Apple acquired Beddit, a Finnish technology company, in 2017. Since then, Apple has filed patents to improve the Beddit system (01, 02, 03 and more). Beyond the Bed device, Apple has launched a sleep app for Apple Watch and iPhone. On Thursday, the US Patent & Trademark Office published an Apple patent application titled “Sleep Staging using Machine Learning.” It’s a patent that medical professionals will appreciate.
One of the inventors who worked on this patent was Matt Bianchi MD PhD, a double resident in neurology and sleep medicine currently with Apple Health Technologies. Another registrant on the patent is Alexander Chan, a Ph.D. in medical engineering with a specialization in neuroscience who led the Health Technology/Data Science Algorithms team that develops machine learning and data processing algorithms. signal to extract health information from sensors for new health/wellness features in new and existing Apple products.
To understand a patient’s sleep patterns, doctors typically perform objective sleep staging by monitoring electroencephalographic (EEG) activity during sleep. An EEG is a test that detects electrical activity in the brain using electrodes attached to the scalp. A patient’s brain cells communicate using electrical impulses and are constantly active, even when the patient is sleeping. Because sleeping with electrodes attached to the scalp can be cumbersome, other sensors to monitor sleep patterns have been developed, such as wearable devices and in-bed sensors.
Wearable devices are usually worn on the wrist, legs, or chest and include motion sensors (eg, accelerometers) to track movement in those locations. In-bed sensors are typically placed under a bed sheet and include sensors that can track breathing and heart rate by measuring tiny bodily movements that occur when a user breathes or their heart beats. Sensor data can be entered into a sleep staging app installed on a smartphone or other device.
The sleep staging app calculates various sleep metrics, such as total sleep/wake time and sleep efficiency, which can be used to quantify sleep to help users improve the amount of sleep they are getting and to allow the sleep/alarm tracker app to coach users on how to get more sleep.
While Apple’s patent application supports “Beddit” sleep tracking, they add a new approach to this system that involves machine learning.
In one embodiment, a method includes: receiving, with at least one processor, sensor signals from a sensor, the sensor signals including at least motion signals and respiratory signals from a user; extracting, with the at least one processor, characteristics of the sensor signals; predicting, with a machine learning classifier, whether the user is asleep or awake based on the characteristics; and calculating, with the at least one processor, an awake or awake metric based on whether the user is predicted to be asleep or awake.
Machine learning is used to improve the prediction of sleep/wake states which can be used by a sleep/wake tracking application to generate a variety of sleep metrics which can be used to quantify sleep to help users to improve the amount of sleep they get and enable the sleep/wake tracker app to help users sleep better.
Apple’s patent FIG. 2 is a conceptual block diagram of a sleep/wake classification system that includes a machine learning classifier; FIG. 4 is a flowchart of a feature extraction process that uses a machine learning sequencer/classifier; FIG. 5 is a flowchart of a classification process for predicting sleep/wakefulness probabilities; FIG. 6 is a flowchart of a sleep/wake process.
Those working in related medical fields will appreciate seeing the details of Apple’s patent application US 20220386944 A1.
#Apple #patent #published #week #describes #machine #learning #system #tracking #sleep #stages