Search Results

The PreemptiveAI Clinical SDK is intended for noninvasive spot measurement of pulse rate. The measurement is based on video recording of the subject's fingertip using a smartphone camera. It is designed for use when the subject is still. It is indicated for individuals 18 years of age or older not requiring critical care or continuous vital signs monitoring. It is not intended for use in patients with known or suspected heart arrhythmias.

The PreemptiveAI Clinical SDK is a Software as a Medical Device (SaMD) that utilizes a smartphone to noninvasively measure pulse rate (PR). Designed for integration into third-party mobile applications, the SDK operates on Android and iOS devices, enabling spot-check pulse rate measurements.

The PreemptiveAI Clinical SDK captures a 30-second fingertip video with the smartphone's back camera. The software extracts a photoplethysmography (PPG) waveform, partitions that 30-second signal into overlapping 10-second windows, and applies an autocorrelation-based signal quality check. A proprietary deep learning algorithm then calculates pulse rate for every window that passes the quality screen. The value shown to the user is the average of all passing windows, and if every window fails the screen, no rate is reported and a "Low Signal Quality" prompt is displayed.

The SDK outputs a single pulse-rate value; it is not intended for automated analysis or arrhythmia detection. The PreemptiveAI Clinical SDK is not intended for use with alarm systems.

Here's a breakdown of the acceptance criteria and study details based on the provided FDA 510(k) clearance letter for the PreemptiveAI Clinical SDK:

Acceptance Criteria and Device Performance

Study Details

1. Sample Size for Test Set and Data Provenance:

• Sample Size: 111 participants.
• Data Provenance: The document states participants were "recruited from a general population," implying the data is from a human clinical study. The country of origin is not explicitly mentioned but is likely the USA given the FDA clearance. The study appears to be prospective as it involved recruiting participants and obtaining new measurements.

2. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications:

• This information is not provided in the document. The document states "The pulse rate obtained by the PreemptiveAI Clinical SDK was compared to the heart rate obtained by the reference device." It doesn't specify how the ground truth from the reference device was established or if experts were involved in its interpretation for this study.

3. Adjudication Method for the Test Set:

• This information is not provided in the document. The reference to comparing the SDK's pulse rate to a "reference device" suggests a direct comparison rather than a human adjudication process for the AI's output against independent expert assessments.

4. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was Done:

• No, an MRMC comparative effectiveness study was not explicitly mentioned or implied. The study focused on the standalone performance of the AI device against a "reference device," not on how human readers improve with AI assistance.
• Effect Size of Human Improvement with AI: Not applicable, as an MRMC study was not conducted.

5. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was Done:

• Yes, a standalone performance study was conducted. The document states, "The pulse rate obtained by the PreemptiveAI Clinical SDK was compared to the heart rate obtained by the reference device," indicating that the SDK's output was directly evaluated without human intervention in the pulse rate determination process.

6. The Type of Ground Truth Used:

• The ground truth was established using a "reference device." While the specific type of reference device is not detailed (e.g., medical-grade ECG, oximeter, etc.), it served as the comparator for the SDK's pulse rate measurements. It is a form of outcomes data in the sense that it provides a measured, objective physiological parameter.

7. The Sample Size for the Training Set:

• This information is not provided in the document. The document describes the clinical study and bench testing, but not the training data specifics for the proprietary deep learning algorithm.

8. How the Ground Truth for the Training Set Was Established:

• This information is not provided in the document.

Summary of Missing Information:

The provided document, being an FDA clearance letter, focuses primarily on the successful outcome and key aspects of the regulatory submission. It does not delve into the detailed methodology of the AI model's development, including:

• Specifics about the reference device used for ground truth in the clinical study.
• The number and qualifications of experts involved in ground truth establishment (if any, beyond the reference device).
• Details on the training set (size, ground truth methods).
• An MRMC study was not stated, so no human-in-the-loop performance change was evaluated.

Ask a specific question about this device