HeartKey® Rhythm

HeartKey® Rhythm is an advanced software library designed to assist healthcare professionals by automatically analyzing single-lead adult ECG data for cardiac arrhythmias. It utilizes a combination of rules-based calculations and AI/ML algorithms, integrated via an API into existing healthcare platforms, to perform ECG signal conditioning, beat detection, and rhythm classification.

HeartKey® Rhythm is a software library designed for qualified healthcare professionals to analyze single-lead electrocardiogram (ECG) data from adult patients. It is intended for integration via an Application Programming Interface (API) into various ECG management systems, analysis platforms, or remote patient monitoring platforms. The system processes ambulatory ECGs from devices like Holters, ECG patches, and event recorders, providing capabilities such as ECG signal conditioning to reduce noise, ECG signal quality assessment, QRS beat detection, and ventricular ectopic beat detection. Its primary function is to automatically analyze ECG data and aid in the assessment of cardiac arrhythmias, offering results on an advisory basis to supplement clinical judgment. The device operates using a combination of rules-based calculations and artificial intelligence/machine learning (AI/ML) technology to process and analyze ECG signals. Its features include a Signal Processing Engine for conditioning and beat detection, and an Arrhythmia Analysis module for identifying rhythms like Normal Sinus Rhythm, Atrial Fibrillation, Cardiac Pause, Tachycardia, and Bradycardia, as well as classifying beat types and providing heart rate. Non-clinical performance testing was conducted using a large proprietary dataset of over 1200 ECGs and standard databases, evaluating metrics such as sensitivity, specificity, and positive predictive values against clinically relevant acceptance criteria. This testing confirmed the software met its design requirements and demonstrated acceptable performance in ECG analysis, including successful cybersecurity assessments.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. What specific AI/ML algorithms are employed by HeartKey® Rhythm, and how were they trained and validated beyond the proprietary dataset mentioned?; How does the single-lead analysis of HeartKey® Rhythm compare in diagnostic accuracy for specific arrhythmias (e.g., Atrial Fibrillation, Ventricular Ectopics) against traditional multi-lead ECG analysis systems?; What are the defined thresholds and criteria for "ECG signal quality assessment," and how does the system handle segments classified as "Unreadable" or "Inconclusive"?; Can the HeartKey® Rhythm API be customized by integrators to prioritize or flag certain types of arrhythmias based on specific clinical needs or patient populations?; What were the specific sensitivity, specificity, and positive predictive values achieved during the non-clinical testing for each of the detectable arrhythmias (Normal Sinus Rhythm, Atrial Fibrillation, Cardiac Pause, Tachycardia, Bradycardia)?