SMART BIOSENSING: AI IN CONTINUOUS GLUCOSE MONITORING

Abstract Continuous glucose monitoring (CGM) has revolutionized the management of diabetes byproviding real time glucose levels and trend data, which is important in overcoming thelimitations of standard capillary glucose measurements. The analysis of the voluminous CGMdata set and the increased accuracy of the sensors have remained a daunting task, however.Artificial intelligence (AI) has emerged as a powerful tool to assist CGM with predictiveanalytics and creation of more advanced biosensors This review examines the workingprinciples of CGM and elucidates the rationale behind the increasing adoption of AI inglucose monitoring.We analyze machine-learning and deep-learning approaches utilized topredict glycaemic processes, alleviate sensor errors, and produce personalized alerts.Nanomaterials-based next-generation biosensors incorporating enzyme-free detectionschemes are examined, including wearable and implantable CGM systems that incorporate AIalgorithms and edge-computing designs to provide real-time decision support. The benefits ofAI-based CGM in terms of improving glycaemic control and convenience to patients areoutlined, and the accompanying issues, namely the accuracy of the hypoglycaemic range,data protection, and the need to regulate AI, are discussed. Future trends show a trendtowards more intelligent and less intrusive glucose monitors and fully automated diabetesmanagement systems. To conclude, AI is expected to significantly improve continuousglucose sensing, making glucose monitoring in the context of diabetes care even smarter,more accurate, and more proactive.