Real-life clinical validation of artificial intelligence-assisted echocardiography for aortic root and left ventricular outflow tract measurements against human readers and cardiac magnetic resonance.

Routine echocardiographic measurements in clinical practice: reproducibility of left and right ventricular indices and agreement with cardiac magnetic resonance

Background Although the reproducibility of individual echocardiographic measurements has been previously investigated, evidence evaluating a comprehensive panel of routinely used left- and right-sided parameters within standardized real-world clinical workflows from Latin American centers remains limited. We aimed to evaluate the reproducibility of routine echocardiographic measurements, including intraobserver and interobserver variability, and to examine their agreement with cardiac magnetic resonance (CMR). Methods Two cohorts of adult participants were included between 2022 and 2023: a clinical cohort consisting of patients referred for transthoracic echocardiography because of cardiovascular symptoms or routine cardiovascular evaluation, and a healthy reference group with no history of cardiovascular disease or self-reported systemic illness. Left- and right-sided echocardiographic measurements, including ventricular, atrial, and left ventricular outflow tract (LVOT) indices, were analyzed. CMR was used to assess left ventricular volumes and ejection fraction. Reproducibility was assessed by two echocardiographers using intraclass correlation coefficients and Bland–Altman analysis. Results Sixty-three participants were included, 20 in the healthy reference group and 43 in the clinical cohort. Left ventricular measurements showed excellent reproducibility, whereas right-sided indices showed overall good reproducibility. LVOT diameter showed the lowest concordance. Compared with CMR, echocardiography underestimated left ventricular volumes, while agreement for left ventricular ejection fraction remained excellent. Conclusions In routine clinical practice, standard echocardiographic measurements showed overall good reproducibility, with the strongest performance for left ventricular indices. Greater variability was observed for selected right ventricular measurements and LVOT diameter, whereas agreement with CMR was excellent for left ventricular ejection fraction despite underestimation of ventricular volumes.

Agreement of AI-assisted semi-automated right ventricular function analysis using 3D transthoracic echocardiography with artificial intelligence.

AI-based 3D TTE semi-automated measurement has the potential to enhance the efficiency and reproducibility of RV function assessment, reducing examiner workload and improving clinical workflow.

Journal of Community Hospital Internal Medicine Perspectives Journal of Community Hospital Internal Medicine Perspectives

Artificial intelligence-based prediction of transthoracic echocardiographic image quality in ICD/CRT-D candidates: A proof-of-concept study.

BACKGROUND High-quality echocardiography is essential for accurate and reproducible assessment of cardiac functional indices, which are highly dependent on adequate image quality and proper probe alignment. An artificial intelligence (AI)-based approach may enable automated image quality assessment. AIMS Our aim was to develop and internally validate an AI-based algorithm to predict image quality from selected echocardiographic frames in candidates for implantable cardioverter-defibrillator and cardiac resynchronization therapy with a defibrillator implantation. METHODS In this retrospective cross-sectional study, 248 patients (297 echocardiographic examinations) were included. Demographic, electrocardiographic, echocardiographic, and clinical data were collected. Apical 2-, 3-, and 4-chamber views were extracted for image quality analysis, yielding a total of 909 echocardiograms. Image quality was assessed using end-diastolic frames. An internally validated scoring framework was applied, demonstrating high interclass correlation. RESULTS Regression models provided more clinically relevant information than classification models. Visual transformer models achieved Pearson correlation coefficients similar to those of convolutional neural networks (up to 0.812 and 0.772, respectively; P = 0.31). Architectures trained on end-diastolic frames achieved comparable Pearson correlation coefficients to those trained on combined end-diastolic and end-systolic frames. Compared with human experts, the models showed significantly higher absolute percentage errors, with values of 11%-12% (median) vs. 7.9% (mean) for the total image quality score and 13%-14% (median) vs. 8.8% (mean) for the border quality score. CONCLUSIONS Regression models demonstrated the highest performance. An internally validated AI model can predict an echocardiographic image quality score in a small cohort of cardiac resynchronization therapy with a defibrillator/implantable cardioverter-defibrillator candidates. However, external and prospective validation will be required to establish its generalizability, reliability, and utility before clinical application.

Quantification of Single Valve Regurgitation in the Pediatric Population Using 3-Dimensional Echocardiography: Reproducibility and Comparison with Cardiac Magnetic Resonance Imaging.

Validation of Feature-Tracking Cardiac CT for Left Ventricular Strain Assessment in Pediatric Acute Myocarditis: Comparison With Cardiac MRI.

OBJECTIVE This study aimed to validate feature-tracking cardiac CT (FT-CCT) for left ventricular (LV) strain assessment against cardiac MRI (CMR) in pediatric acute myocarditis, evaluating its potential as a pragmatic solution when CMR is delayed after a clinically indicated coronary CT angiography (CCTA). MATERIALS AND METHODS Forty-five children with acute myocarditis underwent CCTA and CMR. Global longitudinal (GLS), circumferential (GCS), and radial strain were measured. Intermodality agreement (Bland-Altman, ICC), diagnostic performance for detecting late gadolinium enhancement (LGE) (ROC analysis, DeLong test), and heart rate influence were assessed. RESULTS FT-CCT showed excellent agreement with CMR for LV ejection fraction (bias: -0.1%) and strain (ICCs: 0.750-0.854), independent of heart rate. For detecting LGE, CCTA-derived GCS and GLS demonstrated diagnostic performance equivalent to CMR (AUCs: 0.716 vs. 0.718, P=0.974; 0.705 vs. 0.703, P=0.972). CCTA-derived GLS provided high specificity (93.1%) and PPV (77.8%), serving as a robust "rule-in" tool. Measurements were highly reproducible (ICCs: 0.746-0.809). CONCLUSION FT-CCT applied to clinically indicated CCTA provides a heart rate-robust, reproducible alternative to CMR for LV strain assessment in pediatric myocarditis. Its high specificity and diagnostic equivalence support its role as a complementary tool for initial myocardial injury detection and risk stratification when CMR is not immediately available.