Contributions of Site Heterogeneity in Multi-site Brain-Phenotype Predictions

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Machine learning studies of brain–phenotype associations typically summarize accuracy with a single Pearson correlation, yet in harmonized multi-site data this metric is shaped by training-set size, feature and outcome variability, and true effect heterogeneity. We analyzed structural MRI data from the Reproducible Brain Charts consortium. After preprocessing and quality control, 3486 participants from 9 sites were retained. Three training set compositions were used to understand contributions of site heterogeneity after harmonization: within-site (Local), size-downsampled multi-site (Multi-Down), and full multi-site (Multi-Full). Another site-agnostic strategy (Overall) pooled all sites into a single dataset. All strategies included the target site in the training and testing samples. These analyses separate the contribution of the site effect from that of sample size. Between-site heterogeneity was extensive (age I ² ≈ 95%). The overall correlation of age prediction ( r = 0.73) exceeded the within-site meta-analytic estimate (Multi-Full r = 0.63). The improvement of correlation in the pooled dataset was driven by a mechanical inflation from the larger outcome variance. At matched sample size, within-site training outperformed harmonized multi-site training. Increasing the multi-site training sample size did not yield a significant advantage over within-site training. The apparent advantage of cross-dataset training thus appears to reflect training-set size rather than a benefit of multi-site composition per se. The same pattern generally held for the weaker brain-p-factor signals. Together, these results indicate that between-site heterogeneity in machine learning brain– phenotype prediction persists after harmonization. We recommend reporting site-level prediction accuracy and feature dispersion and standard deviation of the outcome variable alongside any pooled estimate to improve inter-site and inter-study comparison of machine learning prediction accuracy.

Key points

  • Between-site heterogeneity in machine-learning brain–phenotype prediction persists after harmonization: across nine CovBat-harmonized sites ( N = 3,486), within-site models outperformed multi-site models at matched training size, and enlarging the multi-site training set did not consistently surpass within-site training.

  • A single pooled correlation masks this heterogeneity: pooling sites overstated within-site accuracy for age ( r = 0.73 vs. 0.63), largely because pooling widens the outcome’s range.

  • Site-level accuracy tracked imaging-feature variance and outcome variability; we recommend reporting site-level accuracy, feature variance, and outcome standard deviation alongside any pooled estimate.

  • Article activity feed