Fitbit is planning to take its sleep monitoring features to the next level by introducing snoring detection and measurement.
According to a patent approved today, the company highlights processes and methods that can be used to detect when a user is snoring and provide measurement metrics related to it, including snoring intensity and corresponding respiratory metrics.
How would the snoring detection and monitoring work?
The central idea is to use a microphone port on the Fitbit to capture an audio sign while you are sleeping.
The Sleep monitoring algorithm analyzes the audio frame and generates sleep-related information, wherein the information identifies one or more sounds as potential sources of sleep disruption.
The algorithms are responsible for extracting a set of features based on the spectrum of frequencies represented by the audio frame; pass the frequencies to a machine learning model; and obtain information describing the one or more sounds from the machine learning model.
Furthermore, the sleep monitoring improvements tag the sleep stage details corresponding to the time of maximum snoring intensity.
It also tells in which sleeping position did you start snoring.
The sleep-related information generates a baseline noise level while the user was asleep, an average noise level while the user was asleep, the highest noise level, or a total count of sounds detected.
Existing PPG sensor data are used to co-relate the collected information in the form of audio frames.
- Verily Life sciences announces results of Baseline Sleep Mission Study
- Sleep Apnea? The new Withings ScanWatch promises to deliver where it matters most
- Targeted Dream Incubation (TDI) takes sleep tracking and dream research to the next level
How does it know it’s you and not the dog?
If you were wondering how the process distinguish the snoring sound between two different individuals in the room, the following describes their process of identifying the user with the snoring issue 🙂
“In various embodiments, measurements taken by each PPG sensor can be used to determine (or estimate) a respective breathing (or respiratory) phase for the user based on generally known techniques.
In such embodiments, an amount of correlation can be determined between the user’s breathing phase based on measurements captured by a PPG sensor and the breathing phase of the entity based on the audio data, as determined by the audio-based breathing determination module 304.
If a threshold correlation exists, the user of the wearable device can be determined to be the source of the snores captured in the audio data”
This patent, 20200383633, was originally filed on June 2nd 2020 and was approved today.
For users interested in Sleep monitoring and related respiratory metrics, the detailed Fitbit patent provides lot of interesting insights and is worth a read.
Fitbit is not alone to evaluate advances in the area of sleep monitoring. Withings has been working hard to get its FDA approval on the new watch that can detect Sleep apnea.
Other tech companies such as Google’s Verily have also been engaged in different aspects of sleep monitoring.