Related Experiment Video
Updated: Sep 29, 2025
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
Parameters from site classification to harmonize MRI clinical studies: Application to a multi-site Parkinson's
Gemma C Monte-Rubio1,2, Barbara Segura1,2,3,4, Antonio P Strafella5,6,7
1Institute of Neurosciences, University of Barcelona, Barcelona, Catalonia, Spain.
Abstract:
Multi-site MRI datasets are crucial for big data research. However, neuroimaging studies must face the batch effect. Here, we propose an approach that uses the predictive probabilities provided by Gaussian processes (GPs) to harmonize clinical-based studies. A multi-site dataset of 216 Parkinson's disease (PD) patients and 87 healthy subjects (HS) was used. We performed a site GP classification using MRI data. The outcomes estimated from this classification, redefined like Weighted HARMonization PArameters (WHARMPA), were used as regressors in two different clinical studies: A PD versus HS machine learning classification using GP, and a VBM comparison (FWE-p < .05, k = 100). Same studies were also conducted using conventional Boolean site covariates, and without information about site belonging. The results from site GP classification provided high scores, balanced accuracy (BAC) was 98.39% for grey matter images. PD versus HS classification performed better when the WHARMPA were used to harmonize (BAC = 78.60%; AUC = 0.90) than when using the Boolean site information (BAC = 56.31%; AUC = 0.71) and without it (BAC = 57.22%; AUC = 0.73). The VBM analysis harmonized using WHARMPA provided larger and more statistically robust clusters in regions previously reported in PD than when the Boolean site covariates or no corrections were added to the model. In conclusion, WHARMPA might encode global site-effects quantitatively and allow the harmonization of data. This method is user-friendly and provides a powerful solution, without complex implementations, to clean the analyses by removing variability associated with the differences between sites.
Insights
We developed a new method using Gaussian processes (GPs) to harmonize multi-site neuroimaging data, reducing batch effects. This Weighted HARMonization PArameters (WHARMPA) approach improved Parkinson's disease classification and VBM analysis accuracy.
Area of Science:
- Neuroimaging
- Medical Data Analysis
- Machine Learning
Background:
- Multi-site MRI datasets are essential for large-scale neuroimaging research.
- Batch effects arising from different sites pose a significant challenge in data harmonization.
- Existing methods for correcting site effects can be suboptimal.
Purpose of the Study:
- To introduce a novel method for harmonizing multi-site MRI data using Gaussian processes (GPs).
- To evaluate the effectiveness of the proposed Weighted HARMonization PArameters (WHARMPA) in mitigating site-specific batch effects.
- To compare the performance of WHARMPA against conventional site correction methods.
Main Methods:
- Gaussian processes (GPs) were employed for site classification on multi-site MRI data.
- Predictive probabilities from GPs were transformed into Weighted HARMonization PArameters (WHARMPA).
- WHARMPA were used as regressors in machine learning classification (Parkinson's disease vs. healthy subjects) and VBM analysis, compared to Boolean site covariates and no correction.
Main Results:
- Site GP classification achieved high accuracy (98.39% balanced accuracy for grey matter images).
- Parkinson's disease classification showed improved performance with WHARMPA harmonization (78.60% BAC, 0.90 AUC) compared to other methods.
- VBM analysis using WHARMPA yielded larger, more robust clusters in PD-related regions.
Conclusions:
- WHARMPA effectively quantifies and corrects global site effects in neuroimaging data.
- The proposed method offers a user-friendly and powerful solution for data harmonization without complex implementation.
- WHARMPA enhances the reliability and statistical power of analyses in multi-site clinical studies.
More Related Videos
09:06Whole-brain Segmentation and Change-point Analysis of Anatomical Brain MRI—Application in Premanifest Huntington's Disease
Published on: June 9, 2018
08:43Application of Granger Causality Analysis of the Directed Functional Connection in Alzheimer's Disease and Mild Cognitive Impairment
Published on: August 7, 2017