ARTIFICIAL INTELLIGENCE METHODS IN ELECTROCARDIOGRAM AND ELECTROENCEPHALOGRAM DATA CLUSTERING

The paper focuses on the field of artificial intelligence techniques and their use in biomedical data processing. It concerns the clustering techniques inspired by various ant colonies. The behavior of ant colonies shows many interesting properties that have been used in static and dynamic combinatorial problem-solving tasks (mostly since 1990). Also applications to data clustering have been proposed. This branch is a subject of ongoing research. After the introduction into the state-of-the-art of ant-colony-inspired metaheuristics, an overview of ant-colony-inspired clustering metaheuristics is presented, together with the ACO(=)DTree method, developed by the first author, which is based on the autocatalytic collective behavior of real insect colonies. Over the basic algorithm it involves techniques to increase robustness and performance of the method. Application to electrocardiogram and electroencephalogram data processing is also presented, together with comparison to other clustering methods.