Signal quality and measurement confidence

‍

Obtaining a reliable physiological signal depends on many factors: lighting, movement, face positioning, and camera conditions all influence measurement quality.

Rather than waiting until the end of the scan to determine whether conditions were suitable, Shen AI continuously evaluates the environment and provides feedback throughout the measurement. This helps users improve their measurements in real time while giving your application additional information about the confidence of each result.

‍

Before the measurement

Before the scan begins, the SDK analyzes the camera view and checks whether the environment is suitable for measurement.

If conditions aren't optimal, users receive specific guidance based on what the SDK detects. Depending on the situation, they may be asked to:

- move into better lighting,
- avoid strong backlight,
- uncover their forehead,
- remove glasses,
- move closer to or farther from the camera.

This guidance updates in real time as users adjust their position or environment. The goal is to help users start the measurement under the best possible conditions, improving both the experience and the likelihood of obtaining reliable results.

‍

During the measurement

Once the scan starts, Shen AI continuously evaluates the quality of the captured physiological signal. If you're using the built-in interface, users see this as the familiar quality indicator during the measurement.

At the same time, the SDK exposes a Signal Quality value through the API. This is a floating-point value between 0 and 1, where higher values indicate greater confidence in the captured signal.

The value follows the same logic as the on-screen quality indicator and updates continuously as measurement conditions change.

When Blood Pressure is included in the measurement, the Signal Quality value reflects environmental conditions specifically relevant to blood pressure estimation. If Blood Pressure is not selected, the value reflects the requirements for the selected metrics, such as Heart Rate or Heart Rate Variability (HRV).

Because blood pressure estimation requires stricter measurement conditions, it's normal for the Signal Quality value to be lower when Blood Pressure is enabled.

‍

Understanding what affects quality

The SDK evaluates several environmental factors while the measurement is running. These include the quality and consistency of the physiological signal, camera characteristics, face positioning, forehead visibility, and lighting conditions.

Together, these factors determine how confident the SDK is in the captured signal and help explain why measurement quality improves or decreases as conditions change. By identifying specific environmental issues rather than simply reporting a low-quality scan, Shen AI helps users build better measurement habits over time.

‍

After the measurement

Alongside the physiological measurements, Shen AI can provide context about the overall quality of the session and identify the single most important factor that could improve the next measurement.

For example, users may be advised to improve lighting, keep the device steadier, or ensure their forehead remains fully visible. Presenting one clear recommendation instead of multiple warnings makes it easier for users to improve future measurements without overwhelming them.

‍

Using Signal Quality in your application

The Signal Quality value can be useful beyond the built-in user interface.

You can use it to:

- display additional confidence information alongside results,
- decide whether a measurement should be repeated,
- filter low-quality measurements before further analysis,
- monitor measurement quality across users or devices,
- build analytics dashboards that track environmental conditions over time.

The appropriate threshold depends on your use case and the balance you want to achieve between measurement quality and user experience.

‍