A foundation for systematic analysis of transformers and RNNs for tractography

2026-10-01

AI Summary

This study addresses the inherent challenge in diffusion MRI fiber tractography of reconciling local diffusion information with global anatomical plausibility. We systematically evaluate the efficacy of RNNs and Transformers for iterative tractography, optimizing training strategies, input representations, and hyperparameter configurations. Furthermore, we introduce a novel "generation-verification" mechanism that enables streamline-level supervision, effectively resolving the mismatch between local losses and global tract quality. This framework is further enhanced by integrating CNN-based embeddings and end-of-sequence (EOS) tokens to improve modeling capacity. Experimental results demonstrate that the proposed models achieve state-of-the-art performance on the ISMRM 2015 tractography challenge dataset. Additionally, validation on in vivo data confirms the clinical applicability of our approach, highlighting its potential for advancing anatomically accurate white matter reconstruction.

Abstract

Machine learning (ML) has emerged as a promising approach for improving diffusion MRI (dMRI) tractography, a task that remains limited by the intrinsic tension between local diffusion information and global anatomical plausibility. In this work, we systematically evaluate recurrent neural networks (RNNs) and Transformer models for iterative tractography, with particular attention to training strategies, input representations (including convolutional neural network (CNN)-based embeddings and end-of-sequence (EOS) tokens), and hyperparameter selection. We introduce a generation-validation phase enabling supervision at the streamline level during training, allowing supervision despite the mismatch between local loss functions and global streamline quality. Using the ISMRM2015 tractography challenge dataset, our models achieve the highest reported performance to date. Through controlled experiments, we quantify the impact of missing bundles, noisy or imperfect training streamlines, and invalid fibers in the training set. Finally, we demonstrate the applicability of our best-performing models for in vivo data from the Tractoinferno database. Overall, our results highlight both the potential and the limits of sequence-based deep learning models such as Transformers and RNNs for tractography, and emphasize the need for improved phantoms and evaluation methods for in vivo validation. We provide takeaways and recommendations for future researchers training and validating sequence-based supervised methods for tractography.

Problem

• tractography
• diffusion MRI
• transformers
• recurrent neural networks
• streamline quality

Innovation

• tractography
• Transformers
• RNNs
• generation-validation phase
• diffusion MRI