Resources

The 2016 PhysioNet/CinC Challenge aims to encourage the development of algorithms to classify heart sound recordings collected from a variety of clinical or nonclinical (such as in-home visits) environments. The aim is to identify, from a single sho…

Heart sound segmentation code, based on a duration-dependent hidden Markov model, extended with the use of logistic regression for emission probability estimation and an enhanced Viterbi algorithm.