Automatic diagnosis of COVID-19 disease using deep convolutional neural network with multi-feature channel from respiratory sound data: Cough, voice, and breath

Abstract

The problem of respiratory sound classification has received good attention from the clinical scientists and medical researcher’s community in the last year to the diagnosis of COVID-19 disease. The Artificial Intelligence (AI) based models deployed into the real-world to identify the COVID-19 disease from human-generated sounds such as voice/speech, dry cough, and breath. The CNN (Convolutional Neural Network) is used to solve many real-world problems with Artificial Intelligence (AI) based machines. We have proposed and implemented a multi-channeled Deep Convolutional Network (DCNN) for automatic diagnosis of COVID-19 disease from human respiratory sounds like a voice, dry cough, and breath, and it will give better accuracy and performance than previous models. We have applied multi-feature channels such as the data De-noising Auto Encoder (DAE) technique, GFCC (Gamma-tone Frequency Cepstral Coefficients), and IMFCC (Improved Multi-frequency Cepstral Coefficients) methods on augmented data to extract the deep features for the input of the CNN. The proposed approach improves system performance to the diagnosis of COVID-19 disease and provides better results on the COVID-19 respiratory sound dataset.

CovNet: A Transfer Learning Framework for Automatic COVID-19 Detection From Crowd-Sourced Cough Sounds

“…Very deep CNN, known as VGG, were originally designed with up to 19 weight layers and achieved great performance on the large-scale image classification task (40,46). VGG or VGGlike architectures were applied to extract audio features from respiratory sound data for COVID-19 detection and obtained good performances (15,47). -6), (D) MobileNet-6.…”

COVID-19 disease diagnosis with light-weight CNN using modified MFCC and enhanced GFCC from human respiratory sounds

“…This existing pertained model extract features from lower inflammation and upper respiratory inflammation parameters, which helps detect COVID-19 symptoms. Lella and Alphonse [34,35] implemented an Artificial Intelligence (AI) framework to automatically classify COVID-19 symptoms using a 1D CNN network and Deep CNN from respiratory sound data. The 1D CNN model is implemented using a Data De-noising Auto-Encoder (DDAE) mechanism, which helps to give good performance to identify SARS-CoV-2 disease symptoms from various respiratory sounds (cough, voice, and breath).…”

“…This manuscript has associated data in a data repository. [Authors' comment: All data included in this manuscript are not publicly available because the dataset is collected from cited sources in references [14,24,35] for research purposes but is available from the corresponding author on reasonable request. The sample analysis of this work is available in figshare [41]…”

Data Mining untuk Prediksi Status Pasien Covid-19 dengan Pengklasifikasi Naïve Bayes

“…Pasien dengan tingkat kegawatan tinggi memerlukan prioritas penanganan dibanding pasien dengan gelaja sedang atau tanpa gejala [9]. Tenaga medis memerlukan bantuan untuk mengklasifikasi status pasien berdasarkan data pasien secara otomatis untuk mengurangi kelelahan tenaga medis yang harus terus bertugas dan meminimalisir resiko penanganan yang terlambat terhadap pasien [10]. Oleh karena itu dibutuhkan solusi teknologi berbasis data secara otomatis yang dapat membantu mengklasifikasikan status kegawatan berdasarkan data pasien.…”