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

Reporting follow-up budgets for diagnostic and predictive AI.

Henry Bair1, Mak Djulbegovic1

  • 1Wills Eye Hospital, Philadelphia, PA, United States.

Journal of the American Medical Informatics Association : JAMIA
|July 8, 2026
PubMed
Summary

Evaluating diagnostic and predictive artificial intelligence (AI) requires assessing not only classification performance but also the downstream workload generated by AI outputs. This ensures a comprehensive understanding of AI

Related Concept Videos

Health Information Technology and Healthcare Information System01:30

Health Information Technology and Healthcare Information System

Health Information Technology (HIT)
Health Information Technology, commonly called HIT, integrates advanced information systems and technology in healthcare settings. Its primary functions include:
Issues And Trends In Healthcare Delivery System01:29

Issues And Trends In Healthcare Delivery System

The issues and trends in healthcare delivery are constantly changing. The COVID-19 pandemic is one recent issue that wreaked havoc on healthcare systems, causing a shortage of healthcare workers, high demand for medicines and supplies, and increased medical expenditure due to a lack of insurance. Other issues include rising healthcare costs and care fragmentation.
Cost Containment
Payment for healthcare services has historically promoted adoption of costly and often unnecessary or inefficient...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Comment on: Enhancing patient education in cataract surgery using a conversational artificial intelligence chatbot: pilot randomized controlled trial.

Journal of cataract and refractive surgery·2026
Same author

GLP-1 Receptor Agonists and the Ocular Surface: A Narrative Review of Restoration, Remodeling, and Clinical Implications.

Ophthalmology and therapy·2026
Same author

The first reader: Montaigne and the body's particularity.

Palliative & supportive care·2026
Same author

Structured Clinicopathologic Data as a Prerequisite for Generalizable AI in Ocular Oncology.

Ocular oncology and pathology·2026
Same author

When the Employer Is the Gatekeeper: Rethinking Digital Health Task Shifting.

The American journal of bioethics : AJOB·2026
Same author

Digital Identity as Patient Safety Infrastructure.

Applied clinical informatics·2026

Area of Science:

  • Medical Informatics
  • Artificial Intelligence in Healthcare
  • Clinical Decision Support

Background:

  • Current evaluations of artificial intelligence (AI) in healthcare often focus on classification performance.
  • Existing frameworks quantify alert burden or decision utility under resource constraints.
  • There is a need to extend AI evaluation beyond surveillance alerts to encompass broader clinical impact.

Purpose of the Study:

  • To advocate for a dual evaluation approach for diagnostic and predictive AI.
  • To propose the inclusion of downstream workload assessment in AI evaluations.
  • To highlight the importance of considering the practical implications of AI outputs in clinical settings.

Main Methods:

  • The study proposes a "follow-up budget reporting" framework for diagnostic and predictive AI.
Keywords:
artificial intelligenceclinical decision supporthealth care capacityimplementation sciencepredictive models

Related Experiment Videos

  • This framework extends previous work on AI as an intervention.
  • It draws parallels with existing methods for quantifying alert burden and decision utility.

    • Main Results:

    • Similar classification performance in AI tools can lead to vastly different downstream resource demands.
    • Examples include lung nodule detection, sepsis alerting, and diabetic retinopathy screening.
    • These differences impact imaging, specialist access, nursing care, and patient wait times.

    Conclusions:

    • Clinical AI evaluation must incorporate "follow-up capacity" alongside traditional metrics.
    • Metrics should include discrimination, calibration, and decision-analytic measures.
    • A comprehensive "follow-up budget statement" should detail comparators, scope, stopping rules, action volumes, resource types, yield, and pathway assumptions.