JET-Diff: Joint-Encoding Tensor Diffusion Model for Accurate DTI Reconstruction from Sparse DWIs

ICLR 2026 Conference Submission16733 Authors

Keywords: Diffusion Tensor Imaging, Deep Learning, Generative Models, Latent Diffusion Models, Image Reconstruction, Neuroimaging

TL;DR: To overcome the long scan times of conventional DTI, we propose JET-Diff, a latent diffusion model that generates accurate and physically plausible 3D diffusion tensors from a minimal number of DWI scans.

Abstract: Diffusion Tensor Imaging (DTI) is an advanced Magnetic resonance imaging (MRI) technique for characterizing white matter microstructure. Conventional DTI protocols require multiple diffusion-weighted imaging (DWI) acquisitions across numerous directions, resulting in long scan times, motion artifacts, patient discomfort, and reduced clinical utility. Current deep learning approaches frequently yield diffusion tensors that are anatomically inconsistent or physically implausible. We introduce Joint-Encoding Tensor Diffusion (JET-Diff), a framework that synthesizes the full six-component diffusion tensor in 3D. Specifically, we propose a Multi-Tensor Latent Diffusion (MTLD) model that learns a shared latent distribution between DWIs and DTIs, enforcing both anatomical fidelity and physical plausibility. MTLD leverages a novel anatomical autoencoder to disentangle structural information from tensor properties, yielding a compact and expressive latent space optimized for generative performance. Experiments conducted on the Human Connectome Project (HCP) dataset demonstrate that JET-Diff significantly improves reconstruction accuracy and generates diffusion tensors that support more reliable downstream tractography.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 16733