Novel implementation of “Filtered-Reconstruction ICA” enabled wheeze component separation for preliminary diagnosis of COVID-19

Abstract

The outbreak of COVID-19 pandemic has made the whole world at a standstill. With researches advancing towards invention of vaccines, in this project work, we present a faithful and novel approach towards preliminary diagnosis of COVID-19 for infants in the age group of 2-24 months. Wheeze is a prime symptom of COVID-19 infection. Due to cry of babies, many a times the intensity of wheeze cannot be evaluated, which leads to improper medications or even mortality. For this, the concept of Blind Source Separation using Independent Component Analysis (ICA) is utilized. Reconstruction Independent Component Analysis (RICA) is one of the primarily utilized ICA technologies. This technique is a conventionally used mechanism in order to estimate airplanes and also equally utilized in the biomedical field for extracting out desired signals from the interference mixtures like electrocardiogram (ECG), magnetoencephalography (MEG) and electroencephalogram (EEG). Henceforth, RICA is accomplished for the real time processing of signals. In this paper, an algorithm based on RICA and a Least Mean Square (LMS) filter has been developed. The idea behind this paper is to extract the wheeze intensity from a mix of cry and wheeze of infants infected with the disease. The aim of this work is to implement an effective blind component separation algorithm in which a RICA method is used to extract the wheeze components and further an LMS filter is used to enhance the SNR, so that even a remote doctor could prescribe appropriate medication based on the wheeze intensity.