Abstract:Introduction: Cardiac point-of-care ultrasound (POCUS) can evaluate for systolic and diastolic dysfunction to inform care in the Emergency Department (ED). However, accurate assessment can be limited by user experience. Artificial intelligence (AI) has been proposed as a model to increase the accuracy of cardiac POCUS. However, there is limited evidence of the accuracy of AI in the clinical environment. The objective of this study was to determine the diagnostic accuracy of AI for identifying systolic and diastolic dysfunction compared with expert reviewers. Methods: This was a prospective, observational study of adult ED patients aged ≥45 years with risk factors for systolic and diastolic dysfunction. Ultrasound fellowship-trained physicians used an ultrasound machine with existing AI software and obtained parasternal long axis, parasternal short axis, and apical 4-chamber views of the heart. Systolic dysfunction was defined as ejection fraction (EF) < 50 % in at least two views using visual assessment or E-point septal separation >10 mm. Diastolic dysfunction was defined as an E:A < 0.8, or ≥ 2 of the following: septal e' < 7 cm/s or lateral e' < 10 cm/s, E:e' > 14, or left atrial volume > 34 mL/m2. AI was subsequently used to measure EF, E, A, septal e', and lateral e' velocities. The gold standard was systolic or diastolic dysfunction as assessed by two independent physicians with discordance resolved via consensus. We performed descriptive statistics (mean ± standard deviation) and calculated the sensitivity, specificity, positive likelihood ratio (LR+), and negative likelihood ratio (LR-) of the AI in determining systolic and diastolic dysfunction with 95 % confidence interval (CI). Subgroup analyses were performed by body mass index (BMI). Results: We enrolled 220 patients, with 11 being excluded due to inadequate images, resulting in 209 patients being included in the study. Mean age was 60 ± 9 years, 51.7 % were women, and the mean BMI was 31 ± 8.1 mg/kg2. For assessing systolic dysfunction, AI was 85.7 % (95 %CI 57.2 % to 98.2 %) sensitive and 94.8 % (95 %CI 90.6 % to 97.5 %) specific with a LR+ of 16.4 (95 %CI 8.6 to 31.1) and LR- of 0.15 (95 % CI 0.04 to 0.54). For assessing diastolic dysfunction, AI was 91.9 % (95 %CI 85.6 % to 96.0 %) sensitive and 94.2 % (95 %CI 87.0 % to 98.1 %) specific with a LR+ of 15.8 (95 %CI 6.7 to 37.1) and a LR- of 0.09 (0.05 to 0.16). When analyzed by BMI, results were similar except for lower sensitivity in the BMI ≥ 30 vs BMI < 30 (100 % vs 80 %). Conclusion: When compared with expert assessment, AI had high sensitivity and specificity for diagnosing both systolic and diastolic dysfunction.

Internal medicine point-of-care ultrasound assessment of left ventricular function correlates with formal echocardiography

Rapid bedside identification of high-risk population in heart failure with reduced ejection fraction by acoustic cardiography.

Diagnostic accuracy of point-of-care ultrasound with artificial intelligence-assisted assessment of left ventricular ejection fraction

Ensuring competency in focused cardiac ultrasound: a systematic review of training programs

Focused cardiac ultrasound: Competency among pre‐internship medical officers in diagnosing cardiac causes of dyspnea

Hand-held cardiac ultrasound examinations performed in primary care patients by nonexperts to identify reduced ejection fraction

Accuracy of Point‐Of‐Care Cardiac Ultrasound Performed on Patients Admitted to a Pediatric Intensive Care Unit in Shock

Focused cardiac ultrasound in preoperative assessment: the perioperative provider’s new stethoscope?

Assessment of Left Ventricular Function by Emergency Physicians Using Echocardiography

Accuracy of focused cardiac ultrasound interpretation among emergency and critical care medicine residents in Ethiopia: A multi-center cross-sectional study

Agreement of ejection fraction measured by coronary CT and cardiac ultrasound in evaluating patients with chronic heart failure: an observational clinical study

Test Characteristics of Cardiac Point-of-Care Ultrasound in Children With Preexisting Cardiac Conditions

Diagnostic accuracy of artificial intelligence for identifying systolic and diastolic cardiac dysfunction in the emergency department

Agreement of point of care ultrasound and final clinical diagnosis in patients with acute heart failure, acute coronary syndrome, and shock: POCUS not missing the target

Multicenter validation study for automated left ventricular ejection fraction assessment using a handheld ultrasound with artificial intelligence

Point-of-care ultrasonography for cardiovascular conditions in family practice: between risk and opportunity

Artificial Intelligence-Based Left Ventricular Ejection Fraction by Medical Students for Mortality and Readmission Prediction

Comprehensive Assessment of Fluid Status by Point-of-Care Ultrasonography

Multi‐organ clinical ultrasound as a complement to the diagnostic process in an internal medicine consultation

SIMPLIFID LUNG ULTRASOUND IN DIAGNOSIS OF HEART FAILURE WITH PRESERVED EJECTION FRACTION IN ARTERIAL HYPERTENSION

Identifying left ventricular dysfunction using prospective electrocardiogram-triggered coronary computed tomography angiography