Beyond Feature Reliability: Repeat-Informed Multifractal Curve Regression for Brain-Age Prediction

2026-09-24

AI Summary

This study addresses the low test-retest reliability in brain age prediction caused by joint fluctuations of multiple features. To overcome the limitations of single-feature reliability, we propose the RMCR framework, which for the first time performs structured regression modeling on the joint consistency of multi-order fluctuations. Specifically, multifractal analysis is employed to extract curve structural features, and constrained optimization based on test-retest information is introduced to enable joint learning. Experimental evaluations on the HCP-A and Cam-CAN datasets demonstrate that the proposed framework significantly reduces the mean absolute error of predictions while effectively enhancing the consistency of age estimation from single scans.

Abstract

Brain-age prediction from resting-state fMRI provides a quantitative framework for characterizing age-related changes in spontaneous brain dynamics and for identifying functional signatures. Existing studies have linked fractal and multifractal scaling to age and examined the reliability of individual features. However, prediction repeatability depends on how features fluctuate jointly and how a predictor combines them, which feature-wise reliability assessments do not capture. To address this problem, we propose Repeat-informed Multifractal Curve Regression (RMCR), a structured framework for learning stable age-predictive patterns from multifractal curves. By jointly modeling curve structure and repeat-scan variability, RMCR learns predictive combinations of fluctuation orders that target both accuracy and within-subject consistency. Relative to a matched run-level ridge baseline, RMCR reduces single-run MAE by 6.1% on HCP-A and 7.9% on an external Cam-CAN cohort, and within-visit repeat absolute difference by 18.5% on HCP-A, using a single scan at inference.

Problem

• brain-age prediction
• multifractal analysis
• prediction repeatability
• resting-state fMRI

Innovation

• Multifractal Curve Regression
• Brain-Age Prediction
• Prediction Repeatability
• Resting-State fMRI
• Within-Subject Consistency