Cardiothoracic & vascular anesthesia
Pragmatic Evaluation of a Deep-Learning Algorithm to Automate Ejection Fraction on Hand-Held, Point-of-Care Echocardiography in a Cardiac Surgical Operating Room
Authors' conclusion
Applied to acquisitions of adequate image quality, Auto EF produced a numerical EF estimate equivalent to Simpson's method. However, when applied to acquisitions of inadequate image quality, discrepancies arose between EF estimated by Auto EF and Simpson's method.
Study context
- Population
- Adults ≥18 years scheduled for valve, aortic, CABG, heart or lung transplant surgery; apical 4-chamber transthoracic clips on hand-held ultrasound (Lumify)54 clips analysed
- Exposure
- Deep-learning automated ejection fraction (EF) (Auto EF) and experienced reader-estimated EF vs Simpson's method
- Outcome
- Correlation (and bias) of EF estimates vs Simpson's method
Methodological detailssingle-centre, noninterventional
- Single-centre
- Noninterventional
Main finding
As reported in the abstract
EF agreement with Simpson's method
| Group | Bias | R |
|---|---|---|
| Auto EF | −10.17% | 0.56 |
| Experienced reader | −9.82% | 0.80 |
Bias
Auto EF−10.17%
Experienced reader−9.82%
R
Auto EF0.56
Experienced reader0.80
Observational data: associations, not effects.
Further findings
| Outcome | Auto EF |
|---|---|
| Auto EF vs Simpson's method correlation, adequate acquisitions (subgroup) | |
| n acquisitions | 27 |
| R | 0.86 |
| Auto EF vs Simpson's method correlation, inadequate acquisitions (subgroup) | |
| n acquisitions | 27 |
| R | 0.46 |
Original publication
Pragmatic Evaluation of a Deep-Learning Algorithm to Automate Ejection Fraction on Hand-Held, Point-of-Care Echocardiography in a Cardiac Surgical Operating Room. J Cardiothorac Vasc Anesth. Published 11 January 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →