Multi-modal Point-of-Care Diagnostics for COVID-19 Based On Acoustics and Symptoms

Abstract

The research direction of identifying acoustic bio-markers of respiratory diseases has received renewed interest following the onset of COVID-19 pandemic. In this paper, we design an approach to COVID-19 diagnostic using crowd-sourced multi-modal data. The data resource, consisting of acoustic signals like cough, breathing, and speech signals, along with the data of symptoms, are recorded using a web-application over a period of ten months. We investigate the use of statistical descriptors of simple time-frequency features for acoustic signals and binary features for the presence of symptoms. Unlike previous works, we primarily focus on the application of simple linear classifiers like logistic regression and support vector machines for acoustic data while decision tree models are employed on the symptoms data. We show that a multimodal integration of acoustics and symptoms classifiers achieves an area-under-curve (AUC) of 92.40, a significant improvement over any individual modality. Several ablation experiments are also provided which highlight the acoustic and symptom dimensions that are important for the task of COVID-19 diagnostics.

CoughNet-V2: A Scalable Multimodal DNN Framework for Point-of-Care Edge Devices to Detect Symptomatic COVID-19 Cough

“…To have more accurate representation of human hearing, the mel-frequency cepstrum is the way to go, as the frequency bands are evenly spaced on this scale, which is closer to how our ears actually respond. Authors in [26] presented similar approach to ours. However, they have worked on Coswara [18] dataset whereas we have worked on COUGHVID dataset [27].…”

“…Authors in [21]- [23] proposed unimodal machine learning and deep learning based approach to classify whether the cough sound has COVID-19 signature or not. Authors in [24]- [26] proposed multimodal machine learning and deep learning approaches to classify COVID cough audios. However, authors in [24], [26] had to manually select different features out of COVID-19 cough sound whereas our proposed CoughNet-V2 automatically selects features from the mel frequency cepstral coefficients (MFCCs) derived from the cough audios.…”

“…Authors in [24]- [26] proposed multimodal machine learning and deep learning approaches to classify COVID cough audios. However, authors in [24], [26] had to manually select different features out of COVID-19 cough sound whereas our proposed CoughNet-V2 automatically selects features from the mel frequency cepstral coefficients (MFCCs) derived from the cough audios. Mel-frequency cepstrum (MFC) is the short-term power power spectrum representation of a sound, based on the linear cosine transform of a log power spectrum on a nonlinear mel-scale of the frequency.…”

A Comprehensive Review on COVID-19 Detection Based on Cough Sounds, Symptoms, CXR, and CT Images

“…This method takes into account a person's temperature, breathing, and different cough symptoms. Chetupalli et al [144] employed SVM and logistic regression to separate coughing and breathing signals. Canas et al [145] suggested a logistic regression model and the NHS algorithm to predict early indicators of COVID-19 infection in a dataset of 198040 symptoms from patients in the United Kingdom.…”