Cardea: An Open Automated Machine Learning Framework for Electronic Health Records

TL;DR

Cardea is an open-source automated machine learning framework designed for electronic health records, enabling efficient prediction modeling with standardized data handling, automated feature engineering, and comprehensive auditing, demonstrated on real datasets.

Contribution

Introduces Cardea, a flexible, open-source AutoML framework tailored for EHR data, integrating FHIR standards and automated processes for health prediction tasks.

Findings

Effective on MIMIC-III and Kaggle datasets

Matches human performance in prediction tasks

Flexible and user-friendly system

Abstract

An estimated 180 papers focusing on deep learning and EHR were published between 2010 and 2018. Despite the common workflow structure appearing in these publications, no trusted and verified software framework exists, forcing researchers to arduously repeat previous work. In this paper, we propose Cardea, an extensible open-source automated machine learning framework encapsulating common prediction problems in the health domain and allows users to build predictive models with their own data. This system relies on two components: Fast Healthcare Interoperability Resources (FHIR) -- a standardized data structure for electronic health systems -- and several AUTOML frameworks for automated feature engineering, model selection, and tuning. We augment these components with an adaptive data assembler and comprehensive data- and model- auditing capabilities. We demonstrate our framework via 5 prediction tasks on MIMIC-III and Kaggle datasets, which highlight Cardea's human competitiveness, flexibility in problem definition, extensive feature generation capability, adaptable automatic data assembler, and its usability.