Search Results
Found 1 results
510(k) Data Aggregation
(107 days)
FaceHart Corp.
FH Vitals SDK-RR is a software-only respiratory rate measurement tool intended to be integrated into third-party software applications on compatible mobile devices, laptops, or computers. It is intended for spot checking of Respiration Rate (RR) in an automatic contactless manner by analyzing chest wall movement from video input when the subject is still and properly positioned in front of the camera in a well-lit environment.
FH Vitals SDK-RR is intended for use under the supervision of a healthcare professional, either in clinical or home environments, and is not intended for continuous monitoring, apnea detection, or as the sole method for evaluating physical health. It is intended to serve as a supplemental tool to assist in the overall assessment of the patient. The software is indicated for use on individuals aged 18 years and older who do not require critical care or continuous vital sign monitoring.
The FaceHeart Vitals Software Development Kit (FH Vitals SDK-RR) is a software product designed for integration into various software applications, such as those on mobile devices, laptops, or computers. This SDK is strictly a software-only product and is compatible with the following operating systems: Windows 10, Android 14, and iOS 17. It has been tested and is currently only compatible with the following cameras:
- Logitech C930 webcam
- Samsung S24+, S24+ smartphone front-facing cameras
- iPhone 15 Pro, and iPhone 15 Pro Max front-facing cameras
The SDK is non-contact and designed to measure the respiratory rate in breaths per minute based on facial video streams. When the SDK receives a video stream, it first identifies and tracks the face within the video, ensuring consistent positioning of the face. Based on the tracked facial region, the system identifies the chest as the region of interest (RoI) and locates reliable key points. The system then focuses on the movement of the chest within the RoI to further determine the number of respiratory cycles. By analyzing the frequency of chest elevation and depression, the SDK calculates the respiratory rate in breaths per minute.
FH Vitals SDK-RR is designed for non-invasive respiratory rate measurement with a validated measurement range of 5-36 bpm. The SDK is not intended for apnea monitoring or detecting the cessation of breathing. FH Vitals SDK-RR is a prescription-use software device that must be used under the clinical supervision or instruction of a licensed healthcare professional (HCP).
Here's a detailed breakdown of the acceptance criteria and the study proving the device meets them, based on the provided FDA 510(k) Clearance Letter and 510(k) Summary.
Acceptance Criteria and Device Performance
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Respiratory Rate Measurement Range | 5-36 bpm (Predicate: 7-30 bpm; Subject device: 5-36 bpm, validated through clinical testing, aligning with intended use population) |
| Performance (Error Level) | ±2 bpm (Predicate: ±3 bpm; Subject device: Root Mean Square Error (RMSE) for all tested devices was 1.2 bpm (95% CI), demonstrating deviations consistently within ±2 bpm when compared to the gold standard. p-values for RMSE > 2 bpm were below 0.001, indicating statistical significance.) |
| Measurement Window | 60 seconds (Same as predicate) |
| Performance Across Respiratory Rate Levels | RMSE ≤2 bpm across all RR ranges (5-8 bpm: RMSE 1.7-1.8 bpm; 9-11 bpm: RMSE 1.0-1.1 bpm; 12-19 bpm: RMSE 1.0-1.3 bpm; 20-29 bpm: RMSE 0.9-1.0 bpm; 30+ bpm: RMSE 1.1-1.2 bpm for all devices. All p-values 2 bpm) |
| Statistical Significance | p-values for RMSE > 2 bpm were below 0.001, confirming the results were statistically significant. |
| Consistency Across Devices | Coefficient of determination (R²) values exceeded 0.98 across all devices, indicating close alignment with the clinical reference device. |
| Performance Across Demographics | No significant differences in accuracy were found based on sex, age, or BMI, confirming consistent performance across demographic groups. |
| Performance Across Disease Subgroups | No major performance issues were noted in cardiovascular, neurological, or pulmonary patients, with minor underestimation in some conditions (e.g., Alzheimer's) still within acceptable limits. |
| Clothing Impact on Accuracy | Negligible correlation between clothing texture and accuracy, confirming robustness across different clothing types. |
| Performance Under Limiting Conditions (for conditions where motion is not intended use) | Motion: Up-and-down or back-and-forth motion increased errors; left-and-right motion had the least impact. (Device is not intended for motion conditions). |
| Behavior: Coughing caused significant errors; shallow breathing had minimal effect. | |
| User Pose: No difference in accuracy between sitting or lying positions. | |
| Illumination: Stable performance across various lighting conditions. | |
| Measurement Distance: Accuracy reliable within 1.5 meters but decreased beyond. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: 420 participants
- Data Provenance:
- Country of Origin: Taiwan (377 subjects) and United States (43 subjects)
- Retrospective or Prospective: Not explicitly stated as retrospective or prospective, but the description of "evaluation in a clinical study involving 420 participants" generally implies a prospective collection for the purpose of the study.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications of Those Experts
The document does not provide information regarding the number of experts used or their qualifications for establishing the ground truth. It states that the ground truth was established by a "gold standard etCO2 device (manually counted/annotated respiratory rate from analysis of waveform output from the Philips MX100 with Microstream CO2 Extension)." This suggests that human expert annotation may have been involved in the manual counting/annotation, but the specifics are not detailed.
4. Adjudication Method for the Test Set
The document does not specify an adjudication method (such as 2+1, 3+1, none) for the test set. The ground truth was derived from a direct comparison to a gold standard medical device.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and 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 was not done. The study's focus was on the standalone performance of the FH Vitals SDK-RR against a gold standard, not on the improvement of human readers with AI assistance.
6. If a Standalone (i.e., Algorithm Only Without Human-in-the-Loop Performance) Was Done
Yes, a standalone study was done. The entire clinical performance summary describes the FH Vitals SDK-RR (algorithm only) processing video input to measure respiratory rates and comparing these measurements directly against a gold standard device, without human intervention in the measurement process itself.
7. The Type of Ground Truth Used
The ground truth used was a gold standard etCO2 device, specifically the "manually counted/annotated respiratory rate from analysis of waveform output from the Philips MX100 with Microstream CO2 Extension." This is a form of physiological measurement data, potentially with human expert review/annotation of that waveform.
8. The Sample Size for the Training Set
The document does not provide any information about the sample size used for the training set. The descriptions are exclusively for the clinical performance study (test set).
9. How the Ground Truth for the Training Set Was Established
The document does not provide any information on how the ground truth for the training set was established, as the training set details are not included.
Ask a specific question about this device
Page 1 of 1