pyhealth

Build and evaluate clinical AI models for EHR, imaging, and text tasks.

21|1|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/OwnLabAI/ownlab --skill pyhealth-ownlabai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pyhealth
Source: https://github.com/OwnLabAI/ownlab/tree/main/mart/skills/scientific-skills/pyhealth
Command: npx skills add https://github.com/OwnLabAI/ownlab --skill pyhealth-ownlabai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

PyHealth provides a comprehensive, modular toolkit for building, training, and evaluating clinical AI models using diverse healthcare data (EHR, physiological signals, imaging, and text), consolidating data handling, model development, and evaluation into one workflow.

Core Features & Use Cases

  • Modular data loading, task definitions, and model selection for EHR, signals, imaging, and text data
  • A full training/evaluation pipeline with calibration, interpretability, and fairness tooling
  • Extensive reference documentation and tutorials to accelerate clinical ML projects

Quick Start

Launch a mortality-prediction workflow on MIMIC-IV using a Transformer, then evaluate and interpret the results.

Frequently Asked Questions about pyhealth

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

FAQPage Schema
How do I build an end-to-end clinical prediction model using EHR data?

PyHealth enables EHR-driven clinical prediction by providing a modular pipeline covering data loading, task definition, model training, and evaluation. You can launch workflows for mortality, readmission, or length of stay predictions using predefined task functions.

What healthcare AI tasks are supported for model training and evaluation?

PyHealth supports EHR-driven prediction (mortality, readmission, length of stay), clinical coding, sleep staging, medical imaging, and clinical text tasks. It includes 20+ predefined task functions for training and evaluating models across diverse healthcare data formats.

Can I use a Transformer model for mortality prediction on MIMIC-IV data?

Yes, PyHealth allows you to launch a mortality-prediction workflow on MIMIC-IV using a Transformer model. It provides a modular pipeline for model selection, training, evaluation, and result interpretation within a single workflow.

Does this toolkit provide interpretability and fairness features for clinical ML models?

PyHealth provides a full training and evaluation pipeline that includes calibration, interpretability, and fairness tooling for clinical ML models. These features are integrated into the workflow to help evaluate and interpret healthcare AI results.

What is the best way to handle diverse healthcare data formats like physiological signals and clinical text?

PyHealth handles diverse healthcare data formats by providing modular data loading and task definitions for EHR, physiological signals, imaging, and text. This consolidates data handling, model development, and evaluation into one workflow.