The diagnostic accuracy of left ventricular ejection fraction assessment between visual and artificial intelligence-based algorithms on bedside ultrasound

Q: What are the key findings of this study?

AI-assisted FoCUS LVEF assessments showed good agreement with formal TTE but were outperformed by experienced bedside users' visual estimates.

Q: What is the clinical evidence from this study?

Study design: Observational. Population: Patients requiring a transthoracic echocardiogram (n=215). Intervention: AI-based focused cardiac ultrasound (FoCUS) vs. Visual bedside assessment and formal transthoracic echocardiography (TTE). Primary outcome: Agreement with formal TTE for LVEF assessment (Intraclass correlation coefficient) (ICC 0.84, 95% CI 0.80-0.88).

Q: What does this research mean for the field?

AI-assisted focused cardiac ultrasound provides accurate estimates for the presence and severity of left ventricular dysfunction, though it is slightly outperformed by the visual assessments of experienced bedside users. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Nikola Kolobaric; Nickolas Beauregard; William Barbour; Graeme Prosperi‐Porta; Simon Parlow; Pietro Di Santo; Omar Abdel-Razek; Richard Jung; William B. Bradford; Miranda Tsang; Stefano Pacifici; F. Daniel Ramirez; Gordon S. Huggins; Ann Bugeja; Trevor Simard; Rebecca Mathew; Benjamin Hibbert; Pouya Motazedian; Jeffrey A. Marbach

doi:10.1186/s12947-026-00377-3

Key result

AI-assisted FoCUS LVEF assessments agree well with formal TTE but underperform experienced bedside visual estimates.

ICC 0.84
95% CI 0.80-0.88
n=215

Figures

Why the study?

With the integration of AI into focused cardiac ultrasound, there is limited evidence comparing AI to bedside visual assessments by experienced users.

Does AI-assisted FoCUS LVEF assessment improve diagnostic accuracy compared to visual assessment by trained sonographers in patients requiring TTE?

Key Points

To compare the diagnostic accuracy of artificial intelligence-based algorithms versus visual assessment by experienced bedside clinicians for evaluating left ventricular ejection fraction on focused cardiac ultrasound.
Prospective study conducted at Tufts Medical Center from December 2020 to March 2022 enrolling 215 adults (median age 63 years, 38.6% female) referred for formal transthoracic echocardiogram (TTE).
Patients underwent bedside focused cardiac ultrasound (FoCUS) evaluated both by an AI algorithm using Simpson's biplane method of disks and by visual assessment from trained bedside sonographers, with formal TTE serving as reference.
For detecting abnormal LVEF (< 50%), AI-FoCUS showed an AUC of 0.9779 (95% CI: 0.9591–0.9967) with sensitivity 90.9% (95% CI: 88.1–100) and specificity 95.6% (95% CI: 92.6–98.6), whereas visual assessment achieved an AUC of 0.9961 (95% CI: 0.9804–1.0000; p=0.03) with sensitivity 90.9% (95% CI: 81.1–100) and specificity 97.8% (95% CI: 95.7–99.9).
Quantitative agreement with formal TTE was good for AI-FoCUS (intraclass correlation coefficient 0.84, 95% CI: 0.80–0.88; weighted kappa 0.89, 95% CI: 0.84–0.94), but expert visual assessment demonstrated higher concordance (ICC 0.97, 95% CI: 0.96–0.98; weighted kappa 0.96, 95% CI: 0.92–0.99).

Study Design

Type

Observational (n=215)

Multicenter

No

Structured PICO

Does AI-assisted FoCUS LVEF assessment improve diagnostic accuracy compared to visual assessment by trained sonographers in patients requiring TTE?

P

Population

215 adult patients requiring a transthoracic echocardiogram underwent bedside focused cardiac ultrasound to compare visual and AI-based left ventricular ejection fraction assessments.

E

Exposure

Focused cardiac ultrasound (FoCUS) left ventricular ejection fraction (LVEF) classification by artificial intelligence (AI) using the EchoNous KOSMOS device (calculated by modified Simpson's biplane method of disks).

C

Comparator

Bedside visual global assessment of LVEF by trained cardiac sonographers using the same EchoNous KOSMOS device. Both methods were compared against formal cart-based TTE as the reference standard.

O

Outcome

Diagnostic accuracy and agreement of LVEF assessment compared to formal TTE, measured by intraclass correlation coefficient (ICC) for continuous LVEF and Cohen's weighted kappa for severity categorization.surrogate

Main Result

Effect estimate: ICC 0.84 (95% CI 0.80-0.88)

Absolute Event Rate: 0.84% vs 0.97%

While AI-assisted FoCUS provides accurate estimates of LV dysfunction, it is currently outperformed by the visual assessments of experienced bedside sonographers.

Limitations

Participants were enrolled through convenience sampling, which may introduce operator bias.
Single-center design limits the generalizability of the findings.
Small number of participants with severely reduced LVEF (< 30%) limits conclusions for this subgroup.
Bedside visual assessments were completed by trained cardiac sonographers, which may not reflect the performance of all real-world FoCUS users.
Convenience sampling introducing potential operator bias
Single-center design limiting generalizability
Very small number of participants with severely reduced LVEF (n=4)
Bedside visual assessments completed by trained cardiac sonographers, which may not be generalizable to novice FoCUS users

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

AI-assisted FoCUS may support bedside LVEF assessment; leaves open superiority to expert visual estimation in routine care.

Abstract

Focused cardiac ultrasound (FoCUS) has become the standard of care for bedside assessments of cardiac function. With the integration of artificial intelligence (AI), there is limited evidence comparing it to bedside visual assessments by experienced users. In our prospective study conducted at Tufts Medical Center in Boston, Massachusetts from December 2020 to March 2022, patients ≥ 18 years requiring a TTE were recruited by convenience sampling. They each underwent FoCUS LVEF classification by AI and bedside visual assessment, with TTE as reference. LVEF was calculated by Simpson’s biplane method of disks in AI-FoCUS and TTE, and visual global assessment by the bedside sonographer. Data analysis was completed in October 2025. Our 215 participants had a median age of 63 (IQR 49–73) years with 83 (38.6%) being female. AI-FoCUS assessments showed good agreement with TTE (intraclass correlation coefficient 0.84, 95% CI: 0.80–0.88), while visual assessments had stronger concordance (intraclass correlation coefficient 0.97, 95% CI: 0.96–0.98). Categorization of LV dysfunction severity showed excellent agreement with TTE for both AI-FoCUS (kappa 0.89, 95% CI: 0.84–0.94) and visual estimate (kappa 0.96, 95% CI: 0.92–0.99). For AI-FoCUS, the area under the curve (AUC) for identifying an abnormal LVEF (< 50%) was 0.9779 (95% CI: 0.9591–0.9967) with sensitivity 90.9% (95% CI: 88.1–100), specificity 95.6% (95% CI: 92.6–98.6), positive predictive value (PPV) 0.79 (95% CI: 0.66–0.92) and negative predictive value (NPV) 0.98 (95% CI: 0.96–1.00). For visual estimate, the AUC was 0.9961 (95% CI: 0.9915–1.000) with sensitivity 90.9% (95% CI: 88.1–100), specificity 97.8% (95% CI: 95.7–99.9), PPV 0.88 (95% CI: 0.77–0.99) and NPV 0.98 (95% CI: 0.96–1.00). AI-assisted FoCUS LVEF assessments provide accurate estimates for the presence and severity of LV dysfunction but are outperformed in the latter by experienced bedside users.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Kolobaric et al. (2026) conducted an observational in Patients requiring a transthoracic echocardiogram (n=215). AI-based focused cardiac ultrasound (FoCUS) vs. Visual bedside assessment and formal transthoracic echocardiography (TTE) was evaluated on Agreement with formal TTE for LVEF assessment (Intraclass correlation coefficient) (ICC 0.84, 95% CI 0.80-0.88). AI-assisted FoCUS LVEF assessments showed good agreement with formal TTE but were outperformed by experienced bedside users' visual estimates.

synapsesocial.com/papers/6a97e237c562ede874ec636a https://doi.org/10.1186/s12947-026-00377-3

View Full Paper

Ask AI

Bookmark

Share

The diagnostic accuracy of left ventricular ejection fraction assessment between visual and artificial intelligence-based algorithms on bedside ultrasound

Figures

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

The diagnostic accuracy of left ventricular ejection fraction assessment between visual and artificial intelligence-based algorithms on bedside ultrasound

Figures

Key Points

Study Design

Structured PICO

Main Result

Limitations

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading