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: Jan 11, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

ASL 4D MRA Intracranial Vessel Segmentation With Deep Learning U-Nets.

Sang Hun Chung¹, Zihan Wang², Tianrui Zhao^1,3

¹Department of Radiology, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA.

Magnetic Resonance in Medicine

|November 9, 2025

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

Prospective Evaluation of Clavicle Fracture Osteosynthesis: Patient Satisfaction, Clinical, and Radiological Outcomes with Clavipectoral Fascial Plane Block.

Journal of orthopaedic case reports·2026

Same author

Advancements in Skeletal Tumour Management: Linking Metastatic Biology With Surgical Reconstruction.

Cureus·2026

Same author

RNS60 RESCUE Trial in Acute Ischemic Stroke: Post Hoc Analysis in Participants Enrolled <12 Hours Since Last Known Well.

Stroke (Hoboken, N.J.)·2026

Same author

Uncertainty-aware automated labeling of intracranial arteries using deep learning.

BMC medical imaging·2026

Same author

Technical Performance and Interpretation of Digital Subtraction Angiography to Determine Cerebral Circulatory Arrest in Patients Being Evaluated for Brain Death/Death by Neurological Criteria: A Brief Practice Update From the Society of Vascular and Interventional Neurology (SVIN).

Stroke (Hoboken, N.J.)·2026

Same author

DOSTA-Net: Domain-Shuffle Temporal Attention Network for Vessel Extraction in X-Ray Coronary Angiography Using Synthetic Data.

IEEE transactions on medical imaging·2026

Same journal

A Comparison of Tissue Property Values Estimated Using Conventional Cardiac MRF and MT-Cardiac MRF.

Magnetic resonance in medicine·2026

Same journal

Dependence of the Extra-Cellular Diffusion Coefficient on the Fractions of Neurites and Cell Bodies in Gray Matter.

Magnetic resonance in medicine·2026

Same journal

Triple-Pulse <sup>23</sup>Na MRI Sequence (TriNa) for Simultaneous Acquisition of Spin-Density-Weighted and Fluid-Attenuated Images.

Magnetic resonance in medicine·2026

Same journal

Evaluation of Phantom Doping Materials in Quantitative Susceptibility Mapping.

Magnetic resonance in medicine·2026

Same journal

Design of an 8-Channel Transmit 32-Channel Receive 11.7T Head Coil and Evaluation of SNR Gains.

Magnetic resonance in medicine·2026

Same journal

The Potential for Absolute Temperature Imaging Based on Brain Metabolites Using an FID-Shifting Approach in Gradient Echo Planar Spectroscopic Imaging (GREPSI).

Magnetic resonance in medicine·2026

See all related articles

This summary is machine-generated.

The novel 4DST network improves 4D MR angiography (4D MRA) vessel segmentation by effectively using spatial and temporal data. This method enhances accuracy and reduces processing needs for medical imaging analysis.

Area of Science:

Medical Imaging
Artificial Intelligence
Biomedical Engineering

Background:

Accurate segmentation of 4-dimensional (4D) MR angiography (MRA) is crucial for diagnosing vascular diseases.
Pulsed ASL-based 4D MRA offers non-contrast enhanced imaging but presents segmentation challenges.
Existing methods often struggle with computational demands or suboptimal performance.

Purpose of the Study:

To introduce a novel spatio-temporal U-Net based network, termed 4DST.
To enable efficient 4D MRA vessel segmentation by leveraging spatial and dynamic information.
To avoid memory-intensive 4D convolutional layers in the segmentation process.

Main Methods:

Developed and tested the 4DST network architecture for pulsed ASL-based 4D MRA.

Keywords:

4D MRA ASL U‐net arteriovenous malformation deep learning vessel segmentation

More Related Videos

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Jan 11, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

Compared 4DST against spatial-only (2D, 3D) U-Net variations and other state-of-the-art methods (isolation forest, BRAVE-Net).

Evaluated performance using metrics like Dice-Sørensen coefficient (DSC), center-line Dice (clDice), Hausdorff distance (HD), and sensitivity relative to SNR and arterial transit time (ATT).

Main Results:

4DST achieved superior segmentation performance, evidenced by the best DSC (0.876 ± 0.03), clDice (0.865 ± 0.02), and HD (6.241 ± 0.95).
The model demonstrated high sensitivity across a wide range of SNR (1-10) and ATT (500-800 ms).
4DST segmentations more accurately reflected ground truth for total vessel length and branch splits compared to other methods.

Conclusions:

The 4DST network provides a significant advancement in 4D MRA vessel segmentation.
It offers an end-to-end trainable framework suitable for spatio-temporal datasets.
4DST presents an attractive, low-preprocessing solution for pulsed ASL-based 4D MRA vessel segmentation.