Computational and Mathematical Methods in Medicine

Research Article

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

Table 6

The best value, mean (μ) and standard deviation (SD) of AoD, sensitivity, positive predictive value, specificity, and false alarm results obtained by RAAD and BitClusterDiscord with various overlap criteria. Bold figures denote the wining algorithms.


	AoD					Sensitivity					Positive predictive value					Specificity					False alarm rate
	BitClusterDiscord			RAAD		BitClusterDiscord			RAAD		BitClusterDiscord			RAAD		BitClusterDiscord			RAAD		BitClusterDiscord			RAAD
	40%	80%	Car.	80%	Car.	40%	80%	Car.	80%	Car.	40%	80%	Car.	80%	Car.	40%	80%	Car.	80%	Car.	40%	80%	Car.	80%	Car.

MITDB	85.59	85.59	0	*99.10*	*99.10*	*100*	*100*	0	*100*	*100*	*100*	*100*	0	*100*	*100*	*100*	*100*	87.5	*100*	*100*	0	0	12.5	0	0
ITDB
Best	62.79	0	0	89.11	89.11	100	0	0	100	100	100	0	0	100	100	100	77.78	77.78	100	100	0	22.22	22.22	0	0
μ	58.56	0	0	89.11	89.11	100	0	0	100	100	100	0	0	100	100	100	77.78	77.78	100	100	0	22.22	22.22	0	0
SD	4.23	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
ITDB
Best	83.76	83.76	0	89.11	89.11	100	100	0	100	100	100	100	0	100	100	100	100	77.78	100	100	0	22.22	22.22	0	0
μ	69.7	41.88	0	89.11	89.11	100	50	0	100	100	100	50	0	100	100	100	88.89	77.78	100	100	0	11.11	22.22	0	0
SD	14.05	41.88	0	0	0	0	50	0	0	0	0	50	0	0	0	0	11.11	0	0	0	0	11.11	0	0	0
INCART DB02
Best	84.97	61.05	31.87	*96.13*	*96.13*	*100*	66.67	33.33	*100*	*100*	75	25	12.5	*100*	*100*	90.91	63.64	63.64	*100*	*100*	45.46	72.73	72.73	0	0
μ	48.98	9.84	2.66	*96.13*	*96.13*	75	11.11	2.78	*100*	*100*	41.76	4.79	1.04	*100*	*100*	70.45	53.03	50.76	*100*	*100*	29.55	46.97	49.24	0	0
SD	23.74	18.69	8.81	0	0	30.81	20.79	9.21	0	0	18.62	8.69	3.45	0	0	12.92	12.77	14.11	0	0	12.92	12.77	14.11	0	0
INCART DB02
Best	85.82	61.45	0	*96.13*	*96.13*	*100*	66.67	0	*100*	*100*	75	33.33	0	*100*	*100*	90.91	81.82	72.73	*100*	*100*	54.55	81.82	81.82	0	0
μ	47.31	9.79	0	*96.13*	*96.13*	72.22	11.11	0	*100*	*100*	39.73	6.2	0	*100*	*100*	68.94	52.27	49.24	*100*	*100*	31.06	47.73	50.76	0	0
SD	25.57	18.7	0	0	0	35.57	20.79	0	0	0	22.05	11.63	0	0	0	14.11	16.65	16.79	0	0	14.11	16.65	16.79	0	0
INCART DB03
Best	74.85	45.54	0	95.44	95.44	100	50	0	100	100	66.67	25	0	100	100	80	40	40	100	100	50	90	90	0	0
μ	62.2	9.4	0	95.44	95.44	93.75	10.42	0	100	100	52.51	5.36	0	100	100	65	31.67	27.5	100	100	35	68.33	72.5	0	0
SD	9.2	14.5	0	0	0	10.83	16	0	0	0	8.61	8.15	0	0	0	9.57	9.86	10.1	0	0	9.57	9.86	10.1	0	0
INCART DB03
Best	76.34	45.31	0	95.44	95.44	100	50	0	100	100	66.67	25	0	100	100	80	50	50	100	100	60	90	90	0	0
μ	65.67	18.06	0	95.44	95.44	93.75	20.83	0	100	100	52.55	10.91	0	100	100	64.17	35	26.67	100	100	35.83	65	73.33	0	0
SD	10.32	17.37	0	0	0	10.83	19.98	0	0	0	10.06	10.17	0	0	0	12.56	12.58	12.47	0	0	12.56	12.58	12.47	0	0
INCART DB03
Best	79.57	63.48	0	95.44	95.44	100	75	0	100	100	66.67	42.86	0	100	100	80	60	40	100	100	50	80	80	0	0
μ	67.62	38.61	0	95.44	95.44	91.67	45.83	0	100	100	51.34	25.35	0	100	100	64.17	45.83	27.5	100	100	35.83	54.17	72.5	0	0
SD	8.16	19	0	0	0	11.79	22.44	0	0	0	9.67	11.71	0	0	0	9.54	11.15	8.29	0	0	9.54	11.15	8.29	0	0
INCART DB04
Best	95	95	0	*95.07*	*95.07*	*100*	*100*	0	*100*	*100*	*100*	33.33	0	*100*	*100*	*100*	92.86	92.86	*100*	*100*	21.43	28.57	28.57	0	0
μ	57.64	15.56	0	*95.07*	*95.07*	91.67	16.67	0	*100*	*100*	73.61	4.86	0	*100*	*100*	94.05	88.69	87.5	*100*	*100*	5.95	11.31	12.5	0	0
SD	23.22	34.8	0	0	0	27.64	37.27	0	0	0	38.01	11	0	0	0	8.67	6.81	8.31	0	0	8.67	6.81	8.31	0	0
INCART DB05
Best	94.78	94.78	0	86.96	0	100	100	0	100	0	50	16.67	0	50	0	90	80	80	90	80	80	80	80	10	20
μ	54.15	16.44	0	86.96	0	81.82	18.18	0	100	0	19.71	3.03	0	50	0	51.82	45.45	43.64	90	80	48.18	54.55	56.36	10	20
SD	28.97	34.93	0	0	0	38.57	38.57	0	0	0	15.73	6.43	0	0	0	21.67	19.71	19.67	0	0	21.67	19.71	19.67	0	0