Chemometrics and Signal Processing-Assisted Design of a Textile-Based Colorimetric Sensing Platform for Real-Time Monitoring of Glucose

Abstract

Colorimetric analysis of biomarkers in biological fluids, such as sweat and saliva, using textile- and paper-based platforms promises to be a powerful point-of-care analytical tool. Despite exponential advances in material design, device fabrication and image processing over the past years, the majority of paper/fabric-based analytical strips and devices, thus far, are strictly restricted to controlled laboratory settings. In situ and real time monitoring of biological analytes using familiar garments or accessories are centered around electrochemical sensors. In contrast, colorimetric devices that can theoretically enable the quantification of analyte concentration when coupled with image capture devices (such as a smartphone cameras) are comparatively fewer and less advanced. Here, garment design strategies and various embodiments of wireless communication electronics are applied cooperatively to afford a noninvasive platform for quantitative monitoring of a test-bed biomarker, glucose. The electronics-embedded garments contain two functional parts: a functionalized cotton substrate with enzyme/chromogenic reagent and a near-field communication device. The enzyme/chromogenic reagent mixture is directed into a circular reservoir created on the cotton substrate via the capillary force of the cotton and hydrophobic wax barrier created via a cheap and simple wax-drawing technique. This wax barrier limits the irregular flow of the reagent mixture. The second part is a near-field-communication system that causes no irritation and discomfort to the skin. Once a smartphone is put at a close distance, the electronic system can enable the wireless transmission. A built-in application was developed to automatically detect the sensor spot via a signal processing pipeline and extract the R, G and B channel intensities for database construction and analysis. The chemical properties and geometric structures of the glucose sensor were optimized with the aid of a combinatorial method—materials optimization and mathematical optimization. The former aims to overcome the blurring and the loss of the signal due to enzyme washout from the detection zone under capillary forces exerted during wear. This problem was mitigated by immobilizing the enzyme inside a chitosan-carboxymethyl cellulose hybrid via a precisely controlled layer-by-layer deposition protocol. The sensor was further characterized by field-emission scanning electron microscopy (FESEM) to understand long term storage stability and surface morphology-dependent colorimetric response. A mathematical strategy was utilized for the design and selection of an optimal near-field measurement system. This optimization procedure is particularly important for commercialization. It aims to minimize the number of experiments and reduce the quantity of the chromogenic reagent mixture that needs to be deposited on the cotton substrate. Central composite designs were used to estimate the effect of individual variables (i.e. glucose oxidase concentration, horseradish peroxidase concentration and 3,3’,5,5’-Tetramethylbenzidine concentration) and their interactions on the optimization of the color intensity. The significance of the main and interaction effects and the applicability of the model were inferred from the analysis of variance (ANOVA). The optimum value was verified and validated with experimental results. This textile-based colorimetric sensing platform exhibits simple-to-fabricate, and cost-effective nature which might facilitate the broad commercial usage in health monitoring and clinical diagnosis.