StanfordBDHG / PediatricAppleWatchStudy

The Pediatric Apple Watch Study Application
MIT License
4 stars 0 forks source link
apple cs342 digitalhealth fhir pediatric stanford watch

Pediatric Apple Watch Study Application

Beta Deployment codecov DOI

This repository contains the Pediatric Apple Watch Study Application application. The Pediatric Apple Watch Study Application uses the Spezi ecosystem and builds on top of the Stanford Spezi Template Application.

Application Structure

The Spezi Template Application uses a modularized structure using the Spezi modules enabled by the Swift Package Manager.

The application uses the FHIR standard to provide a shared standard to encode data exchanged between different modules.

[!NOTE]
Do you want to learn more about the Stanford Spezi Template Application and how to use, extend, and modify this application? Check out the Stanford Spezi Template Application documentation

Build and Run the Application

You can build and run the application using Xcode by opening up the PAWS.xcodeproj.

Generate Invitation Codes

When signing in to the application for the first time, you will be required to enter a valid invitation code before a user account is created. Use the upload_codes.py script to generate new codes and upload them to a specified Firebase instance or a local file.

export FIRESTORE_EMULATOR_HOST="localhost:8080"
export GCLOUD_PROJECT=<project_id>

python -m scripts.upload_codes --outfile=<local_path> \
--count=<number_of_codes> --length=<code_length> \
--service_account=<service_account_key_file> [--dry]

Generate Screenshots

PAWS uses Fastlane Snapshots to automatically screenshot specific screens in the app during UI tests. To generate new screenshots, you will likewise need to set the proper environment variables for your shell session.

firebase emulators:start --import=./firebase

Then, run fastlane snapshot. By default, results will end up in the .screenshots folder, overwriting previous files.

[!NOTE] Snapshot will run UI tests and concurrently take screenshots on multiple device simulators. As such, multiple new PAWS accounts will be created, possibly in rapid succession, using the same hard-coded testing invitation codes.

The current workaround for simultaneous account registrations during fastlane snapshot is to continually reset invitation codes to an unused state in Firestore by running a designated Python script on repeat (in a shell session with the same environment variables).

for i in {1..360}; do python -m scripts.upload_codes; sleep 10; done

ECG Data Manager

The ECG Data Manager provides capabilities for reviewing and exploring the recorded ECG data. It relies on the spezi_data_pipeline package, which is a library engineered to improve workflows associated with data accessibility and analysis in healthcare environments. In addition to the functions and classes of the spezi_data_pipeline, the two notebooks, namely ECGReviewer.ipynb and ECGExplore.ipynb, offer an environment for interactive data visualization and analysis.

The ECG Data Manager includes:

Notebook Setup Instructions

You can open and run the ECGReviewer.ipynb and ECGExplorer.ipynb notebooks in, e.g., Google Colab. Once the notebook is open, execute the following cell to clone the PediatricAppleWatchStudy repository and navigate into the cloned directory:

# Clone GitHub repository for Spezi ECG Data Pipeline
git clone https://github.com/StanfordBDHG/PediatricAppleWatchStudy.git
cd PediatricAppleWatchStudy/ecg_data_manager

To run the notebooks, add them to Colab Enterprise within the same Google Cloud project as your Firebase setup. For other Python notebook environments, use the Firebase credentials and upload the serviceAccountKey_file.json to the workspace directory to enable Firebase access. This file is essential for authentication and should be securely handled.

Use the Interactive ECG Reviewing Tool

To start reviewing ECG data, execute the cells in your notebook.

This interactive tool allows you to plot ECG data, add diagnoses, evaluate the trace quality, and add notes.

ecg_data_interactive_reviewer.png

Use the Interactive ECG Exploring Tool

To start exploring ECG data, execute the cells in your notebook.

This interactive tool allows you to plot ECG data, filter ECG recordings, and select specific users and timestamps.

ecg_data_interactive_explorer.png

Firebase & Google Cloud Setup

The Google Cloud Setup at Stanford to deploy the project requires the following setup for Google Cloud Firebase and to execute the ECG review and exploration tools.

Firebase Setup

Each Firebase Project for development, staging, and production GitHub environments need the following configurations:

The CI setup requires a github-deployment@PROJECT_ID.iam.gserviceaccount.com account that requires the following rules:

Set up a cloudfunctionsserviceaccount@PROJECT_ID.iam.gserviceaccount.com to execute cloud functions. It needs the following rules:

Notebooks & Colab Enterprise

Created a storage bucket that is used to store the packaged dependencies & code for the Python notebooks in a versioned and isolated state. You need to enable Colab Enterprise.

To secure the data in the notebook, the network access should be retricted to needed Google Services only and using runners that only use a private network.

Run the following commands in the ECGReviewer folder to package the dependencies and upload them, as well as the modules folder to the cloud storage bucket if no outside internet access is enabled.

mkdir packages
pip download -r requirements.txt -d packages
tar -czvf packages.tar.gz packages/

Copy the ECGExporter and ECGReviewer notebooks in Colab Enterprise, uncomment, and adapt the storage bucket paths if the restricted network access is configured.

Continous Integration Setup

The project supports different GitHub environments (development, staging, and production).

Contributing

Contributions to this project are welcome. Please make sure to read the contribution guidelines and the contributor covenant code of conduct first.

License

This project is licensed under the MIT License. See Licenses for more information.

Our Research

For more information, check out our website at biodesigndigitalhealth.stanford.edu.

Stanford Byers Center for Biodesign Logo Stanford Byers Center for Biodesign Logo