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

HIV lipodystrophy case definition using artificial neural network modelling.

John P A Ioannidis1, Thomas A Trikalinos, Matthew Law

  • 1Clinical and Molecular Epidemiology Unit, Department of Hygiene and Epidemiology, University of Ioannina School of Medicine, Ioannina, Greece. jioannid@cc.uoi.gr

Antiviral Therapy
|December 3, 2003
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

Less frequent clinical and viral load assessments during COVID did not increase virological failure.

International journal of STD & AIDS·2026
Same author

Measuring Value in Orthopaedics: The U.S. and U.K. Perspectives.

The Journal of bone and joint surgery. American volume·2026
Same author

Switch to fixed-dose doravirine (100 mg) and islatravir (0·25 mg) once daily in virologically suppressed adults with HIV-1 on oral antiretroviral therapy: 48-week results of a phase 3, multicentre, randomised, open-label, non-inferiority trial.

Lancet (London, England)·2026
Same author

Translational Potential of an Electrospun Polycaprolactone Scaffold for Anterior Cruciate Ligament Reconstruction.

Advanced fiber materials·2026
Same author

Study protocol for a randomised controlled trial to determine the efficacy of lisdexamfetamine for the treatment of acute methamphetamine withdrawal in inpatient settings.

BMJ open·2025
Same author

Anal cancer screening in men who have sex with men living with HIV: The pilot of anal screening study (PASS).

HIV medicine·2025
Same journal

Safety and efficacy of deltoid intramuscular long-acting cabotegravir and rilpivirine, a case series.

Antiviral therapy·2026
Same journal

Factors associated with immunological non-response in HIV patients receiving antiretroviral therapy in Southeast Asia: A systematic review.

Antiviral therapy·2026
Same journal

Epidemiology and clinical management of cytomegalovirus infection in transplant recipients in Central and Eastern Europe - a physician survey.

Antiviral therapy·2026
Same journal

It's about time: The association between abacavir and cardiovascular disease.

Antiviral therapy·2026
Same journal

Combination antiviral and anti-CMV immunoglobulin therapy for CMV infection: A case-control study.

Antiviral therapy·2026
Same journal

Harnessing SuperNatural database to identify VP35 inhibitors as anti-Ebola drug candidates: A multistage <i>In silico</i> study.

Antiviral therapy·2025
See all related articles

Artificial neural networks show improved diagnostic accuracy for HIV lipodystrophy using clinical data alone. Metabolic information did not enhance diagnosis unless body composition data were included.

Area of Science:

  • Medical Informatics
  • Artificial Intelligence in Medicine
  • HIV/AIDS Research

Background:

  • A recent case definition for HIV lipodystrophy combines clinical, metabolic, and body composition variables.
  • Logistic regression models are currently used for diagnosis based on these variables.

Purpose of the Study:

  • To evaluate if artificial neural networks (ANNs) can enhance the diagnostic accuracy of HIV lipodystrophy compared to logistic regression.
  • To assess the impact of different data types (clinical, metabolic, body composition) on diagnostic performance.

Main Methods:

  • Back-propagation neural networks were trained and validated on a dataset of 788 subjects (504 for training, 284 for validation).
  • Performance was compared against logistic regression models using clinical variables only, and then with added metabolic and body composition variables.

Related Experiment Videos

  • Key performance metrics included sensitivity, specificity, overall accuracy, and area under the ROC curve.

    • Main Results:

    • ANNs using only clinical variables (41 items) outperformed logistic regression in specificity (71.2% vs 62.9%) and area under the ROC curve (0.784 vs 0.748).
    • Performance of ANNs was robust even when excluding less readily available clinical parameters.
    • Including metabolic variables did not significantly improve ANN performance over logistic regression; body composition variables yielded similar results for both methods.

    Conclusions:

    • Artificial neural networks demonstrate potential for improving HIV lipodystrophy diagnosis, particularly when relying solely on clinical data for rapid assessment.
    • Metabolic parameters alone do not appear to enhance diagnostic accuracy unless combined with imaging and body composition data.
    • ANNs offer a promising alternative for improving diagnostic discrimination in HIV lipodystrophy management.