DDTracking: A Deep Generative Framework for Diffusion MRI Tractography with Streamline Local-Global Spatiotemporal Modeling

2025-08-06

AI Summary

To address the challenge of jointly modeling local structural details and global temporal dependencies in diffusion MRI (dMRI) tractography, this paper proposes DDTracking—a conditional denoising diffusion-based end-to-end deep generative framework. Methodologically, it introduces (1) a dual-path encoder network that separately captures high-resolution spatial features and long-range temporal dynamics, and (2) a tract propagation process formulated as an anatomy-constrained conditional diffusion process, enabling robust and interpretable whole-brain fiber reconstruction. Evaluated on the ISMRM and TractoInferno benchmarks, DDTracking significantly outperforms state-of-the-art methods in accuracy and reproducibility. Moreover, it demonstrates exceptional generalizability across diverse populations, imaging devices, and acquisition protocols. These advances establish a new paradigm for clinically deployable, physics-informed, and data-driven dMRI tractography.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Machine Learning: Distributed Machine Learning & Federated Learning

Application Category

• Web Mining and Content Analysis: Content-based information diffusion
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

This paper presents DDTracking, a novel deep generative framework for diffusion MRI tractography that formulates streamline propagation as a conditional denoising diffusion process. In DDTracking, we introduce a dual-pathway encoding network that jointly models local spatial encoding (capturing fine-scale structural details at each streamline point) and global temporal dependencies (ensuring long-range consistency across the entire streamline). Furthermore, we design a conditional diffusion model module, which leverages the learned local and global embeddings to predict streamline propagation orientations for tractography in an end-to-end trainable manner. We conduct a comprehensive evaluation across diverse, independently acquired dMRI datasets, including both synthetic and clinical data. Experiments on two well-established benchmarks with ground truth (ISMRM Challenge and TractoInferno) demonstrate that DDTracking largely outperforms current state-of-the-art tractography methods. Furthermore, our results highlight DDTracking's strong generalizability across heterogeneous datasets, spanning varying health conditions, age groups, imaging protocols, and scanner types. Collectively, DDTracking offers anatomically plausible and robust tractography, presenting a scalable, adaptable, and end-to-end learnable solution for broad dMRI applications. Code is available at: https://github.com/yishengpoxiao/DDtracking.git

Problem

• Models streamline propagation as conditional denoising diffusion process
• Integrates local spatial encoding and global temporal dependencies
• Improves tractography accuracy and generalizability across diverse datasets

Innovation

• Deep generative framework for diffusion MRI tractography
• Dual-pathway encoding network for local-global modeling
• Conditional diffusion model for end-to-end streamline prediction