HeartBeam has announced new data demonstrating the effectiveness of its AI combined with vectorcardiography (VCG) in detecting atrial flutter. The study showed that HeartBeam's AI outperformed an expert panel of heart rhythm cardiologists. The AI algorithm, which uses deep learning, was more accurate in identifying heart rhythm abnormalities when compared to traditional methods. Specifically, HeartBeam's AI with VCG showed a 40% improvement over cardiologists in detecting atrial flutter using single-lead ECGs and a 6% improvement using 12-lead ECGs, with zero variability in detection compared to the significant interobserver variability among the cardiologists.
Presented by Dr. Joshua Lampert at the Heart Rhythm Society 2024 in Boston, the study involved analyzing 173 sets of VCGs, single-lead ECGs, and 12-lead ECGs. The results highlighted that HeartBeam's AI combined with VCG was not only highly effective but also provided consistent diagnostic accuracy, outperforming the panel of electrophysiologists in distinguishing atrial flutter from sinus rhythm. This was particularly notable given the variability in diagnoses among the human experts when using single-lead ECGs.
HeartBeam's core technology, vectorelectrocardiography (3D VECG), captures the heart’s electrical activity in three dimensions and synthesizes a 12-lead ECG. The company plans to integrate this technology into the AIMIGo device, a portable, credit card-sized tool for home use, which is currently under FDA review. CEO Branislav Vajdic expressed optimism about the AI program’s potential to enhance cardiac care management outside medical facilities, aiming for commercialization later this year.
Click here to read the original news story.