Peter Robinson, MD, MSc
Professor of computational biology
A sandbox project designed to create a best practices platform for deploying and evaluating clinical machine learning tools and algorithms. Goals include provisioning community-vetted solutions to common clinical machine learning challenges, including data preparation, analysis of bias sources, and evaluation/validation of algorithms.
This project provides a tool to search for biomarkers in EHR data (such as LOINC) of participating CTSA centers for conversion to Human Phenotype Ontology (HPO) codes, and will extend to additional phenotype sources in the EHR.
This ongoing HL7 Community Engagement project focuses on clinical data model harmonization and terminology services for the CTSA hubs. The three areas of work being conducted within this project are:
 Common Data Model Harmonization (CDMH) II, a cross-agency (FDA, ONC, CDC, NCATS, NCI) effort to harmonize the most common clinical data models;  Healthcare Open Terminology (HOT) Ecosystem, a multi-agency (NCATS, NCI, multi-hub) effort to harmonize value sets and terminologies in a sustainable ISO format;  HL7 Vulcan Accelerator, a project that defines or enhances FHIR resources that connect real world data with regulated submissions.