Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 12, 2026

A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia
09:59

A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia

Published on: September 16, 2017

Accelerating Stroke MRI With Diffusion Probabilistic Models Through Large-Scale Pre-Training and Target-Specific

Yamin Arefeen1,2, Sidharth Kumar1, Steven Warach3

  • 1Chandra Family Department of Electrical and Computer Engineering, The University of Texas at Austin, Austin, Texas, USA.

Magnetic Resonance in Medicine
|July 9, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Full robotic endovascular treatment for various head and neck hemorrhagic lesions.

Interventional neuroradiology : journal of peritherapeutic neuroradiology, surgical procedures and related neurosciences·2026
Same author

Transition to tenecteplase is associated with shorter door-to-puncture times: a retrospective study from the Lone Star Stroke consortium TNK registry.

Frontiers in neurology·2026
Same author

Adenoviral Vectors in Gene Therapy: A Detailed Overview.

Iranian biomedical journal·2026
Same author

Volume Encoding Gaussians: Transfer Function-Agnostic 3D Gaussians for Volume Rendering.

IEEE transactions on visualization and computer graphics·2026
Same author

Association Between MR-Based Thrombus Imaging Characteristics and Endovascular Therapy Outcome in Acute Ischemic Stroke: A Systematic Review and Meta-Analysis.

Stroke (Hoboken, N.J.)·2026
Same author

Mechanical Thrombectomy for Acute Ischemic Stroke in Patients With Dementia.

Stroke (Hoboken, N.J.)·2026

Accelerated MRI reconstruction using Diffusion Probabilistic Generative Models (DPMs) is now feasible with limited data. Pre-training DPMs on diverse datasets and fine-tuning on small target datasets enables faster stroke MRI scans with clinically acceptable quality.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Diffusion Models

Background:

  • Accelerated MRI is crucial for clinical applications like stroke detection.
  • Limited fully-sampled data hinders the training of deep learning models for MRI reconstruction.
  • Diffusion Probabilistic Generative Models (DPMs) show promise for image generation and reconstruction.

Purpose of the Study:

  • To develop a data-efficient strategy for accelerated MRI reconstruction using DPMs.
  • To enable faster scan times in clinical stroke MRI with limited available data.
  • To evaluate the performance of DPMs in data-constrained accelerated MRI scenarios.

Main Methods:

  • A simple training strategy involving pre-training a DPM on a large public dataset (fastMRI).
Keywords:
accelerated MRIclinical translationdiffusion probabilistic modelspre‐training and fine‐tuningstroke MRI

More Related Videos

Advanced Diffusion Imaging in The Hippocampus of Rats with Mild Traumatic Brain Injury
10:33

Advanced Diffusion Imaging in The Hippocampus of Rats with Mild Traumatic Brain Injury

Published on: August 14, 2019

Related Experiment Videos

Last Updated: Jul 12, 2026

A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia
09:59

A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia

Published on: September 16, 2017

Advanced Diffusion Imaging in The Hippocampus of Rats with Mild Traumatic Brain Injury
10:33

Advanced Diffusion Imaging in The Hippocampus of Rats with Mild Traumatic Brain Injury

Published on: August 14, 2019

  • Fine-tuning the pre-trained DPM on a small, target clinical stroke MRI dataset.
  • Careful selection of learning rates and fine-tuning durations during the fine-tuning process.
  • Evaluation using controlled experiments and a blinded clinical reader study on stroke MRI data.

    • Main Results:

    • Pre-trained and fine-tuned DPMs achieved reconstruction performance comparable to models trained with substantially more target data.
    • Moderate fine-tuning with reduced learning rates improved performance, while insufficient or excessive fine-tuning degraded quality.
    • Images reconstructed from 2x accelerated data were rated comparably to standard-of-care in a blinded reader study.

    Conclusions:

    • Large-scale pre-training combined with targeted fine-tuning enables DPM-based MRI reconstruction for data-constrained applications.
    • The proposed approach reduces the need for large, application-specific datasets in accelerated MRI.
    • Pre-trained and fine-tuned diffusion models offer a viable strategy for accelerated MRI in targeted clinical applications.