Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Aug 25, 2025

Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla

Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla

Published on: February 19, 2021

Deep learning-regularized, single-step quantitative susceptibility mapping quantification.

Zuojun Wang¹, Henry Ka-Fung Mak¹, Peng Cao¹

¹Department of Diagnostic Radiology, The University of Hong Kong, Hong Kong, China.

NMR in Biomedicine

|October 19, 2022

Summary

This summary is machine-generated.

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

Longitudinal effects of sex differences and apolipoprotein E genotype on white matter engagement among elderly.

Brain communications·2025

Same author

Increased risk of recurrent stroke in patients with impaired kidney function: results of a pooled analysis of individual patient data from the MICON international collaboration.

Journal of neurology, neurosurgery, and psychiatry·2025

Same author

Quantitative myelin water assessment for multiple sclerosis using multi-inversion magnetic resonance fingerprinting.

Medical physics·2024

Same author

Combined Quantitative amyloid-β PET and Structural MRI Features Improve Alzheimer's Disease Classification in Random Forest Model - A Multicenter Study.

Academic radiology·2024

Same author

Dynamic Changes in Long-Standing Multiple Sclerosis Revealed by Longitudinal Structural Network Analysis Using Diffusion Tensor Imaging.

AJNR. American journal of neuroradiology·2024

Same author

A Pilot Study of Short-Course Oral Vitamin A and Aerosolised Diffuser Olfactory Training for the Treatment of Smell Loss in Long COVID.

Brain sciences·2023

Same journal

Liver Diffusion Weighted MRI: Effect of Iron Overload on Apparent Diffusion Coefficient.

NMR in biomedicine·2026

Same journal

In Vivo Assessment of Placental Structure and Perfusion in Late-Gestation Pregnancies and Their Association With Fetal Growth.

NMR in biomedicine·2026

Same journal

Reproducibility of Splanchnic Blood Flow Measured Using Phase-Contrast MRI.

NMR in biomedicine·2026

Same journal

Restriction-Weighted Q-Space Trajectory Imaging (ResQ): Toward Mapping Diffusion-Time Effects With Tensor-Valued Diffusion Encoding in Human Prostate Cancer Xenografts.

NMR in biomedicine·2026

Same journal

In Vivo Quantitative Detection of PEGylated Macromolecules by Magnetic Resonance Spectroscopy.

NMR in biomedicine·2026

Same journal

Metabolic Assessment in Human Pluripotent Stem Cell-Derived Cerebral Organoids Using HR-MAS NMR Spectroscopy.

NMR in biomedicine·2026

See all related articles

A new deep learning model, SS-POCSnet, directly generates quantitative susceptibility mapping (QSM) from phase maps. This method improves accuracy and reduces underestimation of susceptibility values, showing clinical applicability for conditions like cerebral microbleeds and multiple sclerosis.

Area of Science:

Medical Imaging
Artificial Intelligence
Neuroscience

Background:

Quantitative susceptibility mapping (QSM) is crucial for neuroimaging, but traditional methods face challenges with accuracy and susceptibility underestimation.
Deep learning approaches offer potential for improving QSM reconstruction.
Developing a robust, single-step QSM method is essential for clinical translation.

Purpose of the Study:

To develop and validate a deep learning-regularized, single-step quantitative susceptibility mapping (QSM) quantification model (SS-POCSnet).
To directly generate QSM from total phase maps, enhancing accuracy and addressing underestimation issues.
To assess the clinical applicability of SS-POCSnet for various neurological conditions.

Main Methods:

A novel deep learning model, SS-POCSnet, was trained using synthesized QSM datasets.

Keywords:

data-driven deep learning quantitative susceptibility mapping single-step susceptibility quantification

More Related Videos

Quantitative Magnetic Resonance Imaging of Skeletal Muscle Disease

Quantitative Magnetic Resonance Imaging of Skeletal Muscle Disease

Published on: December 18, 2016

A Data-Driven Approach to Quantifying Immune States in Sepsis

A Data-Driven Approach to Quantifying Immune States in Sepsis

Published on: February 7, 2025

Related Experiment Videos

Last Updated: Aug 25, 2025

Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla

Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla

Published on: February 19, 2021

Quantitative Magnetic Resonance Imaging of Skeletal Muscle Disease

Quantitative Magnetic Resonance Imaging of Skeletal Muscle Disease

Published on: December 18, 2016

A Data-Driven Approach to Quantifying Immune States in Sepsis

A Data-Driven Approach to Quantifying Immune States in Sepsis

Published on: February 7, 2025

SS-POCSnet employs a single-step, iterative approach combining spherical mean value kernel and dipole models for data fidelity.

The model was evaluated on synthetic datasets and clinical data including cerebral microbleeds, calcification, and multiple sclerosis lesions.

Main Results:

SS-POCSnet demonstrated superior performance on synthetic datasets compared to other methods, achieving high accuracy metrics.
The model effectively reduced underestimation of susceptibility values in deep brain nuclei.
SS-POCSnet showed sensitivity to cerebral microbleeds, calcification, and multiple sclerosis lesions, indicating clinical utility.
The method supported variable imaging parameters, enhancing its versatility.

Conclusions:

Deep learning-regularized, single-step QSM quantification using SS-POCSnet effectively mitigates underestimation of susceptibility values in deep brain nuclei.
SS-POCSnet shows significant promise for accurate and reliable QSM reconstruction in clinical neuroimaging applications.
The model's ability to handle diverse datasets and imaging parameters underscores its potential for widespread adoption.