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: Jan 9, 2026

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

723

A Pilot Study: PET/CT Cross-Modal-Based Multi-Head Fusion Attention Generative Adversarial Network (MHFA-GAN) for PET

Xin Tian1,2, Xuetai Chen3, Kishore Krishnagiri Manoj Doss4

  • 1School of Medical Imaging, Xuzhou Medical University, Xuzhou, 221004, China.

Journal of Imaging Informatics in Medicine
|December 9, 2025
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

From Single-cell Insights to Clinical Relevance: An M2 Macrophagebased Prognostic Model for Osteosarcoma.

Current medicinal chemistry·2026
Same author

A Foundation Model for Generalisable Detection of Maxillary Sinus Abnormalities: A Multicentre and Clinical Applicability Study.

Journal of oral rehabilitation·2026
Same author

Innovation and Research Master Lecture Series: A Feasible Flexible Voluntary Module Associated With Basic Research Engagement: A 2-Year Retrospective Observational Study.

European journal of dental education : official journal of the Association for Dental Education in Europe·2026
Same author

OsCDR1 Mediates ABA-BR Antagonism to Enhance Drought Tolerance in Rice.

Journal of agricultural and food chemistry·2026
Same author

Stereoselective C1-Arylation of Native Sugars via Ligand-Promoted Nickel Catalysis.

Journal of the American Chemical Society·2026
Same author

From Environmental Exposure to Intervertebral Disc Degeneration: First Evidence of Pro-Degenerative Effects of Polyamide 6 Microplastics.

Biomedicines·2026

This study introduces a novel generative adversarial network (MHFA-GAN) to improve positron emission tomography (PET) image resolution without hardware upgrades. The method enhances image quality and diagnostic accuracy cost-effectively.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence

Background:

  • Positron emission tomography (PET) is crucial for disease detection, but high-resolution (HR) imaging requires expensive equipment and lengthy reconstruction.
  • Current methods face challenges in balancing spatial resolution and reconstruction efficiency in PET imaging.

Purpose of the Study:

  • To introduce a novel multi-head fusion attention generative adversarial network (MHFA-GAN) for enhancing PET image spatial resolution and reconstruction efficiency.
  • To improve PET image quality and diagnostic capabilities without necessitating hardware upgrades.

Main Methods:

  • Developed a multi-head fusion attention generative adversarial network (MHFA-GAN) incorporating multi-head mixed convolution and an enhanced multi-head fusion attention module (EMHFA).
  • Utilized a global residual block (GRB) for balancing local details and global context during reconstruction.
Keywords:
Deep learningGANMultimodal imagingPET/CT imagingSuper-resolution

More Related Videos

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

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

Published on: November 30, 2022

3.3K

Related Experiment Videos

Last Updated: Jan 9, 2026

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

723
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

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

Published on: November 30, 2022

3.3K
  • Integrated complementary features from PET and computed tomography (CT) images for improved cross-modal information alignment.

    • Main Results:

    • MHFA-GAN demonstrated superior performance over state-of-the-art methods in qualitative and quantitative evaluations across various datasets (small animal, phantom, capillary, 22Na).
    • Achieved a spatial resolution of 0.625 mm (FWHM) on the capillary dataset, nearing the physical limit of the scanner and outperforming HR images (0.871 mm).
    • Showcased significant results in super-resolution tasks on different device datasets, confirming cross-device applicability and robustness.

    Conclusions:

    • MHFA-GAN effectively enhances PET image spatial resolution and reconstruction efficiency, offering a cost-effective solution for improved clinical diagnostics.
    • The proposed method provides a viable alternative to hardware upgrades for achieving high-quality PET imaging.
    • The model's robustness and cross-device applicability highlight its potential for widespread adoption in medical imaging.