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 Videos

Lesion size quantification in SPECT using an artificial neural network classification approach

G D Tourassi1, C E Floyd

  • 1Department of Radiology, Duke University Medical Center, Durham, North Carolina 27710, USA.

Computers and Biomedical Research, an International Journal
|June 1, 1995
PubMed
Summary
This summary is machine-generated.

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

Experimental detection of iron overload in liver through neutron stimulated emission spectroscopy.

Physics in medicine and biology·2008
Same author

Neutron stimulated emission computed tomography: a Monte Carlo simulation approach.

Physics in medicine and biology·2007
Same author

Feature and knowledge based analysis for reduction of false positives in the computerized detection of masses in screening mammography.

Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference·2007
Same author

Application of the mutual information criterion for feature selection in computer-aided diagnosis.

Medical physics·2002
Same author

Computerized classification of suspicious regions in chest radiographs using subregion Hotelling observers.

Medical physics·2002
Same author

A neural network approach to breast cancer diagnosis as a constraint satisfaction problem.

Medical physics·2001

A novel artificial neural network (ANN) quantifies lesion sizes in SPECT images. Compared to a Bayesian algorithm, the ANN shows comparable performance when distribution parameters are estimated, demonstrating its potential in medical imaging analysis.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computational Biology

Background:

  • Accurate lesion size quantification is crucial for diagnosis and treatment monitoring in medical imaging.
  • Single photon emission computed tomography (SPECT) is a key nuclear medicine imaging technique.

Purpose of the Study:

  • To develop and evaluate an artificial neural network (ANN) for lesion size quantification in SPECT images.
  • To compare the ANN's performance against an optimal Bayesian algorithm.

Main Methods:

  • A Learning Vector Quantizer (LVQ) artificial neural network was trained using image neighborhoods around detected lesions.
  • The ANN's size quantification performance was evaluated at two noise levels.
  • Performance was benchmarked against a Bayesian algorithm using unreconstructed projection data.

Related Experiment Videos

Main Results:

  • The Bayesian algorithm, with explicit knowledge of distributions, outperformed the ANN.
  • When distributions were estimated from the training data, the ANN achieved performance comparable to the Bayesian algorithm.
  • The ANN's performance was evaluated under varying noise conditions.

Conclusions:

  • Artificial neural networks, specifically LVQ, offer a viable approach for lesion size quantification in SPECT imaging.
  • The ANN demonstrates robust performance, comparable to established Bayesian methods, particularly in realistic scenarios with estimated distributions.
  • This study highlights the potential of ANNs in enhancing quantitative analysis of medical imaging data.