Simulation-based integration of an advanced spectroscopic refractive index biosensor via machine learning algorithms...

Sucrose amounts in a water solution have diverse applications in healthcare, metabolic regulation, and food quality control. It is widely utilized as an alternative for physiological circumstances, fermenting phases, or quality assurance based on its level of concentration in biological and industrial samples. Current detection methodologies, such as enzymatic tests or chromatography, provide remarkable precision but tend to be time-consuming, dependent on reagents, and insufficient for monitoring in real time. To fix these limitations, we have envisioned a Four-Arm Ring-Embedded Refractive Index Sensor (FARERIS) to determine various concentrations of sucrose. Additionally, this study has been elaborated with a machine learning approach, which gives accurate as well as time-saving results. The design of the geometry of a quad-arm pattern leads to multiple pathways of light coupling while simultaneously enhancing interaction between the concentrating optical structures and their surroundings, the analyte. The achieved significant sensitivity values of 1553.39 nm/RIU, 1600.00 nm/RIU, 1518.98 nm/RIU, 1481.48 nm/RIU, 1647.05 nm/RIU, 1477.27 nm/RIU, and significant values for detection limits are 0.000269 RIU, 0.000440, 0.000310, 0.000434, 0.000360, and 0.000332 for the sucrose concentration of 5%, 10%, 15%, 20%, 25%, and 30%, correspondingly. The significant quality factor of 1723.96, and the significant figure of merit of 1283.79 RIU^-1 have been obtained for detecting 5% of sucrose concentration, with the determined significant value of neural network fitting is 0.96475. The fusion of optical sensing and computational intelligence serves as a feasible approach towards advanced, adaptable biosensor devices.