Removal of site effects and enhancement of signal using dual projection independent component analysis for pooling multi-site MRI data.

Researchers developed a better way to combine brain scans from different hospitals and research centres. When studying conditions like autism, combining data from multiple sites creates larger, more reliable datasets. However, different scanners can create confusing differences in the data. This new method better removes these technical differences while keeping the important brain information intact, making autism research more accurate.

This methodological study developed a dual projection independent component analysis (DP-ICA) harmonisation method to improve the pooling of multi-site MRI datasets in neuroscience research. The challenge addressed is that scanner/site variability can confound results when combining neuroimaging data across different research sites. The DP-ICA method aims to more completely remove site effects while preserving signals of interest, even when these are correlated with site variables. Testing on simulated data, autism brain imaging datasets from ABIDE II, and travelling subject data demonstrated superior performance compared to conventional harmonisation methods in removing site effects and enhancing detection sensitivity.

This harmonisation method could improve the reliability of multi-site autism neuroimaging research, potentially leading to more robust findings about brain differences in autism and better reproducibility of research results across different centres.

This is primarily a methodological validation study. The clinical impact and generalisability across different autism populations and neuroimaging protocols requires further validation. Sample sizes for testing datasets were not reported.

Combining magnetic resonance imaging (MRI) data from multi-site studies is a popular approach for constructing larger datasets to greatly enhance the reliability and reproducibility of neuroscience research. However, the scanner/site variability is a significant confound that complicates the interpretation of the results, so effective and complete removal of the scanner/site variability is necessary to realise the full advantages of pooling multi-site datasets. Independent component analysis (ICA) and general linear model (GLM) based harmonisation methods are the two primary methods used to eliminate scanner/site effects. Unfortunately, there are challenges with both ICA-based and GLM-based harmonisation methods to remove site effects completely when the signals of interest and scanner/site effects-related variables are correlated, which may occur in neuroscience studies.

In this study, we propose an effective and powerful harmonisation strategy that implements dual projection (DP) theory based on ICA to remove the scanner/site effects more completely. This method can separate the signal effects correlated with site variables from the identified site effects for removal without losing signals of interest. Both simulations and vivo structural MRI datasets, including a dataset from Autism Brain Imaging Data Exchange II and a travelling subject dataset from the Strategic Research Program for Brain Sciences, were used to test the performance of a DP-based ICA harmonisation method. Results show that DP-based ICA harmonisation has superior performance for removing site effects and enhancing the sensitivity to detect signals of interest as compared with GLM-based and conventional ICA harmonisation methods.