IM007

The IM007 system is a software medical device utilizing a proprietary machine learning algorithm to analyze ECG data from Insertable Cardiac Monitors for the detection and classification of various cardiac arrhythmias. It functions as an API-driven, server-side service that integrates with remote monitoring platforms to provide healthcare professionals with adjunctive diagnostic information.

IM007 is a software medical device designed for qualified healthcare professionals to aid in the assessment of arrhythmias by analyzing electrocardiogram (ECG) data obtained from Insertable Cardiac Monitors (ICMs). The system, developed by Implicity Inc., initially supports ECG data from Medtronic ICMs and is intended to be electronically interfaced with other computer systems, specifically remote monitoring platforms. These platforms supply the ECG data to IM007 and receive its analysis for viewing by healthcare professionals. IM007 itself does not include a user interface for viewing ECG data or a database to store results, as these functionalities are expected to be part of the interfacing 'client' application. The core of IM007 is a proprietary machine learning algorithm that processes ECG signals to detect and classify various cardiac events, including asystole, bradycardia, atrial tachycardia or atrial fibrillation, ventricular tachycardia, normal rhythm, and artifact. The system operates as a server-side service with an Application Programming Interface (API) consisting of input and output messaging queues. It receives ECG files and device parameters from the remote monitoring platform, the algorithm analyzes the data, and the output (formatted as a sequence of waveform labels with start and end times) is sent back. The interpretation results from IM007 are intended for advisory use only, to be considered by physicians in conjunction with their clinical knowledge and other diagnostic information, and it is not for use in life-supporting or sustaining systems. Non-clinical performance testing using ANSI/AAMI EC57:2012 standard databases and proprietary databases confirmed its ability to perform as intended.

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 are the specific performance metrics (sensitivity, specificity, positive predictive value, negative predictive value) of the IM007 algorithm for each type of arrhythmia it detects (asystole, bradycardia, AT/AF, VT) when benchmarked against standardized annotated ECG databases from ICMs?; How does the IM007 algorithm differentiate between true cardiac events and artifacts, particularly those specific to subcutaneous ECG signals from ICMs, and what is its false positive rate for critical arrhythmias like VT or asystole?; What are the technical requirements and data exchange protocols for integrating the IM007 API with various remote monitoring platforms beyond the initially supported Medtronic ICMs?; Could you elaborate on the types of ECG patterns or signal characteristics that would lead to an 'Unspecified abnormality' classification, and what is the recommended clinical workflow when such a classification is provided?; Given that IM007 relies on a 'client' application for ECG data viewing, what are the minimum data visualization and annotation capabilities recommended or required for the interfacing remote monitoring platform to ensure effective clinical use of IM007's analysis results?