Apple is working on technology to detect a user's blood pressure using neural networks and seismocardiogram data, negating the need for a blood pressure cuff.
According to a patent application published Thursday, a wearable device, potentially paired with Apple Watch, might one day be capable of monitoring a user's blood pressure without the need for any additional peripherals.
The patent application, titled "Interpretable neural networks for cuffless blood pressure estimation," explores the use of neural networks to estimate blood pressure using seismocardiogram data. Unlike an electrocardiogram, which relies on electrical signals to monitor a heart rate, a seismocardiogram measures the micro-vibrations produced by the heart beating.
There exist seismocardiogram devices small enough to be considered wearable, though these sensitive systems are typically placed over a user's sternum to measure minute vibrations in or near the heart. How Apple intends to deploy such a device was not discussed in the patent filing.
According to Apple's patent, the system would work by leveraging an individually pruned neutral network accepting a seismocardiogram as input. That neural network would then use the data to estimate blood pressure.
For example, Apple could create a baseline model by training the neural networks with seismocardiogram data and blood pressure measurements from a group of subjects. Using that as a baseline, the system could then prune the model for subsequent users. The initial dataset would be collected from users performing different sedentary activities.
From there, Apple contends that the neural network could be used to accurately estimate a user's blood pressure without a cuff. As the patent points out, the system could use the neutral network "to determine blood pressure based on accelerometer or gyroscope data," which are already present in Apple devices.
Current versions of the Apple Watch can be used to monitor and analyze different types of health data, including heart rates with an optical sensing system, heart rhythms via a built-in electrocardiogram or blood oxygen sensing on the Apple Watch Series 6. Since the Apple Watch release, Apple has been continually expanding the device's health toolkit, and Apple may be looking to build out a set of health-minded accessories.
The patent lists Siddharth Khullar, Nicholas E. Apostoloff, and Amruta Pai as its inventors. Among them, Apostoloff has been named on a previous patent dedicated to facial analysis and emotion detection.
This isn't Apple's first patent focused on blood pressure monitoring. In 2020, the company filed a patent application for an Apple Watch-supported system that could use pulse transit time to measure blood pressure. Another system could use pressure sensor data, and Apple has explored how to monitor blood pressure without a cuff in the past.
According to a patent application published Thursday, a wearable device, potentially paired with Apple Watch, might one day be capable of monitoring a user's blood pressure without the need for any additional peripherals.
The patent application, titled "Interpretable neural networks for cuffless blood pressure estimation," explores the use of neural networks to estimate blood pressure using seismocardiogram data. Unlike an electrocardiogram, which relies on electrical signals to monitor a heart rate, a seismocardiogram measures the micro-vibrations produced by the heart beating.
There exist seismocardiogram devices small enough to be considered wearable, though these sensitive systems are typically placed over a user's sternum to measure minute vibrations in or near the heart. How Apple intends to deploy such a device was not discussed in the patent filing.
According to Apple's patent, the system would work by leveraging an individually pruned neutral network accepting a seismocardiogram as input. That neural network would then use the data to estimate blood pressure.
For example, Apple could create a baseline model by training the neural networks with seismocardiogram data and blood pressure measurements from a group of subjects. Using that as a baseline, the system could then prune the model for subsequent users. The initial dataset would be collected from users performing different sedentary activities.
From there, Apple contends that the neural network could be used to accurately estimate a user's blood pressure without a cuff. As the patent points out, the system could use the neutral network "to determine blood pressure based on accelerometer or gyroscope data," which are already present in Apple devices.
Current versions of the Apple Watch can be used to monitor and analyze different types of health data, including heart rates with an optical sensing system, heart rhythms via a built-in electrocardiogram or blood oxygen sensing on the Apple Watch Series 6. Since the Apple Watch release, Apple has been continually expanding the device's health toolkit, and Apple may be looking to build out a set of health-minded accessories.
The patent lists Siddharth Khullar, Nicholas E. Apostoloff, and Amruta Pai as its inventors. Among them, Apostoloff has been named on a previous patent dedicated to facial analysis and emotion detection.
This isn't Apple's first patent focused on blood pressure monitoring. In 2020, the company filed a patent application for an Apple Watch-supported system that could use pulse transit time to measure blood pressure. Another system could use pressure sensor data, and Apple has explored how to monitor blood pressure without a cuff in the past.
RSS Feed
