pyhealth

Name: pyhealth
Availability: InStock
Author: jaechang-hits

Load EHR data from MIMIC-III/IV and OMOP-CDM for ML model training.

298|27|Updated Feb 18, 2026

Search Tags:#ehr #healthcare data #mimic #omop #clinical machine learning #patient prediction

Ready to run

npx skills add https://github.com/jaechang-hits/SciAgent-Skills --skill pyhealth-jaechang-hits

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pyhealth
Source: https://github.com/jaechang-hits/SciAgent-Skills/tree/main/skills/biostatistics/pyhealth
Command: npx skills add https://github.com/jaechang-hits/SciAgent-Skills --skill pyhealth-jaechang-hits

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pyhealth, torch, pandas, scikit-learn, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill streamlines the process of building machine learning models using Electronic Health Record (EHR) data, enabling clinical prediction and discovery.

Core Features & Use Cases

EHR Data Processing: Load and process data from MIMIC-III/IV, eICU, and OMOP-CDM.
Clinical Prediction Models: Train models for mortality prediction, drug recommendation, readmission, and more.
Medical Code Standardization: Handle and map various medical codes (ICD, ATC, NDC).
Use Case: Predict patient mortality risk using their historical medical records from MIMIC-IV.

Quick Start

Load the MIMIC-III dataset and print the number of patients and visits.

Frequently Asked Questions about pyhealth

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I train clinical machine learning models using MIMIC-IV data?▼

You can train clinical machine learning models using MIMIC-IV by loading the dataset into the workflow, preprocessing patient records, and fitting pre-built models for tasks like mortality prediction and readmission forecasting.

Can I process OMOP-CDM structured EHR databases for patient prediction tasks?▼

Yes, you can process OMOP-CDM structured EHR databases for patient prediction tasks. The workflow supports loading and mapping OMOP-CDM data to train models for mortality risk and drug recommendation.

How do I handle medical code standardization for ICD and NDC codes in healthcare data?▼

Medical code standardization for ICD, ATC, and NDC codes in healthcare data is handled through integrated processing tools that map various medical codes for consistent model training.

Does this workflow require PyTorch and scikit-learn for EHR data processing?▼

Yes, EHR data processing and model training require PyTorch and scikit-learn as core dependencies, alongside pandas for data manipulation to execute the clinical machine learning pipeline.

What is the best way to predict patient mortality risk from historical medical records?▼

The best way to predict patient mortality risk from historical medical records is using pre-built clinical ML models that ingest structured EHR data from sources like MIMIC-III or MIMIC-IV to forecast outcomes.

What clinical prediction tasks are supported for EHR data besides mortality forecasting?▼

Besides mortality forecasting, supported clinical prediction tasks for EHR data include drug recommendation and readmission prediction, utilizing historical patient visits and standardized medical codes.