Prospective evaluation of a deep learning model for classifying pediatric septal defects on grayscale echocardiographic videos: comparison with three pediatric cardiologists
Abstract
Abstract Background Timely recognition of atrial and ventricular septal defects (ASDs and VSDs) remains challenging in settings where access to pediatric echocardiography expertise is limited. We prospectively evaluated a locked deep learning model for standalone classification of pediatric grayscale echocardiographic videos and compared its performance with that of three pediatric cardiologists under the same grayscale-only test condition. Methods Patients younger than 18 years undergoing clinically indicated transthoracic echocardiography at a tertiary pediatric cardiac center were screened prospectively. Participants were enrolled consecutively within each reference-standard diagnostic stratum until 35 were included in each class: ASD, VSD, and normal heart. The locked model and three blinded pediatric cardiologists independently classified identical de-identified grayscale videos without Doppler or clinical information. A single consultant pediatric cardiologist established the reference-standard diagnosis from the complete clinical echocardiographic examination. Accuracy and class-specific diagnostic measures were reported with Wilson 95% confidence intervals (CIs). Agreement was assessed using unweighted Cohen's kappa, and paired differences in accuracy were tested using two-sided exact McNemar tests with Bonferroni correction. Results Of 125 children screened, 105 were included. The locked model correctly classified 94 of 105 examinations (accuracy, 89.5%; 95% CI,…
Read the paper
Related studies
- Comparative effectiveness of GLP-1 receptor agonists on glycaemic control, body weight, and lipid profile for type 2 diabetes: systematic review and network meta-analysis
- Adverse effects of cyclooxygenase 2 inhibitors on renal and arrhythmia events: meta-analysis of randomized trials
- Benefits and harms of drug treatment for type 2 diabetes: systematic review and network meta-analysis of randomised controlled trials
- Comparative effectiveness and safety of statins as a class and of specific statins for primary prevention of cardiovascular disease: A systematic review, meta‐analysis, and network meta‐analysis of randomized trials with 94,283 participants
- Screen time increases overweight and obesity risk among adolescents: a systematic review and dose-response meta-analysis
- Exposure to WHO AWaRe antibiotics and isolation of multi-drug resistant bacteria: a systematic review and meta-analysis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.