Related Experiment Video
Updated: Feb 18, 2026
09:33
Diffusion Tensor Magnetic Resonance Imaging in the Analysis of Neurodegenerative Diseases
Published on: July 28, 2013
29.4K
Integration of routine QA data into mega-analysis may improve quality and sensitivity of multisite diffusion tensor
Peter Kochunov1, Erin W Dickie2, Joseph D Viviano2
1Maryland Psychiatric Research Center, Department of Psychiatry, University of Maryland School of Medicine, Baltimore.
Human Brain Mapping
|November 29, 2017
Summary
This study introduces a new mega-analytical method to reduce variance in diffusion tensor imaging (DTI) data for schizophrenia research. The approach improved the accuracy of white matter integrity findings in schizophrenia patients compared to controls.
Area of Science:
- Neuroimaging
- Psychiatric Neuroscience
- Biostatistics
Background:
- Diffusion tensor imaging (DTI) is crucial for assessing white matter integrity.
- Schizophrenia is associated with alterations in white matter structure.
- Methodological variations across sites can confound multi-site neuroimaging studies.
Purpose of the Study:
- To evaluate a novel mega-analytical approach for reducing methodological variance in multi-site DTI fractional anisotropy (FA) data.
- To compare white matter integrity in individuals with schizophrenia versus controls using this enhanced approach.
- To improve the reliability and comparability of DTI findings across different research sites.
Main Methods:
- Utilized a dataset of 192 individuals (119 patients, 73 controls) from three MRI sites.
- Implemented quality assurance (QA) factor analysis to identify and regress site-specific methodological variances (SNR, FA).
- Applied Marchenko-Pastur Principal Component Analysis (MP-PCA) for data denoising and evaluated its impact.
Main Results:
- Regression of QA factors improved the effect size of schizophrenia on whole-brain average FA (Cohen's d from .53 to .57).
- Agreement between regional FA differences in this study and the ENIGMA consortium improved from r=.54 to .70 after QA regression.
- MP-PCA denoising further enhanced this agreement to r=.81, demonstrating its effectiveness in reducing noise.
Conclusions:
- The novel mega-analytical approach effectively reduces methodological variance in multi-site DTI studies.
- Accounting for QA variance and employing MP-PCA denoising significantly enhances the agreement with large-scale meta-analyses.
- This method offers a more robust framework for investigating white matter integrity in schizophrenia and other neurological disorders.
