Dynamical modeling of respiratory sound an aproach for pulmunary patients classification

Investigating novel efficient feature extraction approach applied to the lung sound signal is a necessary method to improve the performance of respiratory abnormality recognition. Owing to the fact that the Recurrence Quantification Analysis (RQA) is a proper approach for insight into dynamic system, this paper proposed a new method for feature extraction from the lung sound signals. The method is based upon modeling the complexity measure of the system in time domain, by developed Recurrent Fuzzy Function (RFFs) approach. Experimental results show that by considering three subspaces for RFFs model, interactive and recurrent weights can appropriately differentiate lung abnormality and outperform other feature extraction method.