Robust and Accurate Anomaly Detection in ECG Artifacts Using Time Series Motif Discovery

<div>Anomaly detection results among three algorithms on MITDB dataset. (a), (b), (c), and (d) are results of our proposed RAAD, BFDD, HOT SAX, and BitClusterDiscord, respectively. RAAD produced correct results whereas BFDD, HOT SAX, and BitClusterDiscord produced incomplete results along with some false alarm.</div>

Computational and Mathematical Methods in Medicine

fig11

Figure 11

Figure 11: Robust and Accurate Anomaly Detection in ECG Artifacts Using Time Series Motif Discovery 