Machine learning enabled microneedle-based colorimetric pH sensing patch for wound health monitoring and meat spoilage detection

Since pH can alter the biological functions, level of nutrients, wound healing process, and the behavior of chemicals, various healthcare and food industries are showing increased interest in manufacturing low-cost optical pH sensors for meat spoilage detection and wound health monitoring. To meet this demand, we have developed a simple and low-cost machine learning-enabled microneedle-based colorimetric pH sensing patch that can be used for food quality and wound health monitoring applications. The 3D–printed ultrasharp open side channel microneedle array facilitated the autonomous fluid extraction and transportation via surface tension for colorimetric pH sensing. Further, to predict the exact pH value against the obtained color on the pH-test strip, a machine learning model was prepared using experimentally collected different color images obtained from a known pH solution. Furthermore, to make the device user-friendly for older individuals and color-blind individuals, a simple and smartphone-enabled web application was prepared using the developed machine learning model. The proof-of-concept study of the developed patch was demonstrated by determining the pH of real meat samples before and after spoilage and detecting pH in two different skin-mimicking in vitro models (phantom gel and parafilm tape) using a smartphone. The analytical results demonstrated that the developed machine learning-enabled microneedle-based colorimetric pH sensing patch has excellent potential for wound health and food safety applications.