A study in Radiology investigated whether a machine learning model can predict patient survival of heart failure in pulmonary hypertension.
The researchers identified 256 patients who were diagnosed with pulmonary hypertension. Each of these patients underwent cardiac MRI, right-sided heart catheterization and six-minute walk testing with a median follow-up of four years. The researchers used the cardiac MRI to create 3-D models of right ventricular motion, which a machine learning algorithm used to identify patterns predictive of survival.
By the end of the research period, 36 percent of the patients had died. The researchers determined that usage of the machine learning and 3-D cardiac motion model improved survival prediction when used alongside traditional methods, such as imaging and hemodynamic, functional and clinical markers.
The researchers concluded that the machine learning predictions were effective, writing: "A machine-learning survival model that uses three-dimensional cardiac motion predicts outcome independent of conventional risk factors in patients with newly diagnosed pulmonary hypertension."