Michael Rosenberg
University of Colorado Denver/Anschutz Medical Campus, CO·Medicine & Clinical Research
Over a hundred medicines in use today can, in rare cases, disturb the heart's electrical recovery between beats. On an electrocardiogram this shows up as a lengthened QT interval, and it can trigger torsade de pointes, a dangerous rhythm disturbance.
- Fields
Medicine & Clinical Research
Computer Science
Pharmacology
- Topics
QT prolongation
torsade de pointes
drug-induced long QT syndrome
- Grants
- 1 known grant, through Jan 2027
What they work on
- Develop and test a cloud-based deep-learning model that predicts risk of drug-induced long QT syndrome from electronic health record data on hospitalized and outpatient patients, without requiring a prior electrocardiogram.
- Validate genetic predictors of drug-induced QT prolongation using an institutional biobank and an external multi-ethnic population, applying an aggregate polygenic risk score that sums the effects of many DNA variants.
- Genotype participants with a Multi-Ethnic Genotyping Array designed for imputation across non-European ancestries, so that risk prediction holds across diverse patient populations.
- Build and evaluate clinical decision support tools that combine the deep-learning and genetic predictions to reduce prescribing of QT-prolonging drugs to at-risk patients, using data mapped to the Observational Medical Outcomes Partnership common data model for portability to other institutions.
More about the research
Because QT-prolonging drugs are prescribed across virtually every medical specialty, often for conditions unrelated to the heart, clinicians need a reliable way to know in advance which patients are at risk. This project builds and tests software that makes that prediction at the moment of prescribing.
Why it matters
Drug-induced QT prolongation is an uncommon but potentially fatal side effect of widely prescribed medications. A tool that flags high-risk patients before a prescription is written could let clinicians choose a safer alternative when one exists, or monitor more closely when it does not. Because the models are built on a shared data format and a multi-ethnic genotyping platform, they are designed to be reusable at other health systems and across ancestry groups, and to serve as a template for decision support in other conditions.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Jacqueline Stocking
University of California-Davis, CA·Medicine & Clinical Research·Early-career
Optimize and validate an automated, EHR-based model predicting postoperative respiratory failure, using surgical patient data from 2014 to 2021 and pre- and intra-operative risk factors.
Nosheen Reza
University of Pennsylvania, PA·Medicine & Clinical Research·Early-career
Build and validate a machine learning model that detects hypertrophic cardiomyopathy from electrocardiograms in Penn Medicine electronic health record data, and test whether adding other EHR-derived traits improves detection.
Sunny Lou
Washington University in St Louis, MO·Medicine & Clinical Research·Early-career
Embed S-PATH, a personalized AI model that predicts a patient's risk of needing a surgical blood transfusion, into the electronic health record as a clinical decision support tool for presurgical blood orders.
The school
University of Colorado Denver/Anschutz Medical Campus: 41 labs on PI Pages, $766M a year on research, 205 research doctorates a year.
More about University of Colorado Denver/Anschutz Medical Campus
Reach out
A start for your email. Fill in the brackets.