Design and Development of Slotted Ring-Shaped High Efficiency Refractive Index Biosensor for Label-Free Detection of Abused Drugs Using Machine Learning

The abuse of desired prescription drugs instantly and efficiently heightens the operation between different body parts, physiological processes, nerves throughout the body, and other bodily functions, raises the possibility of restful sleep disorders, and produces an intense flow of hormonal substances. In spite of its instantaneous adverse effects, being addicted to drugs has long-term negative impacts on individual well-being and can ultimately result in fatalities. The resulting negative effects indicate the need for more recognition and specificity as well as for the effectively functioning, easy-to-use, and reliable identification of illegal drugs. In response to these factors as the main aims, this work has been nominated to the Twin Curve and Frame Refractive Index Biosensor (TC&FRIB) for the identification of abused drugs which include cocaine (C), ketamine (K), morphine (M), and amphetamine (A). The novelty of this work lies in the design and development of the TC&FRIB integrated with a neural network regression model for label–free detection of abused drugs. In comparison with traditional refractive index sensors, the offered configuration addresses cross-reactivity limitations by carrying out machine learning-supported spectral classification, while simultaneously demonstrating improved sensitivity and figure of merit. The combined use of tailored cavity layout with tungsten as the resonant material and comprehensive regression modelling establishes a novel approach for remarkably precise detection of structurally equivalent drug compounds, which will set an innovative standard for biosensor innovation. This nominated design of TC&FRIB relates to a maximum sensitivity value of 849.24 nm/RIU for amphetamine, 842.63 nm//RIU for morphine, 818.18 nm/RIU for ketamine, and 762.41 nm/RIU for cocaine. The related maximum value of detection limit is 0.000271 for morphine. Here, the maximum regression value of 0.99996, and the mean square error value of 3.13910 × 10 -5 have been achieved at DT R = 220 nm & DT R = 240 nm. The device’s potential as an application in the area of drug assessments is shown by the device’s capacity to recognize differences between multiple indicators acquired with amphetamine and other associated drugs.