Wearable Optical Sensors: Toward Machine Learning-Enabled Biomarker Monitoring

Wearable optical sensors are increasingly being employed for continuous biomarker detection. In contrast to conventional blood testing, the continuous detection of biomarkers in biofluids offers a non-invasive method for disease diagnosis and management. Nanomaterials have proven to be highly efficient in wearable technologies due to their high surface area, exceptional sensitivity, small size, mechanical robustness, and biocompatibility. However, the development of efficient optical skins remains a priority due to the complexity of biofluids, the need for mechanical compatibility with the skin, power sources, wireless communication, the Internet of Things, and data analysis. This mini-review provides an overview of wearable optical sensors, exploring in materials, their wearability characteristics, and their detection performances. This mini-review offers a summary of wearable optical sensors for continuous biomarker analysis, highlighting the challenges faced in this field. It also introduces a new perspective on integration them with IoT and AI toward telemedicine. However, few articles on wearable optical biomarker detection have reported the use of machine learning methods for developing classification models (up to 2024). Machine learning methods are promising approaches to provide real-time colorimetric monitoring with high accuracy.