OpenMAP-BrainAge: Generalizable and Interpretable Brain Age Predictor

2025-06-21

AI Summary

Poor generalizability and limited interpretability hinder brain-age prediction from MRI across diverse populations and scanner platforms. To address this, we propose a linear-complexity stem-Transformer architecture and introduce a novel multimodal representation method that jointly encodes pseudo-3D volumetric MRI (from axial, sagittal, and coronal views) and volumetric embeddings of multiple anatomical brain regions. Our framework integrates self-supervised pretraining, gradient-based attribution visualization, and cross-domain feature alignment to enhance model robustness and clinical interpretability. Evaluated on the ADNI+OASIS test set, our model achieves a mean absolute error (MAE) of 3.65 years; on the independent AIBL cohort, MAE is 3.54 years. Brain-age gap (BAG) significantly differentiates cognitively normal (CN), mild cognitive impairment (MCI), and Alzheimer’s disease (AD) groups (p < 0.001) and exhibits strong negative correlations with MoCA (r = −0.62) and MMSE (r = −0.58) scores, underscoring its potential as a clinically meaningful biomarker.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Purpose: To develop an age prediction model which is interpretable and robust to demographic and technological variances in brain MRI scans. Materials and Methods: We propose a transformer-based architecture that leverages self-supervised pre-training on large-scale datasets. Our model processes pseudo-3D T1-weighted MRI scans from three anatomical views and incorporates brain volumetric information. By introducing a stem architecture, we reduce the conventional quadratic complexity of transformer models to linear complexity, enabling scalability for high-dimensional MRI data. We trained our model on ADNI2 $&$ 3 (N=1348) and OASIS3 (N=716) datasets (age range: 42 - 95) from the North America, with an 8:1:1 split for train, validation and test. Then, we validated it on the AIBL dataset (N=768, age range: 60 - 92) from Australia. Results: We achieved an MAE of 3.65 years on ADNI2 $&$ 3 and OASIS3 test set and a high generalizability of MAE of 3.54 years on AIBL. There was a notable increase in brain age gap (BAG) across cognitive groups, with mean of 0.15 years (95% CI: [-0.22, 0.51]) in CN, 2.55 years ([2.40, 2.70]) in MCI, 6.12 years ([5.82, 6.43]) in AD. Additionally, significant negative correlation between BAG and cognitive scores was observed, with correlation coefficient of -0.185 (p < 0.001) for MoCA and -0.231 (p < 0.001) for MMSE. Gradient-based feature attribution highlighted ventricles and white matter structures as key regions influenced by brain aging. Conclusion: Our model effectively fused information from different views and volumetric information to achieve state-of-the-art brain age prediction accuracy, improved generalizability and interpretability with association to neurodegenerative disorders.

Problem

• Develop interpretable brain age prediction model
• Handle demographic and MRI scan variances
• Improve accuracy and generalizability for neurodegenerative disorders

Innovation

• Transformer-based architecture for MRI scans
• Linear complexity stem for scalability
• Multi-view and volumetric data fusion