Noninvasive Estimation of Glycated Hemoglobin: A Comprehensive Review of Emerging Technologies

Diabetes is a chronic condition in which an individual's body is unable to efficiently break down blood sugar due to insufficient insulin. Glycated hemoglobin (HbA1c) is an important biomarker for diagnosing diabetes. The HbA1c test measures the average blood glucose level over a 3-month period and serves as both a diagnostic tool for diabetes and a measure of glycemic control. Although effective, HbA1c testing often relies on blood-pricking, which can be inconvenient or uncomfortable. This paper explores noninvasive technologies for HbA1c estimation and diabetes management. The study begins by emphasizing the critical need for advanced frameworks to eliminate the need for blood sampling. It then provides a comparative analysis of emerging methodologies, including photoplethysmography, Raman and near infrared (NIR) spectroscopy, physiological data, and biofluid-based sensing. After that, it identifies key challenges related to their accuracy, standardization, and clinical adoption. Furthermore, this study highlights the integral role of machine learning in enhancing noninvasive HbA1c estimation. Finally, we conclude by outlining the limitations and open challenges for noninvasive prediction of HbA1c levels.