Sparse estimation: applications in atrial fibrillation

Finding important links between large numbers of factors is a topical issue. In this study, a mathematical method was developed allowing dominant links between a certain outcome and many possible predictive factors, or key interactions in a large complex network to be discovered by using relatively few data. This can be achieved by assuming that among the many possible links only a few are actually meaningful. This method has been successfully applied to, among other things, discovering key factors playing a role in the effectiveness of treatment strategies for atrial fibrillation, a type of cardiac arrhythmia.