Can this skill help with medical code mapping?

Yes, it provides specialized tools for translating and mapping between various medical coding systems such as ICD-9/10, NDC, RxNorm, ATC, and CCS.

How does PyHealth handle patient privacy during data splitting?

The skill follows best practices by splitting data at the patient level, ensuring no individual patient's data is shared across training, validation, and test sets to prevent data leakage.

What clinical prediction tasks can I perform?

You can implement tasks like mortality prediction, length of stay estimation, hospital readmission risk, drug recommendation, and physiological signal classification.

What datasets does the PyHealth skill support?

PyHealth supports over 10 major healthcare datasets, including MIMIC-III, MIMIC-IV, eICU, OMOP, and various physiological signal datasets like SleepEDF.

Does it include specific clinical models?

Yes, it features over 33 models, including healthcare-specific architectures like RETAIN for interpretability, SafeDrug for medication recommendation, and StageNet.

PyHealth Clinical AI Toolkit

Name: PyHealth Clinical AI Toolkit
Author: BbgnsurfTech

byBbgnsurfTech

•

データサイエンスとML

Builds, tests, and deploys specialized machine learning models for healthcare applications using clinical data and electronic health records.

PyHealth is a comprehensive toolkit designed to streamline the development of healthcare-specific AI models, offering a modular pipeline for data loading, task definition, and model evaluation. It provides specialized support for complex medical datasets like MIMIC-IV and eICU, automates medical code translation (ICD-9/10, NDC, ATC), and includes pre-built implementations of state-of-the-art clinical models such as RETAIN and SafeDrug. This skill is essential for researchers and developers working on mortality prediction, medication recommendation, or physiological signal analysis who require domain-specific best practices and standardized clinical workflows.

主な機能

01Automated medical coding translation between ICD, NDC, RxNorm, and ATC systems

02Advanced evaluation tools for clinical fairness, model calibration, and interpretability

037 GitHub stars

04Library of 33+ pre-implemented models ranging from deep learning baselines to healthcare-specific architectures

05Support for 10+ major clinical datasets including MIMIC-III, MIMIC-IV, eICU, and OMOP

06Modular 5-stage pipeline for healthcare machine learning from data loading to deployment

ユースケース

01Classifying sleep stages and detecting seizures from physiological signals like EEG and ECG

02Predicting ICU patient mortality and hospital readmission risks using longitudinal EHR data

03Developing safe medication recommendation systems that account for drug-drug interactions

主な機能

01Automated medical coding translation between ICD, NDC, RxNorm, and ATC systems

02Advanced evaluation tools for clinical fairness, model calibration, and interpretability

037 GitHub stars

04Library of 33+ pre-implemented models ranging from deep learning baselines to healthcare-specific architectures

05Support for 10+ major clinical datasets including MIMIC-III, MIMIC-IV, eICU, and OMOP

06Modular 5-stage pipeline for healthcare machine learning from data loading to deployment

ユースケース

01Classifying sleep stages and detecting seizures from physiological signals like EEG and ECG

02Predicting ICU patient mortality and hospital readmission risks using longitudinal EHR data

03Developing safe medication recommendation systems that account for drug-drug interactions