US20250209627
2025-06-26
Physics
G06T7/0014
The patent application describes a non-contact, non-invasive health monitoring device that uses advanced AI and machine learning techniques to estimate vital signs and wellness metrics. The system captures real-time image data of a user's face with a high-resolution camera and processes this data to extract physiological signals like Photoplethysmography (PPG) and Ballistocardiography (BCG). These signals are used to predict vital signs such as heart rate, respiratory rate, blood pressure, and oxygen saturation, as well as wellness metrics like stress levels and metabolic health.
This technology is applicable in various fields including health assessment, fitness monitoring, telehealth, public health screening, and automotive safety systems. It involves analyzing physiological signals using deep learning techniques to provide accurate health metrics without physical contact. The device's ability to function in diverse environments makes it especially useful in situations where traditional contact-based monitoring is impractical.
Traditionally, health monitoring required physical contact through sensors or direct measurements, which could be uncomfortable or inaccessible in certain contexts. Recent advancements in AI have enabled the development of non-contact methods that overcome these limitations. By analyzing facial image data, such devices can continuously and unobtrusively provide critical health information. PPG and BCG signals are key components in this process, capturing changes in blood volume and micro-movements associated with the heartbeat.
The device utilizes deep learning models like Convolutional Neural Networks (CNNs) and Transformers to extract meaningful information from PPG and BCG signals. CNNs capture spatial features while Transformers analyze temporal patterns, allowing for robust predictions of vital signs. Additionally, wellness parameters can be estimated using a correlational biocomputational scoring system derived from extensive datasets. This integration enhances the device's accuracy, achieving error margins below 5%.
This technology has significant implications across various industries. In the automotive sector, it can enhance driver safety by monitoring alertness and stress levels. In public health, it facilitates rapid screenings in high-traffic areas like airports and workplaces. For fitness enthusiasts, it provides continuous tracking of physiological responses during exercise. Telehealth applications benefit from the device's ability to perform remote health assessments, particularly useful for patients in underserved areas.