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

Dr. Agent: Clinical predictive model via mimicked second opinions.

Junyi Gao1, Cao Xiao2, Lucas M Glass2,3

  • 1Analytics Center of Excellence, IQVIA, Beijing, China.

Journal of the American Medical Informatics Association : JAMIA
|June 18, 2020
PubMed
Summary

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

Optimizing prochloraz timing for curative control of Fusarium wilt in chili: a multi-target strategy targeting early root infection.

Pest management science·2026
Same author

Fabrication of Breathable Coating and Its Hydrophobization Applied for the Rust Stabilization of Weathering Steels.

Polymers·2026
Same author

Seasonal Variation and Genetic Evaluation of Needle Catechin Content in Half-Sib Families of <i>Pinus taeda</i>.

Plants (Basel, Switzerland)·2026
Same author

Realistic PET image synthesis from MRI for automated inference of brain atrophy and Alzheimer's.

iScience·2026
Same author

Deep Continuous-Time State-Space Models for Marked Event Sequences.

Advances in neural information processing systems·2026
Same author

Exploring the Relationship between Dietary Intake and Clinical Outcomes in Peritoneal Dialysis Patients Stratified by Serum Albumin Levels: A 12-Year Follow-Up Using Fine-Grained Electronic Medical Records Data.

Health data science·2026

Dr. Agent, a novel AI system, improves clinical prediction by using two agents to analyze patient data. This approach enhances disease outcome prediction, outperforming existing models.

Area of Science:

  • Artificial Intelligence in Medicine
  • Clinical Decision Support Systems
  • Machine Learning for Healthcare

Background:

  • Predicting disease phenotypes and outcomes is challenging.
  • Patients often seek multiple expert opinions for complex diagnoses.
  • Current predictive models may not fully capture patient health trajectories.

Purpose of the Study:

  • To develop an AI system, Dr. Agent, that mimics seeking second opinions for improved clinical prediction.
  • To enhance the accuracy of disease outcome prediction by integrating patient history and demographics.
  • To evaluate Dr. Agent's performance against established clinical predictive models.

Main Methods:

  • Dr. Agent utilizes recurrent neural networks augmented with two policy gradient agents.
Keywords:
clinical predictiondeep learningelectronic health recordsintensive carerecurrent neural networkreinforcement learning

Related Experiment Videos

  • A primary agent focuses on recent patient visits, while a second-opinion agent reviews the entire patient history.
  • The model incorporates patient demographics and learns dynamic skip connections for focused information retrieval.

    • Main Results:

    • Dr. Agent demonstrated superior performance across four clinical prediction tasks: in-hospital mortality, acute care phenotype classification, physiologic decompensation, and length of stay prediction.
    • The system outperformed all four baseline clinical predictive models in all evaluated metrics.
    • Dr. Agent achieved up to a 15% increase in the area under the precision-recall curve compared to the best baseline model.

    Conclusions:

    • Dr. Agent effectively models long-term patient health dependencies and demographic influences.
    • The dual-agent approach significantly enhances prediction accuracy in various clinical prediction tasks.
    • This AI system offers a promising advancement in clinical decision support and patient outcome forecasting.