# sunlabuiuc/PyHealth

A Deep Learning Python Toolkit for Healthcare Applications.

Repository: https://github.com/sunlabuiuc/PyHealth
Canonical: https://ross.abutalabs.com/products/pyhealth
Homepage: https://pyhealth.dev/ 
Language: Python
License: MIT
License Family: permissive
Topics: healthcare, data-mining, deep-learning, preprocessing, clinical-data, clinical-research, electronic-medical-record, medical-code, electronic-health-record
Last push: 2026-08-20T18:47:33+00:00

## Health v2 (maintenance only)
Score: 91/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 77, longevity 100
- inputs: {"age_days": 2221, "days_push": 13, "days_rel": 155, "gap_med": 8, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1650, forks 788 (observed 2026-08-28T04:05:16.875365+00:00)

## What it is
PyHealth is an open-source Python toolkit for clinical deep learning, unifying healthcare datasets (EHRs, physiological signals, medical imaging), prediction tasks, and ML models in a simple pipeline. It supports tasks like mortality prediction, readmission, and drug recommendation across 20+ clinical datasets with 50+ models.

## Use cases
- train a mortality prediction model on MIMIC-III EHR data
- predict hospital readmission from electronic health records
- build drug recommendation models from clinical data
- process and preprocess EHR datasets for deep learning
- run clinical ML benchmarks on physiological signals and medical imaging
- prototype healthcare AI pipelines in under 15 lines of code

## When to choose
- you need reproducible clinical deep learning pipelines on standard EHR datasets like MIMIC
- you want prebuilt clinical tasks and models instead of writing data loaders from scratch
- you work in healthcare ML research and need a unified dataset/task/model API

## When to avoid
- you need production clinical deployment with regulatory compliance rather than research tooling
- your data is not healthcare-related
- you need a no-code or GUI tool rather than a Python library

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, etl, data-science
- domain: healthcare, machine-learning, deep-learning, data-science
- platform: python
- tags: ehr, clinical-data, mimic, medical-imaging, healthcare-ai, clinical-deep-learning

## Member repositories
- sunlabuiuc/PyHealth (main) score 91

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:16.875365+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:45:06.196663+00:00, confidence not recorded.
  - readme: https://github.com/sunlabuiuc/PyHealth (fetched 2026-08-28T04:05:16.875365+00:00, sha 6d4dfac02429)
  - homepage: https://pyhealth.dev/  (fetched 2026-08-29T11:18:15.711931+00:00, sha 47242d83e583)
  - registry_pypi: https://pypi.org/pypi/pyhealth/json (fetched 2026-08-29T11:18:15.714983+00:00, sha 7bf3b1e1949b)
- Data as of 2026-08-30T08:39:29.467469+00:00.