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

You might also read

Related Articles

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

Sort by
Same author

Tracer-agnostic diffusion model-based CT-free attenuation correction for brain PET: Comprehensive evaluation across 14 tracers.

Physics in medicine and biology·2026
Same author

Design optimization using GATE Monte Carlo simulations for a sub-0.5 mm resolution PET scanner with 3-layer DOI detectors.

Physics in medicine and biology·2026
Same author

An exploratory study of explainable deep learning for predicting bone mineral density using clavicle features on chest radiographs: A multi-task approach with regression and segmentation.

Journal of applied clinical medical physics·2025
Same author

In-beam range determination using an OpenPET system for helium, carbon, oxygen, and neon ion beams.

Physics in medicine and biology·2025
Same author

α7 nicotinic acetylcholine receptor, activated glia, and cognitive impairment in schizophrenia: a dual-tracer PET study.

Molecular psychiatry·2025
Same author

Development of a submillimeter resolution PET scanner with compact geometry for hybrid multi-photon/PET imaging.

Physics in medicine and biology·2025

Related Experiment Video

Updated: Jul 24, 2025

Author Spotlight: Optimized Lung MRI Protocol with Computationally Efficient Reconstruction Methods
05:07

Author Spotlight: Optimized Lung MRI Protocol with Computationally Efficient Reconstruction Methods

Published on: September 6, 2024

416

Fully 3D implementation of the end-to-end deep image prior-based PET image reconstruction using block iterative

Fumio Hashimoto1,2,3, Yuya Onishi1, Kibo Ote1

  • 1Central Research Laboratory, Hamamatsu Photonics K. K., 5000 Hirakuchi, Hamakita-ku, Hamamatsu 434-8601, Japan.

Physics in Medicine and Biology
|July 5, 2023
PubMed
Summary

Deep image prior (DIP) enables unsupervised positron emission tomography (PET) image reconstruction without training data. This study introduces a 3D PET reconstruction method using DIP, improving image quality and contrast for brain imaging.

Keywords:
deep image priorend-to-end reconstructionfully 3D PET image reconstructionpositron emission tomography (PET)

More Related Videos

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging
10:44

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

546
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

1.5K

Related Experiment Videos

Last Updated: Jul 24, 2025

Author Spotlight: Optimized Lung MRI Protocol with Computationally Efficient Reconstruction Methods
05:07

Author Spotlight: Optimized Lung MRI Protocol with Computationally Efficient Reconstruction Methods

Published on: September 6, 2024

416
Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging
10:44

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

546
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

1.5K

Area of Science:

  • Medical Imaging
  • Computational Imaging
  • Artificial Intelligence in Medicine

Background:

  • Deep Image Prior (DIP) offers unsupervised PET image reconstruction, eliminating the need for prior training datasets.
  • Current practical implementations of fully 3D PET reconstruction face limitations due to GPU memory constraints.

Purpose of the Study:

  • To develop and implement an end-to-end Deep Image Prior (DIP)-based fully 3D Positron Emission Tomography (PET) image reconstruction method.
  • To incorporate a forward-projection model into the loss function for enhanced reconstruction accuracy.
  • To address GPU memory limitations through block iteration and sequential learning.

Main Methods:

  • A novel end-to-end DIP-based fully 3D PET reconstruction algorithm was developed.
  • The optimization process was adapted using block iteration and sequential learning of block sinograms to manage memory limitations.
  • A relative difference penalty (RDP) term was integrated into the loss function to improve quantitative accuracy.

Main Results:

  • The proposed DIP-based method demonstrated superior image quality compared to traditional EM and MAP-EM algorithms in Monte Carlo simulations using [18F]FDG PET data.
  • The method effectively reduced statistical noise and preserved the contrast of brain structures and simulated tumors.
  • Preclinical studies on monkey-brain PET data showed improved fine structure resolution and contrast recovery.

Conclusions:

  • The developed DIP-based method successfully generates high-quality 3D PET images without requiring a prior training dataset.
  • This approach represents a significant advancement for practical and straightforward implementation of end-to-end DIP-based 3D PET image reconstruction.
  • The method holds promise as a key enabling technology for future PET imaging research and clinical applications.