Longitudinal Lesion Inpainting in Brain MRI via 3D Region Aware Diffusion

2026-03-05

AI Summary

This study addresses the challenge posed by evolving lesions in longitudinal brain MRI analysis, which often disrupt automated processing pipelines. To this end, the authors propose a pseudo-3D longitudinal lesion inpainting framework based on Denoising Diffusion Probabilistic Models (DDPM), introducing for the first time a Region-Aware Diffusion (RAD) mechanism. By leveraging multi-temporal, multi-channel conditional inputs, the method accurately reconstructs pathological regions while preserving healthy tissue. The approach effectively balances 3D anatomical continuity, longitudinal consistency, and computational efficiency. Evaluated on data from 93 patients, it significantly outperforms existing methods, reducing the LPIPS distance from 0.07 to 0.03, achieving a temporal fidelity index of 1.024, and requiring only 2.53 minutes per case—approximately a tenfold acceleration over current approaches.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Accurate longitudinal analysis of brain MRI is often hindered by evolving lesions, which bias automated neuroimaging pipelines. While deep generative models have shown promise in inpainting these lesions, most existing methods operate cross-sectionally or lack 3D anatomical continuity. We present a novel pseudo-3D longitudinal inpainting framework based on Denoising Diffusion Probabilistic Models (DDPM). Our approach utilizes multi-channel conditioning to incorporate longitudinal context from distinct visits (t_1, t_2) and extends Region-Aware Diffusion (RAD) to the medical domain, focusing the generative process on pathological regions without altering surrounding healthy tissue. We evaluated our model against state-of-the-art baselines on longitudinal brain MRI from 93 patients. Our model significantly outperforms the leading baseline (FastSurfer-LIT) in terms of perceptual fidelity, reducing the Learned Perceptual Image Patch Similarity (LPIPS) distance from 0.07 to 0.03 while effectively eliminating inter-slice discontinuities. Furthermore, our model demonstrates high longitudinal stability with a Temporal Fidelity Index of 1.024, closely approaching the ideal value of 1.0 and substantially narrowing the gap compared to LIT's TFI of 1.22. Notably, the RAD mechanism provides a substantial gain in efficiency; our framework achieves an average processing time of 2.53 min per volume, representing approximately 10x speedup over the 24.30 min required by LIT. By leveraging longitudinal priors and region-specific denoising, our framework provides a highly reliable and efficient preprocessing step for the study of progressive neurodegenerative diseases. A derivative dataset consisting of 93 pre-processed scans used for testing will be available upon request after acceptance. Code will be released upon acceptance.

Problem

• longitudinal lesion inpainting
• brain MRI
• neurodegenerative diseases
• 3D anatomical continuity
• automated neuroimaging pipelines

Innovation

• Longitudinal Inpainting
• 3D Region-Aware Diffusion
• Denoising Diffusion Probabilistic Models
• Brain MRI
• Temporal Fidelity