Demonstrates Potential for ECG screening for LV dysfunction in asymptomatic patients with ejection fraction > 40%
SOUTHLAKE, Texas, June 28, 2021 – HeartSciences, a medical device company focused on advancing the field of electrocardiology through innovation, announced the results of a study by Australia’s Baker Heart and Diabetes Institute focused on identifying asymptomatic heart failure patients using machine learning based ECG testing.
Increases in healthcare expenditures remains a global widespread issue. The overuse of expensive diagnostic testing has been identified as a contributing factor. More effective low-cost front-line tools are needed that help identify at-risk patients earlier while lowering costs for health systems around the world. Currently no viable screening test exists to improve referral for echocardiography, however recent studies have demonstrated the ability to detect LV dysfunction in patients with LVEF <= 40%. This study evaluated the feasibility of using ECG-based machine learning to effectively screen for patients at-risk for subclinical systolic and diastolic left ventricular dysfunction (LVD) and thus potentially reduce the number being referred on to echocardiography.
The results were published in JACC- Cardiovascular Imaging. The article by Elizabeth L. Potter, MBBS, BSc, Thomas H. Marwick, MBBS, PHD, MPH, Partho P. Sengupta, MD, et. al. is titled “Machine Learning of ECG Waveforms to Improve Selection for Testing for Asymptomatic Left Ventricular Dysfunction Prompt.” The machine learning ECG algorithm using MyoVista wavECG technology was developed by Potter et.al. to determine if it could effectively screen for LV dysfunction in asymptomatic patients with EF > 40%.
A supervised machine-learning approach was used to develop an algorithm using HeartSciences MyoVista® wavECG™ CWT frequency-based features. The ARIC HF score was also included to evaluate the importance of this easily attainable clinical variable against wavECG frequency related variables.
The study included 398 participants (57% female, median age 69 years), and of these, 171 (43%) had LVD. The developed algorithm using CWT features alone was tested in an independent group (n =111; 49% female, median age 61 years), and demonstrated 88% sensitivity and 70% specificity (area under the receiver-operating curve or AUC) of 0.78. The ARIC HF risk score had an AUC of 0.72 for LVD discrimination. An optimized cutoff point for sensitivity was identified as an ARIC HF risk score of 2.6, providing 90% sensitivity and 40% specificity. The AUC for NT-proBNP was 0.53, with an optimized cutoff of 21 pg/ml providing a sensitivity of 88% and specificity of 14%. An abnormal ECG, as determined by conventional ECG software-based analysis, provided a sensitivity of 36% and a specificity of 85%. In those with available NT-proBNP, adding NT-proBNP to the ARIC HF risk score (AUC 0.65) did not significantly improve discriminatory ability versus ARIC alone (AUC 0.63). Furthermore, the addition of both NT-proBNP and abnormal conventional ECG results did not significantly improve discriminatory ability (AUC 0.67)
Results of the study: “Conventional candidates for LVD screening (ARIC score, N-terminal pro–B-type natriuretic peptide, and standard automated ECG analysis) had inferior discriminative ability. Integration of ewECG [HeartSciences wavECG] in screening of people at risk of HF would reduce need for echocardiography by 45% while missing 12% of LVD cases. “
“This new study by Baker Institute using MyoVista’s wavECG Technology further demonstrates the ability of MyoVista wavECG Technology to detect LV dysfunction.” stated Mark Hilz, President and CEO of HeartSciences.
Andrew Simpson, Chairman of HeartSciences, stated “This study further indicates that HeartSciences MyoVista wavECG’s innovative technology has the potential to significantly improve low-cost front-line testing for heart disease”.
HeartSciences sits at the forefront of innovation and technological development focused on advancing the field of electrocardiology to provide early heart disease detection. Its first product, the MyoVista® Wavelet ECG (wavECG™) Cardiac Testing Device, is a resting 12-lead electrocardiograph that uses AI and continuous wavelet transform (CWT) signal processing to provide cardiac information associated with left ventricular diastolic dysfunction (LVDD), a condition which has previously not been possible to detect using conventional electrocardiology. LVDD is associated with almost all forms and co-morbidities of heart disease and may include hypertension, diabetes, valvular disease, ischemia, and reduced systolic function, among others.
The MyoVista Device additionally provides all the information and capabilities of a full-featured conventional resting 12-lead ECG within the same test and follows the same clinical AHA/IEC lead placement protocol.
HeartSciences is a privately held U.S. corporation based in Southlake, Texas.
The MyoVista Device is not currently FDA cleared and is not available in the United States.
For more information visit www.heartsciences.com.