Non-Invasive Glucose Monitoring Technologies: A ComprehensiveReview of Wearables, AI, and Emerging Technologies

Diabetes mellitus, a global health challenge affecting millions, necessitates advanced solu-tions for continuous glucose monitoring (CGM) to improve disease management and mitigatecomplications. Conventional methods such as glycated hemoglobin (HbA1c) and fingerstickblood glucose tests are limited by invasiveness, intermittent data, and patient discomfort. Thiscomprehensive review explores cutting edge non invasive glucose monitoring technologies,emphasizing their potential to transform diabetes care through wearable sensors, artificialintelligence (AI), and predictive analytics. We systematically evaluate emerging approaches thatleverage alternative biofluids (e.g., sweat, saliva, tears) and optical sensing modalities, includingnear infrared (NIR) spectroscopy, Raman spectroscopy, and smart contact lens biosensors. Theseinnovations address key challenges in patient safety, usability, and real time data accuracy.Furthermore, we highlight the role of machine learning (ML) and deep learning (DL) inenhancing glucose prediction models, enabling adaptive monitoring and personalized insights.The integration of Large Language Models (LLMs) into digital health platforms is also discussed,showcasing their potential to support clinical decision making and patient empowerment.By bridging engineering advancements with clinical applications, these technologies promiseseamless integration into wearable health devices and e-health ecosystems, fostering proactive,data-driven diabetes management. This review underscores the transformative impact of noninvasive monitoring on health technology assessment, disease management, and patient centeredcare, while identifying future directions for research and implementation.