Researchers in Japan have developed a flexible sensor patch that could transform healthcare by monitoring and analyzing vital health data right from the surface of the skin. Developed by a team at Hokkaido University, the patch uses a type of processing known as “edge computing,” which eliminates the need for cloud-based data processing. This compact, flexible device can detect early signs of health issues like arrhythmia, coughing, falls, and even heat stress.
The modern era of health monitoring
Wearable health monitoring has advanced tremendously, with over half a billion people worldwide wearing things like smartwatches or fitness trackers. These wearable devices track things like heart rate, steps, and sleep patterns. But usually, they rely heavily on cloud computing, where data must be sent to remote servers for processing. This results in delayed feedback and potential privacy risks. Additionally, conventional wearable devices are often limited to monitoring a limited range of metrics, often with questionable accuracy.
The new sensor takes a different approach by using edge computing.
Edge computing is a processing approach where data is analyzed locally, on or near the device where it’s generated, rather than being sent to a centralized server or cloud for processing. In wearable health technology, for example, edge computing allows devices like smart patches or watches to analyze health data directly on the device or on a nearby smartphone. This local processing reduces response time, since data doesn’t have to travel far, and enhances privacy by keeping sensitive information closer to the user.
“Our goal in this study was to design a multimodal sensor patch that could process and interpret data using edge computing, and detect early stages of disease during daily life,” explains study author Professor Kuniharu Takei at Hokkaido University.
The sensor’s data is transmitted via Bluetooth to a smartphone where it’s processed in real-time. Initial trials demonstrated its ability to detect abnormal heart rhythms, coughing, and even falls — key indicators that can signal underlying health issues. The success of this device opens the door for wide-ranging applications in personal health monitoring, particularly for early detection of disease.
Durable design with advanced sensors
Creating a wearable sensor patch that could monitor several vital signs without compromising comfort or durability is an impressive engineering feat.
The patch essentially integrated different sensors, and each sensor uses different materials, Takei tells ZME Science. This means you have to deal with the different temperatures and chemical environments in which the sensors work — and then make it so that it can be integrated on a wearable, flexible patch.
<!– Tag ID: zmescience_300x250_InContent_3
–>
“Because of this, to integrate different sensors, we need to carefully consider the fabrication process. In addition, importantly, developing different sensors on a flexible film is another challenge,” the researcher says.
The device was initially tested on three volunteers who wore the patch on their chests while the sensors monitored their vital signs in varying conditions. Even when exposed to high temperatures (like you’d have in a heatwave), the sensors could detect changes in vital signs. This achievement, though preliminary, shows that the patch can help to identify early-stage health conditions that might otherwise go unnoticed.
“By monitoring continuous vital change, we can most likely find small changes corresponding to the beginning of a disease, which we cannot often feel by ourselves. By analyzing this small change, we believe that we can find the early-stage disease. This continuous and automatic vital monitoring is very important for future medical and healthcare fields,” the researcher explained in an email.
Hardware and software
In addition to designing the physical sensors, the team developed a machine-learning algorithm to process the recorded data. With this, they could identify patterns and anomalies in the readings, crucial for detecting specific health events like arrhythmia or a sudden fall.
The device can detect even subtle changes that could allow the patch to identify illnesses like respiratory infections, early cardiovascular issues, or even stress levels before they become symptomatic.
With further testing, the researchers hope to identify more diseases and refine the device’s accuracy in these early detections. “We need physiological tests with a much larger group of volunteers — over 100 — to confirm that we can reliably identify early-stage diseases or other abnormalities,” Takei told ZME Science, underlining the ongoing research and development efforts.
Associate Professor Kohei Nakajima of The University of Tokyo, a co-researcher, emphasized the value of edge computing.
Processing data on the device itself rather than transferring it to a cloud server provides a dual advantage. It minimizes data lag, making the sensor patch’s insights almost instantaneous, and it enhances privacy by keeping sensitive health data on the user’s device. However, the machine learning model still requires initial training on a computer. This could be solved in future versions with a more streamlined data processing model that allows all computations to be completed directly on the smartphone.
Working with hospitals
The research team has ambitious goals for this technology. They plan to extend its detection capabilities to monitor other diseases and improve the overall accuracy of the device. To do this, they are collaborating with medical professionals and institutions, aiming to tailor the sensor patch to specific health conditions.
“We’re currently working with medical doctors and hospitals to explore monitoring a range of symptoms,” Takei explained. While details are not yet public, the team is focused on diseases that could potentially be diagnosed early through continuous vital monitoring.
Another potential upgrade for the device is a more user-friendly interface and system. Since the patch relies on a smartphone for data processing and alerts, a well-designed app interface could further streamline the experience, making it easier for users to understand their health data and respond to any anomalies. This could be particularly useful for elderly or less tech-savvy users who may benefit most from the patch’s real-time health monitoring.
For now, the researchers are focused on expanding their datasets, testing the patch in various real-world settings, and refining their machine learning models. They are optimistic that, with larger-scale testing, the sensor patch will eventually reach the point where it can be used in routine healthcare settings or even as a personal health monitoring tool at home.
Journal Reference: Real-time personal healthcare data analysis using edge computing for multimodal wearable sensors. 10.1016/j.device.2024.100597