebrospinal fluid (CSF) estimation in MRI scans.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Biomedical Engineering
Background:
- Accurate brain tissue quantification is crucial in neuroimaging.
- Existing methods like SPM and FAST struggle with reliable cerebrospinal fluid (CSF) volume estimation.
- Magnetic resonance (MR) multispectral classification is a key technique.
Purpose of the Study:
- To develop and evaluate a novel algorithm, TRIO (TRIOA), for accurate MR brain tissue classification and volume measurement.
- To compare the performance of TRIOA against established methods (SPM, FAST) for gray matter (GM), white matter (WM), and CSF volume estimation.
- To address the limitations of current methods in reliably quantifying CSF volumes.
Main Methods:
- Development of the TRIO algorithm (TRIOA), a new MR multispectral classification method.
- Evaluation of TRIOA, SPM8, SPM12, and FAST using the BrainWeb database and real MRI data from 140 healthy volunteers.
- Preprocessing steps included skull stripping, motion, and inhomogeneity correction.
Main Results:
- TRIOA demonstrated superior effectiveness compared to SPM and FAST in extensive experimentation.
- All tested methods showed age-associated declines in GM and WM volume fractions in participants aged 20-83.
- TRIOA provided more consistent and accurate estimations for GM, WM, and particularly CSF volumes compared to SPM and FAST.
Conclusions:
- The proposed TRIOA algorithm offers significant advantages over SPM and FAST for MR brain tissue classification.
- TRIOA provides more accurate volume measurements, especially for CSF, improving upon existing neuroimaging analysis tools.
- This advancement enhances the reliability of quantitative neuroimaging studies.
