An extensible, open-source, and web-based signal quality assessment pipeline for ambulatory cardiovascular data

---

HeartView is a Python-based signal quality assessment pipeline and dashboard that visualizes and summarizes segment-by-segment quantification of missing and invalid beats in wearable electrocardiograph (ECG) and photoplethysmograph (PPG) data obtained in research contexts.

In contrast to other existing tools, HeartView provides a graphical user interface intended to increase efficiency and accessibility for a wider range of researchers who may not otherwise be able to conduct rigorous quality assessment on their data. As such, HeartView is meant as a diagnostic tool for pre- and post-artifact correction of cardiovascular data. We aim to help researchers make more informed decisions about later data cleaning and processing procedures and the reliabiltiy of their data when wearable biosensor systems are used.

Currently, HeartView works with data collected from the Actiwave Cardio, Empatica E4, and other ECG devices outputting data in comma-separated value (CSV) format.

Features

• File Reader
  Read and transform raw ECG, PPG, and accelerometer data from European Data Format (EDF), archive (ZIP), and CSV files.
• Configuration File Exporter
  Define and save pipeline parameters in a JSON configuration file that can be loaded for use on the same dataset later.
• Signal Filters
  Filter out noise from baseline wander, muscle (EMG) activity, and powerline interference from your ECG or PPG data.
• Beat Detection
  Extract heartbeats from your ECG [1] and Empatica E4 data.
• Visualization Dashboard
  View and interact with our signal quality assessment chart and ECG/PPG, interbeat interval (IBI), and acceleration signal plots by segment.
• Signal Quality Metrics
  Generate segment-by-segment information about missing and invalid data.
• Manual Beat Editor
  Ability to update an existing beat and manually inputting beat(s)

Citation

If you use this software in your research, please cite this paper. 💛

@inproceedings{Yamane2024,
  author    = {Yamane, N. and Mishra, V. and Goodwin, M.S.},
  title     = {HeartView: An Extensible, Open-Source, Web-Based Signal Quality Assessment Pipeline for Ambulatory Cardiovascular Data},
  booktitle = {Pervasive Computing Technologies for Healthcare. PH 2023},
  series    = {Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering},
  volume    = {572},
  year      = {2024},
  editor    = {Salvi, D. and Van Gorp, P. and Shah, S.A.},
  publisher = {Springer, Cham},
  doi       = {10.1007/978-3-031-59717-6_8},
}

Latest Release [HeartView 2.0]

Pipeline Enhancements:

• Added beat detection algorithms for ECG and PPG pre-processing.
• Added default filters for ECG, PPG, and ACC.
• Implemented artifact identification methods in the SQA.py module.
• Added progress bars for real-time feedback.

Dashboard Improvements:

• Enhanced handling of uploaded data files.
• Added artifactual beats visualization and metrics.
• Improved segment view selection options.
• Enabled data downsampling for faster chart rendering.

Structural Refactoring:

• Refactored functions into new classes for clarity.
• Updated function and package names in documentation.

For a full list of changes, see the full changelog.

Roadmap

We're constantly working on improving HeartView. Here's a glimpse of what's in store in the near future:

• Automated beat correction functionality
• Manual beat correction feature
• Data imputation techniques
• Electrodermal activity (EDA) data pre-processing and quality assessment

Installation

1. Clone the HeartView GitHub repository into a directory of your choice.

cd <directory>  # replace <directory> with your directory

git clone /~https://github.com/cbslneu/heartview.git

2. Set up and activate a virtual environment in the heartview project directory.
   ❗️Note: If you do not have virtualenv installed, run pip3 install virtualenv before proceeding below.

cd heartview

virtualenv venv -p python3

If you are a Mac/Linux user:

source venv/bin/activate

If you are a Windows user:

venv\Scripts\activate

3. Install all project dependencies.

pip3 install -r requirements.txt

Installation for Manual Beat Editor

Manual Beat Editor requires Node (v20.x.x +), please be sure to install before proceeding with the installation below

Run the following code below to check if Node is installed on your machine:

node --version

If an error occurs, please refer to this link to install Node on your machine: https://nodejs.org/en/download/package-manager

1. Go to the beat-editor directory

cd beat-editor

2. Install the required modules for beat-editor

npm install

3. Go to the server folder

cd server

4. Install the required modules for beat-editor's backend

npm install

HeartView Dashboard

Executing

1. Within the activated virtual environment (i.e., source <directory>/heartview/venv/bin/activate), run the command:

python3 app.py

2. Open your web browser and go to: http://127.0.0.1:8050/

Executing Manual Beat Editor

1. Navigate to the server directory

cd server

2. Run the following line to start the backend

npm run

3. Open up another terminal tab or window and navigate back to beat-editor, once there run the line from the step above to start the frontend code.

Terminating

1. Kill the dashboard program: press CTRL+c.
2. Exit the virtual environment: deactivate.