Researchers at the National University of Singapore (NUS)and A*STAR’s Institute of Materials Research and Engineering (IMRE) have developed a paper-like, battery-free, AI-enabled sensor that can assesses the status of wound healing in 15 minutes.
Advertisement
Currently, wound healing is typically examined visually by a clinician. Wound infections are mostly diagnosed via swabbing followed by a bacteria culture, which takes a considerable length of time. This makes accurate prediction of wound healing challenging in the clinical setting. In addition, assessment typically requires frequent manual removal of the wound dressing, which elevates the risks of infection and may cause additional pain and trauma for patients.
‘To address this challenge, NUS researchers combined our expertise in flexible electronics, artificial intelligence and sensor-data processing with the nano-sensor capabilities of IMRE researchers to develop an innovative solution that could benefit patients with complex wound conditions,’ said associate professor Benjamin Tee from the Department of Materials Science and Engineering under the NUS College of Design and Engineering, and the NUS Institute for Health Innovation & Technology.
Advertisement
Most wearable wound sensors measure only one or a small number of parameters and require bulky printed circuit boards and batteries. In contrast, the battery-free PETAL sensor patch is comprised of five colorimetric sensors that can determine the patient’s wound healing status within 15 minutes by measuring a combination of biomarkers – temperature, pH, trimethylamine, uric acid and moisture of the wound. These biomarkers were carefully selected to effectively assess wound inflammation and infection, as well as the condition of the wound environment. More biomarkers can be added if required.
‘We designed the paper-like PETAL sensor patch to be thin, flexible and biocompatible, allowing it to be easily and safely integrated with wound dressing for the detection of biomarkers. We can thus potentially use this convenient sensor patch for prompt, low-cost wound-care management at hospitals or even in non-specialist healthcare settings such as homes,’ explained Su Xiaodi, principal scientist in the Soft Materials Department at IMRE.
The sensor patch can be customised for different types and sizes of wounds, and is able to operate without an energy source; sensor images are captured by a mobile phone and analysed by AI algorithms to determine the patient’s healing status.
‘Our AI algorithm is capable of rapidly processing data from a digital image of the sensor patch for very accurate classification of healing status,’ Tee said. ‘This can be done without removing the sensor from the wound. In this way, doctors and patients can monitor wounds more regularly with little interruption to wound healing. Timely medical intervention can then be administered appropriately to prevent adverse complications and scarring.’
Each PETAL sensor patch consists of a fluidic panel patterned in the form of a five-petal pinwheel flower, with each ‘petal’ acting as a sensing region. An opening in the centre of the fluidic panel collects fluid from the wound and distributes it evenly via five sampling channels to the sensing regions for analysis. Each sensing region uses a different colour-changing chemical to detect and measure the respective wound indicators.
The fluidic panel is sandwiched between two thin films. The top transparent-silicone layer allows for normal skin functions of oxygen and moisture exchange, and also enables image display for accurate image capture and analysis. The bottom wound-contact layer gently attaches the sensor patch to the skin and protects the wound bed from direct contact with the sensor panel, to minimise wound-tissue disruption.
After sufficient wound fluid is accumulated (usually within a few hours or over a few days), the PETAL sensor patch will complete the detection of biomarkers within 15 minutes. Images or a video of the sensor patch can be recorded on a mobile phone for classification using the proprietary AI algorithm.
There are no apparent signs of adverse reactions observed on the skin surface in contact with the PETAL sensor patch over four days, demonstrating the biocompatibility of the patch for ambulatory wound monitoring.
In the current study, the performance of the PETAL sensor patch was demonstrated on chronic wounds and burn wounds, but it can be adapted and customised for other wound types by incorporating different colorimetric sensors, such as glucose, lactate or Interleukin-6 for diabetic ulcers. The number of detection zones can also be easily reconfigured to detect different biomarkers concurrently, so its application can be broadened for different wound types.
An international patent has been filed for the invention and the researchers plan to advance to human clinical trials.
The research has been published in Science Advances.
