icdcodex 0.3.0

icd embedding for machine learning

icdcodex

ICD embedding for machine learning, created for MedHacks2020 ❤️.

• Free software: MIT license

• Documentation: https://icdcodex.readthedocs.io.

What is Medhacks?

MedHacks hosted by Johns Hopkins University aims to unite talented and diverse minds from all backgrounds in order to foster a collaborative environment that aims to solve the world’s medical obstacles and issues.

The Problem

ICD coding is a laborous, but difficult to automate by machine learning because the output space if intractably large. (ICD-10CM has over 70,000 codes.) icdcodex creates a vector embedding for this input space, making it simpler for machine learning practioners to efficiently adapt algorithms for ICD coding.

Our Solution

We rely on the word2vec model to generate this embedding. In this set up, each ICD code represents a “word,” whereas a path sampled from breadth-first or depth-first search represents the “sentence.”

The Team

• Jeremy Adams Fisher

• Alhusain Abdalla

• Natasha Nehra

• Tejas Patel

• Hamrish Saravanakumar

Features

• Curated networkX graphs representing ICD9 and ICD10 hierarchies

• A simple API to generate continuous embeddings for these hierarchies

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

History

0.1.0 (2020-09-04)

• First release on PyPI.

0.3.0 (2020-09-05)

• Finesse API, now consistent between documentation and implementation