Markov Chain Approach to Electrocardiogram Modeling and Analysis

摘要

A novel class of models of the electrocardiogram using interacting Markov chains is developed and used as a basis for signal processing. The modeling methodology emphasizes a balance between the inclusion of physiological detail and practicality for signal processing. In order for signal processing algorithms based on the model to achieve accurate, detailed classification of the electrocardiogram, it is necessary to include physiological detail in the model. On the other hand, in order to make the signal processing practical, the models are restricted by imposing spatial, temporal, and hierarchical decompositions. A signal processing algorithm for a wave tracking problem relevant to rhythm classification is proposed. The algorithm is decomposed to mirror the spatial decomposition of the model. On the other hand, in order to make the signal processing practical, the models are restricted by imposing spatial, temporal, and hierarchical decompositions. A signal processing algorithm for a wave tracking problem relevant to rhythm classification is proposed. The algorithm for a wave tracking problem relevant to rhythm classification is proposed. The algorithm for a wave tracking problem relevelant to rhythm classification is proposed. The algorithm for a wave tracking problem relevant to rhythm classification is proposed. The algorithm is decomposed to mirror the spatial decomposition of the model. Limited simulations indicate that reasonable performance may be attainable.