Diagnostic or prognostic study
Artificial Intelligence–Guided Lung Ultrasound by Nonexperts
- Diagnostic validation study
- multicenter
- 4 clinical sites
- masked expert readers
Authors' conclusion
In this multicenter validation study, THCPs with AI assistance achieved LUS images meeting diagnostic standards compared with LUS experts without AI.
Study context
- Population
- Adults aged ≥21 years with shortness of breath, July–December 2023each had both examinations
- Participants
- 176 adults
- Comparison
- Artificial intelligence (AI)-guided lung ultrasound (LUS) by trained health care professionals (THCPs) vs LUS experts without AI
- Outcome
- Proportion of THCP-acquired examinations of diagnostic quality per panel of 5 masked expert LUS readers
Performance
Examinations of diagnostic quality
Difference
1.7%
95% CI −1.6% to 5.0%
By group
THCP
98.3%
95% CI 95.1%–99.4%
THCP vs LUS expert
difference 1.7% (95% CI −1.6% to 5.0%)
As reported in the abstract
Original publication
Artificial Intelligence–Guided Lung Ultrasound by Nonexperts. JAMA Cardiol. Published 1 March 2025. doi:10.1001/jamacardio.2024.4991
Summarised from the published abstract; every number is checked against it