Biomedical time series clustering based on non-negative sparse coding and probabilistic topic model

• Published in: Computer methods and programs in biomedicine Vol. 111; no. 3; pp. 629 - 641
• Main Authors: Wang, Jin, Liu, Ping, F.H.She, Mary, Nahavandi, Saeid, Kouzani, Abbas
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ireland Ltd 01.09.2013; Elsevier
• Subjects: Applied sciences; Artificial intelligence; Bag-of-Words; Biological and medical sciences; Cluster Analysis; Computer science; control theory; systems; Connectionism. Neural networks; Data processing. List processing. Character string processing; Electrocardiography; Electrocardiography. Vectocardiography; Electrodiagnosis. Electric activity recording; Electroencephalography; Exact sciences and technology; Humans; Internal Medicine; Investigative techniques, diagnostic techniques (general aspects); Medical sciences; Memory organisation. Data processing; Models, Statistical; Nervous system; Other; Probabilistic topic model; Probability; Signal-To-Noise Ratio; Software; Sparse coding; Unsupervised learning; Sparse coding; Probabilistic topic model; Bag-of-Words; Unsupervised learning; Cluster analysis; Dictionaries; Text analysis; Modeling; Bag of words; Classification; Electrocardiography; Sparse representation; Diagnosis; Document analysis; Biomedical engineering; Statistical analysis; Probabilistic approach; Time series; Cluster; Information retrieval; Text; Document management; Neural network; Similarity solution; Long term; Biosensor
• ISSN: 0169-2607; 1872-7565
• DOI: 10.1016/j.cmpb.2013.05.022

Abstract Biomedical time series clustering that groups a set of unlabelled temporal signals according to their underlying similarity is very useful for biomedical records management and analysis such as biosignals archiving and diagnosis. In this paper, a new framework for clustering of long-term biomedical time series such as electrocardiography (ECG) and electroencephalography (EEG) signals is proposed. Specifically, local segments extracted from the time series are projected as a combination of a small number of basis elements in a trained dictionary by non-negative sparse coding. A Bag-of-Words (BoW) representation is then constructed by summing up all the sparse coefficients of local segments in a time series. Based on the BoW representation, a probabilistic topic model that was originally developed for text document analysis is extended to discover the underlying similarity of a collection of time series. The underlying similarity of biomedical time series is well captured attributing to the statistic nature of the probabilistic topic model. Experiments on three datasets constructed from publicly available EEG and ECG signals demonstrates that the proposed approach achieves better accuracy than existing state-of-the-art methods, and is insensitive to model parameters such as length of local segments and dictionary size.