Graphs Applied in Healthcare AI
(GrApH AI) working group
The Mission
Apply graph technology
to make healthcare data
AI-friendly from the ground up
The People
We are an open, sharing community of physicians, data scientists, data engineers, software engineers, students, and others inspired by a common vision
The Plan
We are working use-case by use-case with real patient data to build a foundation of knowledge, a toolkit, and a corps of colleagues to act as a springboard for real-world implementations
IN THE QUEUE
Recommendation engine to present plan items in real time as you type in the problem
Appropriate for this patient, with this problem, at this moment
Use Cases
Workflow
Clinicians and data scientists collaborate
to break down use cases into bite-sized challenges, which are fed through technical sub-groups until the use case is fulfilled
Data
Data
MIMIC-III is a dataset of de-identified data from over 40,000 patients admitted to the intensive care unit of the Beth Isreal Deaconess Medical Center in Boston, Massachusetts. The dataset is curated by MIT and is openly available on physionet.org.
Code
Our code is pushed to GitHub under the MIT license,
which means we allow anyone to build upon our work
even in proprietary, commercial products
We support those who want to go out and
make things happen in the real world
visit
us on
github
Data Models
Data models can be shared across use cases
and institutions to reduce the development workload
Learn more at our GitHub.io pages
Get Involved
Come introduce yourself on
and join us at any of our group meetings
(video conference links in the calendar)