Prospective Human Validation of Artificial Intelligence Interventions in Cardiology: A Scoping Review
Authors' conclusion
This review identifies the potential of AI in cardiology, with models often performing equally well as human counterparts for specific and clearly scoped tasks suitable for such models.
Study context
- Population
- Peer-reviewed studies of prospective human validation of artificial intelligence (AI)-based interventions in cardiology, January 2015 to December 2023 (Scopus, MEDLINE)studies included
- Participants
- 64 studies
- Comparison
- AI interventions vs human experts or gold standard practices
- Outcome
- Characteristics of prospectively human-validated AI interventions in cardiology (included studies)
Pooled result
As reported in the abstract
- Studies included (of 2,351 initial records)
- 64
- Published after 2020
- 59 (92.2%)
- RCTs
- 11 (17.2%)
- AI reported definite clinical or operational improvements over human experts
- 44 articles (68.75%)
- Application area: imaging
- n = 14, 21.9%
- Application area: EF
- n = 10, 15.6%
- Application area: arrhythmia
- n = 9, 14.1%
- Application area: CAD
- n = 12, 18.8%
- Convolutional neural networks (most common model)
- n = 44, 69%
- Images (most used data type)
- n = 38, 54.3%
- Models or data made accessible
- 22 (34.4%)
Original publication
Prospective Human Validation of Artificial Intelligence Interventions in Cardiology: A Scoping Review. JACC Adv. Published 28 August 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →