Can PyHealth handle medical code translation?

Yes, PyHealth includes tools to translate and map codes between systems like ICD-9, ICD-10, NDC, RxNorm, ATC, and CCS.

Does it support non-EHR data?

Yes, it supports physiological signals (EEG, ECG), clinical text (medical notes), and medical imaging tasks in addition to structured EHR data.

What kind of models are available in PyHealth?

It supports over 30 models, ranging from baselines like Logistic Regression to advanced healthcare-specific models like RETAIN, SafeDrug, GAMENet, and StageNet.

Which datasets does this skill support?

It provides standardized loading for major healthcare datasets including MIMIC-III, MIMIC-IV, eICU, OMOP, and various physiological signal datasets like SleepEDF.

PyHealth is a specialized Python library for healthcare AI that simplifies processing clinical data, medical coding, and building predictive models for electronic health records.

PyHealth Healthcare AI

Name: PyHealth Healthcare AI
Author: Activer007

byActiver007

データサイエンスとML

Builds and deploys healthcare machine learning models using specialized tools for EHR data, clinical tasks, and medical coding systems.

概要

PyHealth is a comprehensive Python library specifically designed for healthcare AI, offering a modular 5-stage pipeline for clinical machine learning. This skill enables developers to easily process complex healthcare datasets like MIMIC and eICU, translate between medical coding systems (ICD, NDC, ATC), and implement state-of-the-art models for tasks such as mortality prediction and medication recommendation. It streamlines the entire workflow from data loading to deployment, incorporating clinical best practices for interpretability, fairness, and uncertainty quantification.

主な機能

0 GitHub stars
Modular pipeline for efficient data loading, task definition, and model training
33+ pre-built models including healthcare-specific architectures like RETAIN and SafeDrug
Standardized interfaces for 10+ major healthcare datasets including MIMIC and eICU
Integrated clinical evaluation tools for model fairness, calibration, and uncertainty
Automatic translation between ICD, NDC, ATC, and RxNorm medical coding systems

ユースケース

Classifying sleep stages and detecting seizures from physiological EEG signals
Developing medication recommendation systems while avoiding drug-drug interactions
Predicting patient mortality and 30-day hospital readmission risk from EHR data

概要

主な機能

0 GitHub stars
Modular pipeline for efficient data loading, task definition, and model training
33+ pre-built models including healthcare-specific architectures like RETAIN and SafeDrug
Standardized interfaces for 10+ major healthcare datasets including MIMIC and eICU
Integrated clinical evaluation tools for model fairness, calibration, and uncertainty
Automatic translation between ICD, NDC, ATC, and RxNorm medical coding systems

ユースケース

Classifying sleep stages and detecting seizures from physiological EEG signals
Developing medication recommendation systems while avoiding drug-drug interactions
Predicting patient mortality and 30-day hospital readmission risk from EHR data