Development of an intelligent system for the detection of corona virus using artificial neural network

TL;DR

This paper develops an artificial neural network-based system for COVID-19 detection using temperature data, achieving high accuracy and reliability, and implements it on FPGA hardware for potential real-time application.

Contribution

The paper introduces a novel neural network model trained on temperature data and its hardware implementation on FPGA for COVID-19 detection.

Findings

Accuracy of 97% in detection

Regression value of 0.967 indicating high model fit

MSE of 0.00100Mu demonstrating model precision

Abstract

This paper presents the development of an intelligent system for the detection of coronavirus using artificial neural network. This was done after series of literature review which indicated that high fever accounts for 87.9% of the COVID-19 symptoms. 683 temperature data of COVID-19 patients at >= 38C^o were collected from Colliery hospital Enugu, Nigeria and used to train an artificial neural network detective model for the detection of COVID-19. The reference model generated was used converted into Verilog codes using Hardware Description Language (HDL) and then burn into a Field Programming Gate Array (FPGA) controller using FPGA tool in Matlab. The performance of the model when evaluated using confusion matrix, regression and means square error (MSE) showed that the regression value is 0.967; the accuracy is 97% and then MSE is 0.00100Mu. These results all implied that the new detection system for is reliable and very effective for the detection of COVID-19.