A research group from Hokkaido University, the University of Tokyo, Juntendo University, and Osaka Metropolitan University has developed a bandage-type sensor patch that can continuously measure multiple vital signs at all times.
In recent years, watch-type wearable devices capable of measuring heart rate and activity have become popular. However, these devices prioritize comfort, but the sensor does not fit the body closely enough, making them unsuitable for stable and highly accurate measurement of various vital signs from the skin surface.
In this study, we aimed to develop a wearable device that can be attached to the skin without the user feeling a foreign body sensation and can measure various vital information with high accuracy without the user even realizing it. First, we developed a flexible sensor device that can be attached to the skin like a bandage by integrating sensors for electrocardiogram, skin temperature, respiration, and skin humidity on a soft polyester film. We equipped this sensor patch with a wireless system, transmitted vital signals to a smartphone using Bluetooth, and developed an algorithm to instantly analyze the vital data. Here, we applied reservoir computing, a type of machine learning, to achieve high-speed real-time analysis of vital information that changes every moment. For example, by analyzing noise in an electrocardiogram with a reservoir computer, we succeeded in determining whether the noise was caused by coughing, arrhythmia, or body movement with an accuracy rate of around 80%.
By implementing all of these algorithms within a smartphone, an "edge AI system" has been created for the first time, capable of wirelessly measuring vital signs, analyzing the data, and displaying the results at all times, even in environments without an Internet connection.
The company says it has currently begun demonstration tests for a variety of illnesses using this sensor patch, and as its applications progress, it is expected that this will lead to the early detection of pre-illness through remote diagnosis, and a significant reduction in the number of people dying alone through remote monitoring.
Paper information:【Device】Real-time personal healthcare data analysis using edge computing for multimodal wearable sensors
