International Research Journal of Education and Technology

Non-invasive blood glucose monitoring offers a pain-free alternative to traditional finger-prick methods, enabling continuous glucose tracking and improved diabetes management. Photoplethysmography (PPG), a cost-effective optical technique traditionally used for heart rate and oxygen saturation monitoring, has emerged as a promising approach for glucose measurement. Glucose levels influence blood optical properties, such as scattering and absorption, which affect PPG waveforms. Researchers have developed algorithms to analyze PPG features—such as amplitude and pulse transit time—to estimate glucose levels. Machine learning techniques, including neural networks and support vector machines, have enhanced prediction accuracy. Despite promising initial results, challenges like motion artifacts, skin tone variations, and physiological differences must be addressed. Further clinical validation is needed to achieve regulatory approval and ensure reliability in real-world applications.

Keywords: Non-invasive glucose monitoring, diabetes management, photoplethysmography, machine learning, signal processing