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

Validation procedures in radiologic diagnostic models. Neural network and logistic regression.

E Arana1, P Delicado, L Martí-Bonmatí

  • 1Department of Radiology, Hospital Casa de Salud, Valencia, Spain.

Investigative Radiology
|October 6, 1999
PubMed
Summary

The neural network (NN) classifier outperformed logistic regression (LR) in predicting calvarial lesions, particularly when using three-fold cross-validation. Other resampling methods failed to show significant performance differences between the models.

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

A Fully Automated Pipeline for Vertebral Structural Assessment From Medical Images. Application Under Metastatic Conditions.

International journal for numerical methods in biomedical engineering·2026
Same author

Results of the Distributed Biomedical Imaging Network: Ten years serving the scientific community.

Radiologia·2026
Same author

Donate data, build health: Data altruism in the European Health Data Space.

Radiologia·2025
Same author

Analysing the level of evidence of publications in the journal Radiología.

Radiologia·2025
Same author

The importance of data: Monitoring variables in causal inference with medical imaging.

Radiologia·2025
Same author

Pain Management in Pediatrics: What the IR has to Offer.

Cardiovascular and interventional radiology·2025

Area of Science:

  • Radiology
  • Machine Learning
  • Medical Imaging

Background:

  • Calvarial lesions present a diagnostic challenge.
  • Predictive models can aid in diagnosis.
  • Comparing logistic regression (LR) and neural network (NN) models is crucial.

Purpose of the Study:

  • To compare the performance of logistic regression (LR) and neural network (NN) models for calvarial lesion prediction.
  • To evaluate the impact of five different resampling methods on model performance.

Main Methods:

  • One hundred sixty-seven patients with calvarial lesions were analyzed.
  • Clinical and CT data were used to develop LR and NN models.
  • Models were validated using cross-validation, leave-one-out, and bootstrap methods.

Related Experiment Videos

Main Results:

  • The neural network (NN) model achieved statistically higher area under the curve (Az) values than the logistic regression (LR) model when using three-fold cross-validation.
  • No statistically significant differences in performance were observed between LR and NN models with leave-one-out or bootstrap resampling methods.

Conclusions:

  • The neural network (NN) classifier demonstrates superior performance compared to the logistic regression (LR) model for calvarial lesion prediction.
  • Three-fold cross-validation effectively highlights the NN's advantage, whereas other resampling techniques do not.
  • The choice of validation strategy significantly influences the perceived performance difference between predictive models.