Combination of Hedge Algebra and Type-2 Fuzzy System for Electrocardiogram Signal Recognition and Classification

Abstract

The paper presents a new application of Hedge Algebra in Type 2 (HAT2) Fuzzy System (FS) for electrocardiographical signal recognition and classification. The HAT2 integrates the capability of fuzzy logic in modeling uncertainties of data and the linguistic variables to describe and manipulate the knowledge in a way closer to the way humans describe facts and rules. An adaptive learning algorithm will be also proposed to train the system’s parameters for better fitting to the data samples. The proposed approach will be tested with the heart beat classification problem with different types of arrhythmias in the ECG signals taken from the Arrhythmia Databases of Massachusetts Institute of Technology and Boston's Beth Israel Hospital. The numerical results will be compared with other methods to show the high quality of proposed solution. The proposed solution is also shown being simple with low complexity of computation, which makes it suitable for use in IoT or portable devices for intelligent solutions used in modern healthcare systems.