Abstract
This paper presents a computational cost analysis for dynamic modeling methods considering its application to real-time biomedical applications. The analyzed methods are Dynamic Bayesian Networks (DBN) and Sequential Independent Component Analysis Mixture Modeling (SICAMM). The results show that the ICA-based methods have a lower computational cost than the BN-based methods. The applicability of these methods to patient monitoring using EEG signals is discussed besides the improvement of the time response by means of parallelization techniques.
