Optimization Methods Can Increase the Durability of Smart Electrochemical Biosensors

Abstract

Biosensors are devices equipped with bioreceptors for specific and selective detection of various analytes. They work generally based on either redox or inhibition reactions. Due to the fact that the activity of the bioreceptor immobilized on the working electrode of electrochemical biosensors reduces over time, it should be replaced frequently, resulting in undesirable effects on costs and commercialization of biosensors. In this research, machine learning was used as the decision-making unit of an electrochemical nitrate biosensor considering electrochemical data, pH of the samples, and enzyme activity decrement due to its lifespan and storage temperature. Genetic algorithm (GA), particle swarm optimization (PSO), Harris hawks optimization (HHO), and whale optimization algorithm (WOA) were used to optimize the parameters of artificial neural networks (ANNs) and support vector machine (SVM) to predict nitrate concentration. The findings showed that WOA-ANN and GA-SVM resulted in a promising performance with coefficients of determination (R <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ) equal to 0.93 and 0.91, respectively, even three weeks after the enzyme immobilization. Comparing the results of the present study and those of former works on the development of smart biosensors revealed the importance of improving the performance of smart biosensors by using metaheuristic optimization methods.