Anatomically Guided Latent Diffusion for Brain MRI Progression Modeling

2026-01-21

AI Summary

Existing methods for modeling longitudinal progression in brain MRI are often limited by architectural complexity, insufficient integration of clinical covariates, and challenges in preserving anatomical consistency. This work proposes the AG-LDM framework, which enables end-to-end fusion of baseline anatomy, noisy follow-up images, and clinical covariates at the input level, while incorporating a lightweight WarpSeg segmentation model to provide anatomical supervision—eliminating the need for auxiliary control networks and enabling unified training. The approach substantially enhances sensitivity to temporal and clinical factors (up to 31.5-fold), achieves state-of-the-art image quality on the ADNI dataset, reduces volumetric errors by 15–20%, and generates plausible counterfactual progression trajectories consistent with Alzheimer’s disease pathology in zero-shot evaluation on OASIS-3.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multimodal Learning
• Multiagent Systems: Adversarial Agents

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Accurately modeling longitudinal brain MRI progression is crucial for understanding neurodegenerative diseases and predicting individualized structural changes. Existing state-of-the-art approaches, such as Brain Latent Progression (BrLP), often use multi-stage training pipelines with auxiliary conditioning modules but suffer from architectural complexity, suboptimal use of conditional clinical covariates, and limited guarantees of anatomical consistency. We propose Anatomically Guided Latent Diffusion Model (AG-LDM), a segmentation-guided framework that enforces anatomically consistent progression while substantially simplifying the training pipeline. AG-LDM conditions latent diffusion by directly fusing baseline anatomy, noisy follow-up states, and clinical covariates at the input level, a strategy that avoids auxiliary control networks by learning a unified, end-to-end model that represents both anatomy and progression. A lightweight 3D tissue segmentation model (WarpSeg) provides explicit anatomical supervision during both autoencoder fine-tuning and diffusion model training, ensuring consistent brain tissue boundaries and morphometric fidelity. Experiments on 31,713 ADNI longitudinal pairs and zero-shot evaluation on OASIS-3 demonstrate that AG-LDM matches or surpasses more complex diffusion models, achieving state-of-the-art image quality and 15-20\% reduction in volumetric errors in generated images. AG-LDM also exhibits markedly stronger utilization of temporal and clinical covariates (up to 31.5x higher sensitivity than BrLP) and generates biologically plausible counterfactual trajectories, accurately capturing hallmarks of Alzheimer's progression such as limbic atrophy and ventricular expansion. These results highlight AG-LDM as an efficient, anatomically grounded framework for reliable brain MRI progression modeling.

Problem

• brain MRI progression
• anatomical consistency
• clinical covariates
• longitudinal modeling
• neurodegenerative diseases

Innovation

• anatomically guided diffusion
• latent diffusion model
• brain MRI progression
• segmentation supervision
• clinical covariate conditioning