skills.rest
silverstein
silversteinCOMMUNITY
Explore all skills by this author →
Website𝕏@matsilversteinGitHub

pyhealth

Build and evaluate healthcare prediction models using standardized clinical datasets.

21|2|Updated Dec 8, 2025
Search Tags:#model training#interpretability#healthcare ai#clinical prediction#medical coding#ehr machine learning#clinical datasets
One-click install
npx skills add https://github.com/silverstein/claude-scientific-skills-desktop --skill pyhealth-silverstein
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pyhealth
Source: https://github.com/silverstein/claude-scientific-skills-desktop/tree/main/corpus/pyhealth
Command: npx skills add https://github.com/silverstein/claude-scientific-skills-desktop --skill pyhealth-silverstein

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

PyHealth streamlines the process of building, training, and deploying machine learning models on clinical data by providing standardized datasets, tasks, models, and evaluation workflows tailored to healthcare scenarios.

Core Features & Use Cases

  • Healthcare AI toolkit for EHR and clinical ML: supports mortality, readmission, drug recommendation, medical coding, and physiologic-signal modeling using common clinical prediction pipelines.
  • Modular data-to-model pipeline: handles dataset loading, task definition (input/output schemas), model selection (baselines to healthcare-specific), training, evaluation, calibration, and interpretability.
  • Reference-driven guidance: includes organized documentation for datasets, medical code translation, tasks, models, preprocessing, and training/evaluation best practices.

Quick Start

Use the pyhealth skill to build an ICU mortality prediction pipeline on MIMIC-IV by loading the dataset, selecting the appropriate mortality task, choosing a model such as Transformer or RETAIN, training with the Trainer class, and evaluating with AUROC/AUPRC plus calibration and interpretability checks.

Frequently Asked Questions about pyhealth

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

FAQPage Schema
How do I train machine learning models for clinical prediction tasks using EHR data?▼

Yes, you can perform ICD, NDC, ATC, and CCS medical coding translation by mapping clinical datasets to coding tasks via standardized input/output schemas and selecting compatible healthcare-specific models for training and evaluation.

What healthcare prediction tasks are supported for EHR machine learning?▼

Supported EHR machine learning tasks include mortality and readmission prediction, medication and drug recommendation, medical coding translation, sleep staging, EEG event detection, and clinical text classification using standardized datasets.

Can I build an ICU mortality prediction pipeline on MIMIC-IV?▼

You can build an ICU mortality prediction pipeline on MIMIC-IV by loading the dataset, selecting the mortality task, choosing a model like Transformer or RETAIN, training with the Trainer class, and evaluating with AUROC, AUPRC, calibration, and interpretability checks.

Do I need interpretability and calibration checks for clinical ML models?▼

Interpretability and calibration checks are required components of the clinical ML workflow, ensuring that trained healthcare prediction models produce reliable, fair, and interpretable results for clinical decision-making scenarios.

How do I select compatible models for different clinical data modalities?▼

Model selection requires matching the model architecture to the specific task type and data modality, ranging from baseline models to healthcare-specific architectures designed for physiologic signals, imaging, text, and EHR datasets.

Related Skills

Explore More Education & Research Skills→
rubenslivrubensliv
15

pyhealth

Build, train, and deploy clinical prediction models on EHR, imaging, and text data.

Community
Advanced
HolobiomicsLabHolobiomicsLab
4

pyhealth

Develop and deploy clinical machine learning models for EHR data.

Official
Advanced
ovachieverovachiever
52

pyhealth

Convert raw clinical data and task specifications into model-ready healthcare AI inputs.

Community
Advanced
Zeyuyang-0420Zeyuyang-0420

pyhealth

Automate 5-stage clinical deep learning pipelines with PyHealth and PyTorch.

Community
Advanced