Machine learning to predict rapid progression of carotid atherosclerosis in patients with impaired …

The novel finding of our study is that machine learning methods can be applied to a limited dataset typical of a clinical trial in order to predict impaired glucose tolerance subjects who will develop rapid carotid plaque progression with overall good performance. Our results demonstrate the potential utility of sophisticated Bayesian approaches in predicting clinical events from limited clinical datasets. In 2010, approximately 1 in 3 adults in the USA or about 79 million people had prediabetes [23], which includes IGT and impaired fasting glucose. Aside from the risk for developing diabetes, prediabetes by itself is also independently associated with future risk of stroke [24]. It is therefore critical that we develop tools for early identification of at-risk patients who might benefit from targeted early intervention, both non-pharmacologic and pharmacologic.…


Link to Full Article: Machine learning to predict rapid progression of carotid atherosclerosis in patients with impaired …