Compact near-infrared photoacoustic spectroscopy glucose detection system based on a triangular prism-shaped photoacoustic cell

Diabetes mellitus, characterized by impaired glucose regulation, poses a major global health burden. Consequently, the development of non-invasive blood glucose monitoring systems is of high clinical importance. Photoacoustic spectroscopy has gained attention as a promising technique due to its high sensitivity and favorable signal-to-noise ratio (SNR). The photoacoustic cell (PAC), as the core component, amplifies weak signals and reduces noise. Currently, most resonant designs are bulky with long cavities, which limit system integration and increase optical attenuation, thereby degrading the SNR. In this study, a small triangular prism-shaped PAC (tPAC) was developed with a volume of only 74.29<i>m</i><i>m</i>³ and a resonance frequency of 29.43 kHz. Based on this cell, a low-cost, compact near-infrared photoacoustic spectroscopy system was built for detecting glucose solutions across normal and hyperglycemic ranges. The system achieved a detection limit of 28.63 mg/dL and a strong correlation (96.58%) between photoacoustic signal amplitude and glucose concentration. Partial least-squares regression and multiple linear regression have been utilized to establish concentration prediction models, achieving accuracies of 87.13% and 92.1%, respectively. All predicted values were located in zones A and B of Clarke's error grid analysis. These results indicate that the proposed tPAC is capable of effectively detecting photoacoustic signals.