(117 days)
The ViSI Mobile Monitoring System is intended for use by clinicians and medically qualified personnel for single or multi-parameter vital signs monitoring of adult patients (18 years or older). It is indicated for ECG (3 or 5 lead-wire), respiration rate (RESP), heart rate (HR), noninvasive blood pressure (NIBP), continuous noninvasive blood pressure (cNIBP), noninvasive monitoring of functional oxygen saturation of arterial hemoglobin (SpO2), pulse rate (PR), skin temperature (TEMP), posture tracking and basic arthythmia (Ventricular Tachycardia, Ventricular Fibrillation, Asystole, Atrial Fibrillation) analysis and alarm in hospital-based facilities including general medical-surgical floors, intermediate care floors, and emergency departments.
Continuous non-invasive blood pressure (cNIBP) measurements have not been evaluated on patients during ambulation. The basic arrhythmia analysis feature is intended for use on patients 18 years of age and older; it has not been evaluated on pediatric patients.
The arrhythmia analysis feature is intended for use by healthcare professionals trained in the identification and treatment of arrhythmia events. Automated arrhythmia analysis is an adjunct to clinical assessment; clinician review of the analysis should precede any therapeutic intervention.
The ViSi Mobile Monitoring System may be used as standalone devices or networked to ViSi Mobile Remote Viewers through wireless 802.11 communication.
The Visl Mobile Insight is an optional secondary notification system that communicates alarms directly to an assigned caregiver. It is intended to supplement the primary alarming devices which originate in the ViSI Mobile patient-worn device.
The ViSi Mobile Monitoring System is a patient worn, portable, battery operated, continuous physiological monitoring device intended for the monitoring of ECG (3 or 5 lead-wire), respiration rate (RESP), heart rate (HR), non-invasive blood pressure (NIBP), continuous non-invasive blood pressure (cNIBP), non-invasive monitoring of functional oxygen saturation of arterial hemoglobin (SpO2), pulse rate (PR), skin temperature (SKIN TEMP), posture tracking and alarms, basic arrhythmia analysis (ventricular fibrillation, ventricular tachycardia, asystole, atrial fibrillation) and alarms.
The ViSi Mobile Monitoring System consists of the patient worn devices, disposables, backup battery, patient data server and remote viewer.
This document (K180472) is a 510(k) premarket notification for the ViSi Mobile Monitoring System. The focus of the provided text is to demonstrate substantial equivalence to a predicate device, primarily by highlighting changes and comparing technical characteristics and performance, especially concerning arrhythmia detection.
Based on the provided text, here's a detailed breakdown of the acceptance criteria and the study proving the device meets them:
1. A table of acceptance criteria and the reported device performance
The document does not explicitly present a "table of acceptance criteria" in the traditional sense for the acceptance of the device as a whole. Instead, it presents performance metrics against recognized standards (ANSI/AAMI EC57: 2012) for its arrhythmia detection algorithm, comparing the improved "Proposed Device" (Subject Device) with a "Reference Device A" (K142827), which is a previously cleared version of the ViSi Mobile Monitoring System.
Here's a table derived from the provided "C. Atrial Fibrillation Comparison" and "E. ALARM TEST COMPARISON" sections, focusing on the arrhythmia detection performance:
| Metric | Acceptance Criteria (Implicitly, the performance of the Reference Device K142827 or general industry standards like ANSI/AAMI EC57:2012) | Reported Device Performance (Proposed Device K180472) | Commentary (from document) |
|---|---|---|---|
| AF Episode Sensitivity (%) | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | 0‡ | - | Null result (test done, but statistic cannot be calculated due to absence of test or reference annotation) |
| MIT-BIH Gross | 90 | 87 | Same or Better |
| MIT-BIH Average | 92 | 87 | |
| AF Episode Positive Predictivity (%) | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | 0‡ | - | Null result |
| MIT-BIH Gross | 100 | 65 | Same or Better |
| MIT-BIH Average | 100 | 99 | |
| AF Duration Sensitivity (%) | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | ‡ | - | Null result |
| MIT-BIH Gross | 75 | 80 | Same or Better |
| MIT-BIH Average | 71 | 84 | |
| AF Duration Positive Predictivity (%) | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | 0‡ | 0 | Same or Better |
| MIT-BIH Gross | 0 | 75 | |
| MIT-BIH Average | 0 | 40 | |
| AF False Positive Report | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | 1‡ | 0 | Same or Better |
| MIT-BIH Gross | 0 | 137 | |
| MIT-BIH Average | 0 | ||
| AF False Negative Report | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | 0‡ | 0 | Same or Better |
| MT-BIH Gross | 0 | 10 | |
| MIT-BIH Average | 0 | ||
| AF Time to Detection (sec) | RA (Reference Device A - K142827): | S (Subject Device - K180472): | |
| NST | - | - | Null result |
| MIT-BIH Gross | 00:20.94 | 0:06.4 | Same or Better |
| MIT-BIH Average | |||
| VF/VT Sensitivity (%) | Reference Device A (K142827) | Subject Device (K180472) | |
| AHA | 90 | 91 | |
| MIT | 100 | 100 | |
| CU | 94 | 97 | |
| VF/VT Positive Predictivity (%) | Reference Device A (K142827) | Subject Device (K180472) | |
| AHA | 60 | 100 | |
| MIT | 67 | 100 | |
| CU | 46 | 94 | |
| AFIB Sensitivity (%) | Reference Device A (K142827) | Subject Device (K180472) | |
| MIT | 97 | 90 | |
| AFIB Positive Predictivity (%) | Reference Device A (K142827) | Subject Device (K180472) | |
| MIT | 65 | 100 |
The study that proves the device meets the acceptance criteria is primarily non-clinical, using established databases.
2. Sample size used for the test set and the data provenance
The test set utilizes publicly available, established databases:
- AHA (American Heart Association)
- MIT-BIH (Massachusetts Institute of Technology - Beth Israel Hospital Arrhythmia Database)
- CU (Creighton University Ventricular Tachyarrhythmia Database)
- NST
The specific sample sizes (e.g., number of ECG recordings or patients) from these databases are not explicitly stated in the provided text. The provenance is from these known databases, which are widely accepted for arrhythmia algorithm testing. The data within these databases is retrospective.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
The document states: "The arrhythmia analysis is based on the same technology as the currently cleared Monebo Technologies, Inc Arrhythmia library (K062282), which has been adapted for real time analysis in the ViSi System. The ViSi Monitoring System arrhythmia analysis has been validated by comparison to the AHA, MIT-BIH, CU, and NST databases as prescribed in ANSI/AAMI EC57: 2012."
The ground truth for these databases (AHA, MIT-BIH, CU, NST) was established by experts, typically cardiologists or electrophysiologists, during the creation of the databases. The specific number and qualifications of these experts are not detailed within this 510(k) document but are inherent to the accepted nature of these standard databases.
4. Adjudication method for the test set
The document does not describe an adjudication method for the test set data itself. The ground truth for the test sets (AHA, MIT-BIH, CU, NST databases) was established independently by the creators of those databases.
5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance
No. A multi-reader multi-case (MRMC) comparative effectiveness study involving human readers is not described in this document. The study focuses purely on the algorithm's performance against established gold-standard databases, comparing an updated algorithm to a previous version and recognized standards. The device is not presented as an AI-assistance tool for human readers but rather as a standalone monitoring system with an integrated arrhythmia detection algorithm.
6. If a standalone (i.e. algorithm only without human-in-the loop performance) was done
Yes. The performance evaluation presented is a standalone (algorithm only) performance assessment. The tables show the algorithm's sensitivity and positive predictivity for various arrhythmia events (VF/VT, AFIB) when processing the specified ECG databases. The context further reinforces this, stating "Automated arrhythmia analysis is an adjunct to clinical assessment; clinician review of the analysis should precede any therapeutic intervention," indicating it acts as an aid rather than requiring human modification of its output for the test.
7. The type of ground truth used
The ground truth used for the performance testing is expert consensus from established, recognized ECG databases (AHA, MIT-BIH, CU, NST). These databases contain expert-annotated ECG waveforms, which serve as the gold standard for evaluating arrhythmia detection algorithms.
8. The sample size for the training set
The document does not explicitly state the sample size of the training set used to develop or refine the arrhythmia detection algorithm. It mentions that "The arrhythmia analysis is based on the same technology as the currently cleared Monebo Technologies, Inc Arrhythmia library (K062282)," which implies that the training might have occurred prior to this submission (or the core technology was pre-trained).
9. How the ground truth for the training set was established
The document does not explicitly detail how the ground truth for the training set was established. Given that the algorithm is based on (or an adaptation of) the Monebo Technologies, Inc Arrhythmia library, it's highly probable that the training also utilized similar expert-annotated ECG datasets or proprietary clinical data. The document does state that the arrhythmia algorithm was "Rewritten to improve specificity and detection metrics" and later "New arrhythmia detection algorithm to improve arrhythmia detection accuracy and positive predictivity," implying development and testing that would have involved ground truth data.
§ 870.1025 Arrhythmia detector and alarm (including ST-segment measurement and alarm).
(a)
Identification. The arrhythmia detector and alarm device monitors an electrocardiogram and is designed to produce a visible or audible signal or alarm when atrial or ventricular arrhythmia, such as premature contraction or ventricular fibrillation, occurs.(b)
Classification. Class II (special controls). The guidance document entitled “Class II Special Controls Guidance Document: Arrhythmia Detector and Alarm” will serve as the special control. See § 870.1 for the availability of this guidance document.