Can I use this for real-time clinical data processing?

PyHealth is primarily designed for building and validating models on healthcare datasets, but its standardized pipelines can be adapted for clinical decision support systems.

What is the PyHealth Claude Code Skill?

It is a specialized extension for Claude Code that provides guidance, code patterns, and best practices for using the PyHealth library to develop healthcare-focused machine learning applications.

Which medical coding systems are supported?

The skill supports comprehensive mapping and translation between ICD-9, ICD-10, NDC, RxNorm, ATC, CCS, and several other standardized medical terminologies.

Does it support physiological signal analysis?

Yes, it includes specialized tools for loading and processing signals such as EEG and ECG for tasks like sleep staging and seizure detection.

How does it handle model interpretability in healthcare?

It provides implementation patterns for interpretable architectures like RETAIN and tools for visualization, SHAP analysis, and uncertainty quantification to ensure clinical trust.

PyHealth AI Toolkit

Name: PyHealth AI Toolkit
Author: pur3v4d3r

bypur3v4d3r

•

Ciencia de Datos y ML

Develops, tests, and deploys healthcare-specific machine learning models using clinical data and electronic health records.

Acerca de

PyHealth is a comprehensive framework designed to streamline the end-to-end pipeline for clinical research and healthcare AI development. This skill enables Claude to assist with processing complex electronic health records (EHR), translating between medical coding systems like ICD and NDC, and implementing state-of-the-art clinical models such as RETAIN or SafeDrug. Whether you are working with large-scale datasets like MIMIC-IV and eICU or analyzing physiological signals like EEG and ECG, PyHealth provides the specialized data structures and evaluation metrics necessary to ensure clinical rigor, fairness, and interpretability in medical AI applications.

Características Principales

Predefined clinical tasks for mortality prediction, hospital readmission, and length of stay analysis.
Seamless translation and mapping between medical coding systems like ICD-9/10, NDC, RxNorm, and ATC.
Standardized data loading for 10+ clinical datasets including MIMIC-III/IV, eICU, and OMOP.
Access to 33+ specialized healthcare models including GNNs, Transformers for EHR, and drug recommendation architectures.
Advanced clinical evaluation tools for fairness, calibration, and model interpretability.
1 GitHub stars

Casos de Uso

Building medication recommendation systems that account for drug-drug interaction constraints.
Predicting patient mortality and 30-day readmission risks using longitudinal EHR data.
Classifying sleep stages and detecting seizures from multi-channel physiological signals.

Acerca de

Características Principales

Predefined clinical tasks for mortality prediction, hospital readmission, and length of stay analysis.
Seamless translation and mapping between medical coding systems like ICD-9/10, NDC, RxNorm, and ATC.
Standardized data loading for 10+ clinical datasets including MIMIC-III/IV, eICU, and OMOP.
Access to 33+ specialized healthcare models including GNNs, Transformers for EHR, and drug recommendation architectures.
Advanced clinical evaluation tools for fairness, calibration, and model interpretability.
1 GitHub stars

Casos de Uso

Building medication recommendation systems that account for drug-drug interaction constraints.
Predicting patient mortality and 30-day readmission risks using longitudinal EHR data.
Classifying sleep stages and detecting seizures from multi-channel physiological signals.