Chi Hwan Lee and colleagues have developed a smart neckband that tracks eating and drinking habits. This machine learning-powered device uses sensors to distinguish between these behaviors and similar behaviors such as talking and walking, offering the potential to manage diabetes and improve overall health. .The study was published in the journal PNAS Nexus.
Smart neckband wearers can monitor the amount of food they consume. If you want to maximize your fitness or manage diseases like diabetes or obesity, automatically tracking your food and hydration intake can be helpful. However, eating and drinking must be distinguished from related activities such as talking and walking via wearable technology.
Chi Hwan Lee and colleagues propose a neckband with machine learning capabilities that can distinguish between voice, body language, hydration, and food intake. The sensor module in the neckband includes a microphone, 3-axis accelerometer, and surface electromyography sensor.
Together, these sensors can record auditory data, body movements, and muscle activity patterns in the thyrohyoid muscles of the neck. Machine learning algorithms accurately identified which behaviors were associated with eating and drinking in a study of six participants with nearly 96% accuracy for individual activities and 89% accuracy for concurrent activities .
The neckband is made with a breathable, mesh construction, flexible and twistable fabric, features 47 active and passive components, and can operate on battery power for over 18 hours without recharging .
Neckbands are useful for athletes and others interested in improving their general health. It can also be used in closed-loop systems with continuous blood glucose meters and insulin pumps to help diabetics decide when to take their insulin doses.
Reference magazines:
Park, T. other. (2024) Machine learning-enabled smart his neckband for monitoring dietary intake. PNAS Nexus. doi.org/10.1093/pnasnexus/pgae156.
Source: https://academic.oup.com/pnasnexus
