Continuous Detection of Abnormal Heartbeats from ECG Using Online Outlier Detection

Detecting abnormal heartbeats from an electrocardiogram (ECG) signal is an important problem studied extensively and yet is a difficult problem that defies a viable working solution, especially on a mobile platform which requires computationally efficient and yet accurate detection mechanism. In this project, a prototype system has been built to test the feasibility and efficacy of detecting abnormal ECG segments from an ECG data stream targeting a mobile device, where data are arriving continuously and indefinitely and are processed online incrementally and efficiently without being stored in memory. The processing comprises three steps: (i) segmentation using R peak detection, (ii) feature extraction using discrete wavelet transform, and (iii) outlier detection using incremental online microclustering. Experiments conducted using real ambulatory ECG datasets showed satisfactory accuracy. In addition, comparing personalized detection (tuned separately for each patient’s ECG datasets) and non-personalized detection (tuned aggregated over all patients’ datasets) confirms a definite advantage of personalized detection for ECG.

[1]  Fabrizio Angiulli,et al.  Distance-based outlier queries in data streams: the novel task and algorithms , 2010, Data Mining and Knowledge Discovery.

[2]  Yannis Manolopoulos,et al.  Continuous monitoring of distance-based outliers over data streams , 2011, 2011 IEEE 27th International Conference on Data Engineering.

[3]  Ira Assent,et al.  AnyOut: Anytime Outlier Detection on Streaming Data , 2012, DASFAA.

[4]  Saurabh Singh,et al.  ECG Arrhythmia Classification Using Spearman Rank Correlation and Support Vector Machine , 2011, SocProS.

[5]  H. Bazett,et al.  AN ANALYSIS OF THE TIME‐RELATIONS OF ELECTROCARDIOGRAMS. , 1997 .

[6]  Szi-Wen Chen,et al.  A moving average based filtering system with its application to real-time QRS detection , 2003, Computers in Cardiology, 2003.

[7]  A. Bensaid,et al.  Classification of ECG patterns using fuzzy rules derived from ID3-induced decision trees , 1998, 1998 Conference of the North American Fuzzy Information Processing Society - NAFIPS (Cat. No.98TH8353).

[8]  Lovekesh Vig,et al.  Anomaly detection in ECG time signals via deep long short-term memory networks , 2015, 2015 IEEE International Conference on Data Science and Advanced Analytics (DSAA).

[9]  Raymond T. Ng,et al.  Distance-based outliers: algorithms and applications , 2000, The VLDB Journal.

[10]  Raymond T. Ng,et al.  Algorithms for Mining Distance-Based Outliers in Large Datasets , 1998, VLDB.

[11]  Matthew O. Ward,et al.  Neighbor-based pattern detection for windows over streaming data , 2009, EDBT '09.

[12]  Bharadwaj Veeravalli,et al.  Real-Time, Personalized Anomaly Detection in Streaming Data for Wearable Healthcare Devices , 2017, Handbook of Large-Scale Distributed Computing in Smart Healthcare.

[13]  Pablo Laguna,et al.  Multilead Analysis of T-Wave Alternans in the ECG Using Principal Component Analysis , 2009, IEEE Transactions on Biomedical Engineering.

[14]  Cyrus Shahabi,et al.  Distance-based Outlier Detection in Data Streams , 2016, Proc. VLDB Endow..

[15]  Yannis Manolopoulos,et al.  Continuous outlier detection in data streams: an extensible framework and state-of-the-art algorithms , 2013, SIGMOD '13.

[16]  Jan Macek,et al.  Incremental learning of ensemble classifiers on ECG data , 2005, 18th IEEE Symposium on Computer-Based Medical Systems (CBMS'05).

[17]  Peter W. Macfarlane,et al.  Artificial neural networks for the diagnosis of atrial fibrillation , 1993 .

[18]  Marta Karczewicz,et al.  ECG data compression by spline approximation , 1997, Signal Process..

[19]  P. Karthigaikumar,et al.  ECG Signal Preprocessing and SVM Classifier-Based Abnormality Detection in Remote Healthcare Applications , 2018, IEEE Access.

[20]  A. Houghton,et al.  MAKING SENSE OF THE ECG A HANDS-ON GUIDE , 2008 .

[21]  Abhilasha M. Patel,et al.  Real Time ECG Feature Extraction and Arrhythmia Detection on a Mobile Platform , 2012 .