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 Concept Videos

Propagation of Uncertainty from Random Error00:59

Propagation of Uncertainty from Random Error

An experiment often consists of more than a single step. In this case, measurements at each step give rise to uncertainty. Because the measurements occur in successive steps, the uncertainty in one step necessarily contributes to that in the subsequent step. As we perform statistical analysis on these types of experiments, we must learn to account for the propagation of uncertainty from one step to the next. The propagation of uncertainty depends on the type of arithmetic operation performed on...

You might also read

Related Articles

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

Sort by
Same author

Multi-cancer framework with cancer-aware attention and adversarial mutual-information minimization for whole slide image classification.

Medical image analysis·2026
Same author

A self-attention model for robust rigid slice-to-volume registration of functional MRI.

Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society·2025
Same author

Radiomics-based Machine Learning Prediction of Neoadjuvant Chemotherapy Response in Breast Cancer Using Physiologically Decomposed Diffusion-weighted MRI.

Radiology. Imaging cancer·2025
Same author

Improved MRI detection of inflammation-induced changes in bone marrow microstructure in mice: a machine learning-enhanced T2 distribution analysis.

European radiology experimental·2025
Same author

MBSS-T1: Model-based subject-specific self-supervised motion correction for robust cardiac T1 mapping.

Medical image analysis·2025
Same author

Non-parametric Bayesian deep learning approach for whole-body low-dose PET reconstruction and uncertainty assessment.

Medical & biological engineering & computing·2025

Related Experiment Video

Updated: Jul 14, 2026

Topographical Estimation of Visual Population Receptive Fields by fMRI
06:02

Topographical Estimation of Visual Population Receptive Fields by fMRI

Published on: February 3, 2015

9.3K

NPB-REC: A non-parametric Bayesian deep-learning approach for undersampled MRI reconstruction with uncertainty

Samah Khawaled1, Moti Freiman2

  • 1The Interdisciplinary program in Applied Mathematics, Faculty of Mathematics, Technion - Israel Institute of Technology, Israel.

Artificial Intelligence in Medicine
|March 10, 2024
PubMed
Summary

NPB-REC, a new Bayesian framework, enhances MRI reconstruction from undersampled data. It improves image quality and quantifies uncertainty, paving the way for safer clinical applications of deep learning.

Keywords:
MRIReconstructionUncertainty

More Related Videos

Author Spotlight: Integrated Photoacoustic, Ultrasound, and Angiographic Tomography (PAUSAT) for NonInvasive Whole-Brain Imaging of Ischemic Stroke
06:45

Author Spotlight: Integrated Photoacoustic, Ultrasound, and Angiographic Tomography (PAUSAT) for NonInvasive Whole-Brain Imaging of Ischemic Stroke

Published on: June 2, 2023

1.3K
Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
07:53

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

1.4K

Related Experiment Videos

Last Updated: Jul 14, 2026

Topographical Estimation of Visual Population Receptive Fields by fMRI
06:02

Topographical Estimation of Visual Population Receptive Fields by fMRI

Published on: February 3, 2015

9.3K
Author Spotlight: Integrated Photoacoustic, Ultrasound, and Angiographic Tomography (PAUSAT) for NonInvasive Whole-Brain Imaging of Ischemic Stroke
06:45

Author Spotlight: Integrated Photoacoustic, Ultrasound, and Angiographic Tomography (PAUSAT) for NonInvasive Whole-Brain Imaging of Ischemic Stroke

Published on: June 2, 2023

1.3K
Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
07:53

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

1.4K

Area of Science:

  • Medical Imaging
  • Machine Learning
  • Bayesian Inference

Background:

  • Accelerating MRI acquisition is crucial for clinical workflow and patient comfort.
  • Deep learning (DL) methods show promise for reconstructing MRI from undersampled data.
  • Quantifying uncertainty in DL-based MRI reconstruction is essential for clinical adoption.

Purpose of the Study:

  • Introduce NPB-REC, a non-parametric fully Bayesian framework for MRI reconstruction with uncertainty estimation.
  • Improve reconstruction accuracy and provide reliable uncertainty quantification for undersampled MRI data.
  • Enhance the clinical applicability of DL in accelerated MRI.

Main Methods:

  • Utilized Stochastic Gradient Langevin Dynamics for training to characterize posterior distribution of network parameters.
  • Developed a non-parametric fully Bayesian framework (NPB-REC) for MRI reconstruction.
  • Validated on the fastMRI challenge dataset, comparing against End-to-End Variational Network (E2E-VarNet).

Main Results:

  • NPB-REC achieved superior reconstruction accuracy (PSNR: 34.55, SSIM: 0.908) compared to E2E-VarNet (PSNR: 33.08, SSIM: 0.897) at R=8.
  • Uncertainty measures from NPB-REC showed a stronger correlation with reconstruction error (R=0.94) than the baseline (R=0.91).
  • Demonstrated improved generalization across anatomical distributions (brain to knee data).

Conclusions:

  • NPB-REC effectively reconstructs high-quality MRI from undersampled data while providing reliable uncertainty estimates.
  • The framework offers enhanced accuracy and better generalization, addressing key limitations of current DL methods.
  • NPB-REC facilitates the safe and reliable clinical use of DL for accelerated MRI acquisition.