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: Nov 19, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.1K

LCPR-Net: low-count PET image reconstruction using the domain transform and cycle-consistent generative adversarial

Hengzhi Xue1,2, Qiyang Zhang2,3, Sijuan Zou4

  • 1College of Medicine and Biological Information Engineering, Northeastern University, Shenyang, China.

Quantitative Imaging in Medicine and Surgery
|February 3, 2021
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

Corrigendum: Referenceless MR thermometry-a comparison of five methods (2017<i>Phys. Med. Biol</i>.<b>62</b>1-16).

Physics in medicine and biology·2026
Same author

[Protecting effect of Quan Duzhong Capsule() on the cartilaginous endplate so as to improve acupuncture-induced intervertebral disc degeneration].

Zhongguo gu shang = China journal of orthopaedics and traumatology·2026
Same author

Diff5T: Benchmarking human brain diffusion MRI with an extensive 5.0 Tesla k-space and spatial dataset.

Scientific data·2025
Same author

A Near-Field Coupling Array Enables Parallel Imaging and SNR Gain in MRI.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2025
Same author

Transcranial focused ultrasound stimulation alleviates NLRP3-related neuroinflammation induced by ischemic stroke via regulation of the Nespas/miR-383-3p/SHP2 pathway.

International immunopharmacology·2024
Same author

Magnetic resonance imaging classification in a percutaneous needle injury rat model of intervertebral disc degeneration.

Journal of orthopaedic surgery and research·2024

A new deep learning method, LCPR-Net, directly reconstructs high-quality full-count (FC) positron emission tomography (PET) images from low-count (LC) data. This approach enhances image quality and reconstruction speed for PET imaging.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence

Background:

  • Reducing radiation tracer dose and scanning time in PET imaging is crucial for cost reduction, artifact minimization, and scanner efficiency.
  • Low-count (LC) PET data often results in noisy images, necessitating high-quality reconstruction methods.
  • Developing effective techniques for reconstructing high-quality images from LC data is essential.

Purpose of the Study:

  • To propose LCPR-Net, a deep learning method for direct reconstruction of full-count (FC) PET images from LC sinogram data.
  • To improve the quality and efficiency of PET image reconstruction using low-dose protocols.

Main Methods:

  • Utilized a generative adversarial network (GAN) framework with a cyclic consistency constraint and least-squares loss.
  • Integrated convolutional neural networks (CNNs) and residual networks for feature extraction and image reconstruction.
Keywords:
Positron emission tomography (PET)adversarial learningdeep learningimage reconstruction

More Related Videos

Photorealistic Learned Landscapes for Augmented Reality
06:54

Photorealistic Learned Landscapes for Augmented Reality

Published on: June 27, 2025

457

Related Experiment Videos

Last Updated: Nov 19, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.1K
Photorealistic Learned Landscapes for Augmented Reality
06:54

Photorealistic Learned Landscapes for Augmented Reality

Published on: June 27, 2025

457
  • Employed domain transform (DT) to incorporate prior information into the cycle-consistent GAN (CycleGAN), optimizing computational resource usage.

    • Main Results:

    • LCPR-Net directly reconstructs FC PET images from LC sinogram data, offering faster reconstruction speeds compared to model-based iterative methods.
    • The CycleGAN framework significantly improves the quality of reconstructed PET images.
    • The method effectively handles low-count data to produce high-fidelity images.

    Conclusions:

    • The proposed LCPR-Net method demonstrates accuracy and effectiveness in FC PET image reconstruction.
    • Quantitative and qualitative evaluations confirm its superiority over other state-of-the-art techniques.
    • This deep learning approach offers a promising solution for low-dose PET imaging.