A python library for digital health measurement

What is OpenWillis?

OpenWillis is a python library for digital measurement of health.

We developed OpenWillis to establish standardized methods in digital phenotyping.

OpenWillis lives entirely on Github; the repo can be found under bklynhlth/openwillis. It ensures code used to calculate digital measures is accessible to all researchers.

The official documentation is housed within the Github wiki. It includes technical instructions on how to install and use OpenWillis and details the methods used to calculate each measure.

The documentation also includes helpful research guidelines that can assist OpenWillis users collect high quality measurements from their data.

What is OpenWillis for?

Bridging the gap between academic research and digital health applications

OpenWillis allows academic researchers easy access to methods in digital measurement of health. This enables them to use this code in their research and in doing so build trust in a common library of methods. This library can then serve as the foundation of digital health tools everywhere, standardizing measurement methods used across the field.

What can OpenWillis do?

OpenWillis is a collection of python functions designed to quantify clinically meaningful behaviors.

Overall expressivity in the face regardless of emotion

Presence and intensity of individual emotions

Acoustic properties of voice related to health and functioning

Transcription of speech into text in multiple languages

Linguistic markers of health and functioning

Speech diarization to separate voice from multiple speakers

We're continuing to add to this list of measures. Want to learn more or suggest your own? Get in touch.

If you're an OpenWillis user or just want to stay updated on new functions, future releases, trainings, and workshops, sign up for our listserv.