npj Digital Medicine paper Supplementary Material

This repository contains codes accompanying the npj digital medicine paper named: Sounds of COVID-19: exploring realistic performance of audio-based digital testing.

Introduction

To identify Coronavirus disease (COVID-19) cases efficiently, affordably, and at scale, recent work has shown how audio (including cough, breathing, and voice) based approaches can be used for testing. However, there is a lack of exploration of how biases and methodological decisions impact these tools' performance in practice. Due to these issues, many researchers raised concerns about the feasibility and effectiveness of such models if deployed in real settings. In this work, we investigate the limits of audio-based COVID-19 testing to create the foundation of realistically applicable audio tools. The aim of this study is two-fold: first, to investigate the performance of an audio-based COVID-19 testing method while addressing the issues noted in the previous studies, by using a large crowd-sourced dataset, to the best of our knowledge, unbiased data, with a methodological design based on realistic assumptions (e.g. independent user split). Second, to explore the impact of biases and design pipeline on the performance.

Dataset

For the above purpose, we developed an app to crowd-source audio data. More details about this project can be found at https://www.covid-19-sounds.org/. Naturally, data of such origin are sensitive in nature. Thus, to gain access to the full dataset, a Data Transfer Agreement (DTA) needs to be signed. To obtain this or to contact the team, please email us: covid-19-sounds@cl.cam.ac.uk.

Code

Our models are implemented in Python 3 (tested using 3.6) using Tensorflow 1.15. Please note that due to the requirement of Tensorflow 1.15 versions of python above 3.7 are not supported. Before you start, please ensure that you have requested and downloaded the task files as they are not included when cloning the repository. After receiving the files and in order to reproduce the results, you need to create an appropriate virtual environment and activate it. You can create such an environment, as is shown below:

# create a virtual environment for this project (assuming you are in covid19-sounds-neurips)
python3 -m venv ./venv
# now activate it
source ./venv/bin/activate
# and finally, install the dependencies
pip install -r ./requirements.txt

As described in the paper, we develop a VGGish based model. We will now describe how to reproduce the results presented in the paper using the data provided. To reproduce, please follow the steps below:

1. Navigate to the cloned repository (normally, covid19-sounds-npjDM)

2. Ensure you have downloaded the npjDM21_data files from Google Drive

   • Move the drive file data_0426_en_all and preprocess to path COVID19_prediction/data

   • Prepare input

      cd ./COVID19_prediction/data
      python pickle_data_from_csv.py

   • Go to model's path cd ./COVID19_prediction/COVID_model

     • Train the model sh run_train.sh
       • After training, models will be stored in path ./COVID19_prediction/data/train/
     • Test the model sh run_test.sh

3. Baselined on model output, we also provide the codes to generate table and figure in our paper in covid19-sounds-npjDM/results:

   • Move the drive file output/* to path COVID19_prediction/results/output/
   • Main experimental findings and figures: Experiment_results_figures.ipynb
   • Results in tables: Experiment_results_tables.ipynb
   • Results with bias in the data: Experiment_with_bias_table.ipynb

Issues

The code was initially created by Tong Xia (tx229@cl.cam.ac.uk). For any questions, please either contact the authors directly or create an issue.