Background: Mastery of cardiac auscultation is a core component of cardiology training. New computer-based auscultation teaching tools have been developed to help teach this skill to trainees.
Objective: To assess incoming pediatric cardiology fellows’ auscultation skills and evaluate a phone-based application (app) to teach auscultation.
Conclusions: The use of an app-based teaching tool is an effective method to teach auscultation skills to incoming pediatric cardiology fellows. The app can improve fellows’ comfort with identifying the most common lesions seen in pediatric cardiology.
Objective: Assess the feasibility of using an electronic stethoscope with eMurmur AI and remote cardiologist interpretation of recorded heart sounds to screen for pathologic murmurs.
Results: Among 137 subjects, 96% were successfully recorded (80 male, average age 6.1 years, SD 6.1, 5 not recorded due to network connectivity or failed quality check). Of the 132 obtained, 5 recordings were uninterpretable (1 by AI and 4 by remote cardiologist due to noise). Overall, net feasibility was 93%. 53 subjects were referred to the cardiology clinic by primary providers because of a murmur, of which 10 had pathology. For these, accuracy of the AI algorithm was similar to that of the remote cardiologist.
Objectives: This study explores whether incorporating an artificial intelligence (AI) algorithm (eMurmur ID) and telemedicine to assist referral decision-making can decrease the number of inappropriate echocardiograms and improve clinical outcomes for patients with heart murmurs when compared to current referral behavior and outcomes.
Conclusions: Accuracy of the AI-algorithm is similar to a trained cardiologist (132 patients, 84% accuracy). A new paradigm to screen for pathologic heart murmurs in primary care settings is feasible and could reduce the number of patients with only innocent heart murmurs being referred inappropriately to cardiology or echocardiography by up to 50-75%. It could help confirm the clinician’s suspicion of important pathologic heart murmurs while providing evidence to reassure parents of children with innocent murmurs that their children’s hearts are normal.
Results from a study called “Artificial Intelligence-Assisted Auscultation of Heart Murmurs: Validation by Virtual Clinical Trial” have been published in the peer-reviewed journal of Pediatric Cardiology. This is the first reported comprehensive and objective evaluation of an AI-based murmur detection algorithm, and included 3180 heart sound recordings from 603 outpatient visits. Algorithm estimation of heart rate was similar to gold standard. Sensitivity and specificity for detection of pathologic cases were 93% (CI 90–95%) and 81% (CI 75–85%), respectively, with accuracy 88% (CI 85–91%). In conclusion, the eMurmur algorithm tested has high sensitivity and specificity for detection of pathologic murmurs, similar to levels reported for specialist auscultation, making it a potentially useful screening tool for heart disease.
Dr. Lillian Lai of Children’s Hospital Eastern Ottawa presented a study today titled “Performance Evaluation of Two Heart Murmur Detection Applications by Prospective Clinical Trial” at the AHA Scientific Session 2017 today.
The study included 93 patients and demonstrated a sensitivity and specificity for automated detection of pathologic murmurs of 84.4% and 85.9%, respectively, with the gold standard being echocardiography, with eMurmur significantly outperform the other device.
In the AHA Scientific Session “Best Abstracts in Health Tech ”, Dr. W. Reid Thompson, associate professor of pediatrics at Johns Hopkins University, today presented a study titled “Validation of a Heart Murmur Algorithm by Virtual Clinical Trial” which used the eMurmur algorithm to blindly analyze 3,180 heart sound recordings from the Johns Hopkins Cardiac Auscultatory Recording Database (CARD). Study results showed a sensitivity of 93% and a specificity of 81%, as well as an accuracy of 88%.
eMurmur announces completion of a clinical trial evaluating the performance of its heart murmur detection algorithms on an elderly population with congenital heart disease. The algorithm’s sensitivity for autonomous detection of pathologic murmurs was 89%. No high severity cases were missed by the algorithm. Further, it yielded accurate heart rate estimation and S1/S2 detection, despite the presence of significant environmental noise.
eMurmur announces impressive pilot study of its AI platform on 106 children at Children’s Hospital of Eastern Ontario, Canada, where it differentiated pathologic from innocent heart murmurs with high sensitivity (87%) and specificity (100%), a positive predictive value of 100%, a negative predictive value of 90%, and high accuracy (94%) when compared with echocardiography as the gold standard for diagnosing murmurs.
Sign up for our newsletter!