ASSIST© Published in the European Heart Journal

Led by Evangelos Oikonomou, MD, DPhil, an internal medicine resident in the Physician-Scientist Training Program at Yale, the novel tool ASSIST© was published in the European Heart Journal. The tool applies machine learning to the clinical trial data from the PROMISE trial - the largest clinical trial comparing functional and anatomical testing in suspected coronary artery disease. The functional tests include stress tests, and anatomical tests include cardiac CT angiography, a non-invasive CT scan. The study embedded local data experiments in the multidimensional representations of the clinical trial population. These patient representations were based on patient characteristics before they were randomized to the testing strategy, ensuring these local data experiments were unbiased for the choice of the diagnostic strategy. The tool is based on 12 commonly available patient characteristics and is available here.