Jonas Lynggaard Isaksen - Postdoctoral Fellowship 2023
Project summary:
AI–triaging of patients with possible acute myocardial infarction
We hypothesize, that we currently perform too many and too few invasive coronary angiograms – in other words we sometimes perform angoigrams on the wrong patients.
Undertreatment: Up to 1300 patients a year without so-called ST-elevations on the electrocardiogram (ECG) may have a total occlusions of one or more coronary arteries.
Overtreatment: ST-elevations may arise from well-known conditions that are not acute myocardial infarction. Up to 1200 patients yearly may receive unnecessary “rule-out” angiograms.
Project Title
AI–triaging of patients with possible acute myocardial infarction
Background
Undertreatment – or delayed treatment – may lead to a higher incidence of reinfarction, heart failure and mortality. Overtreatment – on the individual level – means exposure to unnecessary procedure-related risks (e.g. stroke) whereas for the society overtreatment is unnecessary costs.
Aim
We aim to identify a large proportion of those currently being undertreated or overtreated. To do so, we will apply machine learning methods to a large dataset of ECGs. The product will be an algorithm that provides decision support for clinicians to solve today’s problems in acute myocardial infarction.
Methods
Phase 1: Retrospective, cross-sectional
Phase 2: Blinded prospective evaluation
Phase 3: Prospective randomized controlled trial
Jonas Lynggaard Isaksen
- MSc and PhD
- Department of Biomedical Sciences, University of Copenhagen
- The Heart Centre, Department of Cardiology, Copenhagen University Hospital Rigshospitalet
Mentors:
Jørgen K. Kanters, University of Copenhagen
Henning Bundgaard, Rigshospitalet