ehr-phenotyping

Repository for the GA4GH phenopacket schema

[ML4H 2022] This is the code for our paper `Counterfactual and Factual Reasoning over Hypergraphs for Interpretable Clinical Predictions on EHR'.

How to: create SNOMED CT codelists

using AI model to infer patient phenotypes from identified named entities (instances of biomedical concepts)

Code accompanying our ACM CHIL paper: "TASTE: Temporal and Static Tensor Factorization for Phenotyping Electronic Health Records"

The OMOP Common Data Model(v6) implemented in Django (v3)

How to: find acute exacerbation of COPD events in UK primary care electronic healthcare records

Cast different EHR (electronic health record) layers to a shared latent space to identify patient subtypes

How to: create drug codelists for recorded prescriptions

RETIRED - Clinical phenotyping and ETL system

Lecture slides on Electronic Health Record Data Quality

Repositorio del proyecto "Detección automática de consultas para apoyar la atención de la salud sexual y reproductiva de personas con discapacidad"

Introduction to PERMIT project resources

This project shows a method with a topic modeling approach: non-negative matrix factorization (NMF), used in EHR for subphenotyping and identifying the relationship between phenotypes and genetic variants

Quality and Outcomes Framework (QOF) notes

Synthetic EHR cancer cohort phenotyping and validation workflow using longitudinal healthcare data and Python analytics.

Diffusion-based representation learning for extracting meaningful patient embeddings from longitudinal EHR data.

PheMINE is a reproducible, automated workflow for analyzing electronic health records. It streamlines cohort construction, performs permutation-based phecode enrichment analysis, and trains supervised learning models (Random Forest, XGBoost, Neural Networks) for disease prediction.

A research framework for longitudinal EHR sequence modeling, exploring multiple temporal representations, learning objectives, and model classes for disease progression, survival analysis, and temporal phenotyping under censoring and irregular follow-up.