Answers are grounded in this paper and cite their sources.
Experimental approach analyzes glucose levels in μPADs through smartphone imaging, indicating effective AI classification.
In this study, an AI-driven approach was implemented for μPAD glucose analysis using smartphone images captured under indoor and outdoor conditions with flash on and off. Prior to app integration, three-zone μPADs were fabricated via a wax printing protocol, whose detection areas were modified with 3,3’,5,5’-tetramethylbenzidine (TMB), glucose oxidase (GOx), and horseradish peroxidase (HRP) for glucose detection. Color changes on the μPADs were recorded, and a small dataset was generated by photographing the devices with three different smartphones indoors and outdoors with flash control. MATLAB’s Classification Learner was used to analyze the dataset, and the best performing model (accuracy: 0.95) was exported as .mat files. These numerical outputs were then converted into .c and .h functions using MATLAB Coder and integrated into a custom Android application called "GlucoObserver", enabling real-time AI-based glucose classification under practical imaging scenarios.
No takes yet. Share an insight, caveat, or question.
Yıldız et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: