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: Jul 7, 2026

Anatomically Inspired Three-dimensional Micro-tissue Engineered Neural Networks for Nervous System Reconstruction, Modulation, and Modeling
10:45

Anatomically Inspired Three-dimensional Micro-tissue Engineered Neural Networks for Nervous System Reconstruction, Modulation, and Modeling

Published on: May 31, 2017

An artificial neural network for SPECT image reconstruction.

C R Floyd1

  • 1Dept. of Radiol., Duke Univ., Durham, NC.

IEEE Transactions on Medical Imaging
|January 1, 1991
PubMed
Summary

A novel artificial neural network reconstructs quantitative single photon emission computed tomographic (SPECT) images. This AI approach learns filters for projections, producing results comparable to traditional filtered backprojection methods.

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

Bayesian reconstruction and use of anatomical a priori information for emission tomography.

IEEE transactions on medical imaging·1996
Same author

Reconstruction of SPECT images using generalized matrix inverses.

IEEE transactions on medical imaging·1992
See all related articles

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computational Science

Background:

  • Quantitative imaging is crucial for accurate medical diagnosis.
  • Traditional image reconstruction methods like filtered backprojection have limitations.
  • Developing advanced algorithms for improved image reconstruction is an ongoing need.

Purpose of the Study:

  • To develop and evaluate an artificial neural network for quantitative SPECT image reconstruction.
  • To enable the network to learn optimal filters for projection data.
  • To compare the performance of the neural network method with conventional techniques.

Main Methods:

  • An artificial neural network was designed for SPECT image reconstruction.
  • The network was trained using ideal projection-image pairs.
  • A backpropagation algorithm was employed to minimize mean squared error during training.
  • The trained network applies learned filters to acquired projection data and then backprojects to form an image.

Main Results:

  • The trained neural network learned a shift-invariant weighting (filter) for projections.
  • The network's impulse response mimicked the ramp filter used in filtered backprojection.
  • Reconstructed images using the neural network showed similarity to those from filtered backprojection.
  • The method allows for quantitative SPECT image reconstruction.

Conclusions:

  • Artificial neural networks can effectively reconstruct quantitative SPECT images.
  • The developed network provides an alternative to traditional filtered backprojection.
  • This AI-driven approach shows promise for enhancing medical imaging analysis.

Related Experiment Videos

Last Updated: Jul 7, 2026

Anatomically Inspired Three-dimensional Micro-tissue Engineered Neural Networks for Nervous System Reconstruction, Modulation, and Modeling
10:45

Anatomically Inspired Three-dimensional Micro-tissue Engineered Neural Networks for Nervous System Reconstruction, Modulation, and Modeling

Published on: May 31, 2017