US EditionApple AI Model Flags Health Conditions with up to 92% Accuracy
NO LOC, JUL 11 – The Wearable Behavior Model analyzes patterns from over 160,000 participants to improve detection of static and dynamic health conditions, achieving up to 92% accuracy for pregnancy detection.
- Apple created the Wearable Behavior Model , an AI system developed using extensive data gathered from Apple Watch and iPhone users, which can predict various health conditions with accuracy reaching as high as 92%.
- Apple facilitated this research by gathering data from over 161,000 individuals participating in a large-scale study that leverages Apple Watch and iPhone sensor information to enhance health prediction models beyond basic sensor measurements.
- The Wearable Behavior Model leverages aggregated behavioral indicators derived from Apple Watch data—such as daily movement, sleep patterns, heart rhythm fluctuations, and overall physical activity—to identify both persistent and changing health conditions, demonstrating superior performance compared to models relying solely on raw sensor inputs.
- The hybrid model that integrated behavioral and biometric data reached 92% accuracy in detecting pregnancy and demonstrated notable improvements across various health-related tasks, including those involving sleep quality, respiratory infections, injuries, and cardiovascular conditions such as atrial fibrillation.
- This advancement enhances Apple's AI health technology while adhering to its 2030 carbon-neutral goal by running AI processes on 100% renewable energy at all third-party computing sites.
14 Articles
14 Articles
der Standard AT
der Standard DEThe new model should take into account not only cardiovascular data, but behavioural habits of users. However, AI has a bias problem.
Can a smart watch predict if a person is pregnant, has a respiratory infection, or sleeps badly? According to a new study supported by Apple, the answer is yes. And it does so thanks to an artificial intelligence model that is not based so much on traditional sensors (such as those that measure the pulse), but on how we move, how much we sleep, how much energy we spend or how fast we walk.The system is called WBM (Wearable Behavior Model) and wa…
Is Apple Building an AI Doctor? A New Study Says Maybe
Reports of Apple’s work on a new AI-powered health service have been heating up over the past several months. While nobody expects the project to be ready until next year at the earliest, a new Apple-supported health study is providing some insights into ways that AI models can analyze data from the Apple Watch and […]Read More...
Coverage Details
Bias Distribution
- 50% of the sources lean Left, 50% of the sources are Center
To view factuality data please Upgrade to Premium
