Machine-learning to estimate cardiac death risk
Machine learning can be used to predict a patient’s risk of cardiac death based on a combination of clinical variables.
Researchers from the Illinois Institute of Technology (Chicago, USA) and Cedars-Sinai Medical Center (Los Angeles, USA) have developed machine-learning models to estimate a patient’s risk of cardiac death based on adenosine myocardial perfusion SPECT and associated clinical data. The group demonstrated an approach to visually convey the reasoning behind a patient’s risk, thus providing insight to clinicians beyond that of a “black box.” Combined with presenting the reasoning behind the risk scores, their results suggest that machine learning can be more effective than logistic regression when estimating a patient’s risk of cardiac death. The findings appear in “Prediction of Cardiac Death after Adenosine Myocardial Perfusion SPECT Based on Machine Learning” in the Journal of Nuclear Cardiology.
By: Ben Bishop, Journal of Nuclear Cardiology