Wearable health technology is becoming smarter and more powerful. Smartwatches, fitness bands and medical patches can already track important health signals such as heart rate, muscle activity and blood flow. However, there is still a major problem: different body signals usually require different sensors.
This means that a device designed to monitor several health signals may need multiple sensors, separate electronic circuits and more space on the skin. The result can be a larger, less comfortable device that consumes more power.
Researchers at the National University of Singapore (NUS) have developed a new technology that could change this approach.
Led by Assistant Professor Liu Yuxin from the Department of Biomedical Engineering at the NUS College of Design and Engineering, the research team has created a cross-modal skin sensor called X-Sig. The device can combine electrical and mechanical signals from the body into a single signal, allowing different types of health information to be collected through one sensing system.
The research was published in Nature Sensors.
A new approach to wearable health monitoring
The human body produces many different signals. Electrical signals include the electrical activity of the heart and the impulses that control muscles. Mechanical signals include the pressure waves created by blood flow and the physical forces produced when muscles contract.
Traditional wearable devices generally measure these signals separately.
For example, electrodes are used to measure an electrocardiogram (ECG), which records the electrical activity of the heart. Similar technology can measure electromyography (EMG), which detects electrical activity associated with muscle movement.
Mechanical signals require different types of sensors. These can measure signals such as radial pulse waves or force myography (FMG), which detects pressure changes caused by muscle contractions.
Because these sensors work differently, conventional systems often require separate electronic circuits to process each type of signal.
X-Sig takes a different approach.
Instead of measuring electrical and mechanical signals separately and combining them later through additional electronics, the new sensor combines them directly at the skin.
Two sensing layers, one signal
The X-Sig sensor uses a special layered design.
One layer contains a conductive electrode that detects electrical signals from the body. Another layer contains an extremely thin piezoelectric material that produces an electrical voltage when mechanical pressure is applied.
Because both layers produce voltage signals, their outputs can naturally be combined into a single waveform.
This eliminates the need for separate processing circuits for every type of signal. In simple terms, the sensor does some of the signal-combining work before the information even reaches the electronics.
This could make future wearable health devices smaller, lighter and more energy efficient.
The researchers also addressed another common problem with wearable electrodes: balancing electrical performance with comfort and skin adhesion.
The X-Sig electrode is made using three specially engineered polyurethane layers. Each layer performs a different function. One provides elasticity, another helps the sensor stick to the skin, and the third provides electrical conductivity.
This layered approach allows the electrode to maintain strong contact with the skin while also offering low electrical resistance.
According to the researchers, the material performed better than commercial gel electrodes in both adhesion and electrical performance.
The sensor is also perforated with tiny openings. These openings allow the adhesive material to bond directly with the skin, helping the device remain attached even on curved parts of the body.
Importantly, the researchers reported no skin irritation after 24 hours of continuous use. The electrode materials can also be recycled by washing them sequentially with water and ethanol.
Tracking the heart and blood pressure with one patch
One of the most promising demonstrations involved placing X-Sig on the wrist, over the radial artery.
At this location, the sensor can simultaneously detect ECG signals and pulse waves.
The researchers then used machine learning to analyse the combined waveform. From it, the system extracted heart rate and pulse arrival time—the short delay between the heart's electrical activity and the arrival of the resulting pressure wave at the wrist.
These measurements were then used to estimate blood pressure without requiring a traditional inflatable cuff.
The results were highly promising.
Compared with a standard cuff-based blood pressure monitor, the predicted values showed mean differences of only 0.27 mm Hg for systolic blood pressure and 0.33 mm Hg for diastolic blood pressure.
The system achieved the highest accuracy grade, Class A, under the relevant Institute of Electrical and Electronics Engineers (IEEE) standard for cuffless blood pressure devices.
The sensor also detected expected changes in blood pressure when volunteers changed positions, held their breath, or performed short periods of exercise.
This could eventually contribute to more convenient systems for continuous blood pressure monitoring.
Better control of prosthetic limbs
The researchers also tested X-Sig for detecting hand movements.
For this experiment, the sensor was placed on the forearm to capture both EMG and FMG signals. These signals provide different information about muscle activity.
When the two signals were combined, the system could identify 10 different hand gestures with 96.4% accuracy.
That was significantly better than using either signal alone. EMG achieved 72.1% accuracy, while FMG achieved 82.9%.
The combination also reduced the amount of training data needed. The machine-learning model required only 70 training samples per gesture to achieve high accuracy.
This could be particularly useful for technologies such as prosthetic limbs, rehabilitation devices and human-machine interfaces, where accurately understanding muscle movements is important.
One sensor, four types of body signals
The researchers have gone even further. They demonstrated that X-Sig can capture four different types of signals from a single location: ECG, EMG, pulse waves and FMG.
That means one small patch could potentially provide information about heart activity, muscle activity, blood flow and muscle force at the same time.
The technology is still at the research stage, so further testing will be needed before it can become a widely used medical product. Real-world health monitoring also requires strong validation across larger and more diverse groups of people.
Nevertheless, the concept could address one of the biggest challenges facing wearable healthcare technology: how to collect more information without making devices more complicated.
A step toward simpler continuous health monitoring
The importance of X-Sig is not simply that it can measure more signals. Its biggest advantage may be that it can do so using a simpler hardware architecture.
By combining electrical and mechanical information directly at the sensor, researchers could reduce the number of components needed in future wearable devices. This could lower power consumption, reduce device size and make continuous monitoring more comfortable.
The potential applications are broad, ranging from cuffless blood pressure monitoring to prosthetic control and physical rehabilitation.
As wearable health technology moves beyond fitness tracking and toward continuous medical monitoring, devices will need to become both more capable and easier to wear.
X-Sig offers a possible solution to that challenge: instead of adding more sensors to collect more health information, combine multiple signals at the source.
If this approach can be successfully developed and validated for everyday use, a single lightweight patch could one day monitor several important functions of the human body continuously—making advanced health monitoring more comfortable, convenient and accessible outside hospitals and clinics.
Reference: Wu, X., Zhu, C., Zheng, L. et al. A cross-modal epidermal sensor enables single-channel fusion of biopotential and biomechanical signals. Nat. Sens. 1, 315–327 (2026). https://doi.org/10.1038/s44460-026-00044-0
Comments
Post a Comment