Table 5.

Validation AccID (mean ± CI) of the optimal ID models for different feature extraction methods using 12-lead ECG (Table 3). AccID is evaluated for different number of reference subjects (10 to 230 subjects included in the validation set). The colour map is provided for visual identification of the ID models’ ability to keep the mean AccID within different tolerance ranges: ≥90% in dark green, [80–90)% in light green, [70–80)% in yellow, [50–70)% in orange, <50% in grey.

Nb Subjects	COR-PQRST	COR-QRS	AMP-QRS	MATCH-QRS	All Features
10	93.7 ± 2.1	94.1 ± 2.4	56.2 ± 4.7	98.0 ± 0.8	98.4 ± 0.8
20	90.4 ± 2.9	91.9 ± 2.9	44.5 ± 5.0	96.4 ± 1.4	97.2 ± 1.3
30	88.1 ± 3.3	90.3 ± 3.2	38.1 ± 4.9	95.1 ± 1.9	96.1 ± 1.7
40	86.3 ± 3.6	89.1 ± 3.4	33.7 ± 4.8	94.0 ± 2.2	95.3 ± 2.0
50	84.9 ± 3.9	88.0 ± 3.6	30.5 ± 4.8	93.0 ± 2.5	94.5 ± 2.2
60	83.7 ± 4.1	87.2 ± 3.8	28.0 ± 4.7	92.1 ± 2.7	93.8 ± 2.4
70	82.7 ± 4.2	86.5 ± 3.9	25.9 ± 4.7	91.3 ± 2.8	93.2 ± 2.6
80	81.8 ± 4.4	85.9 ± 4.1	24.2 ± 4.6	90.6 ± 3.0	92.7 ± 2.7
90	81.0 ± 4.5	85.3 ± 4.2	22.7 ± 4.6	90.0 ± 3.2	92.2 ± 2.9
100	80.2 ± 4.6	84.8 ± 4.3	21.5 ± 4.5	89.4 ± 3.3	91.7 ± 3.0
110	79.6 ± 4.7	84.4 ± 4.3	20.4 ± 4.5	88.8 ± 3.4	91.3 ± 3.1
120	78.9 ± 4.8	84.0 ± 4.4	19.5 ± 4.5	88.2 ± 3.6	90.9 ± 3.3
130	78.4 ± 4.9	83.6 ± 4.5	18.7 ± 4.5	87.7 ± 3.7	90.5 ± 3.4
140	77.8 ± 5.0	83.3 ± 4.6	18.0 ± 4.5	87.1 ± 3.8	90.1 ± 3.5
150	77.3 ± 5.1	82.9 ± 4.6	17.3 ± 4.5	86.6 ± 3.9	89.8 ± 3.6
160	76.8 ± 5.1	82.6 ± 4.7	16.7 ± 4.5	86.2 ± 4.0	89.4 ± 3.7
170	76.3 ± 5.2	82.4 ± 4.7	16.2 ± 4.4	85.7 ± 4.2	89.1 ± 3.8
180	75.9 ± 5.3	82.1 ± 4.8	15.7 ± 4.4	85.2 ± 4.3	88.8 ± 3.9
190	75.5 ± 5.4	81.8 ± 4.9	15.3 ± 4.4	84.8 ± 4.4	88.5 ± 3.9
200	75.0 ± 5.5	81.6 ± 4.9	14.9 ± 4.5	84.3 ± 4.5	88.2 ± 4.0
210	74.7 ± 5.5	81.3 ± 5.0	14.6 ± 4.5	83.9 ± 4.6	87.9 ± 4.1
220	74.3 ± 5.6	81.1 ± 5.0	14.2 ± 4.5	83.5 ± 4.7	87.7 ± 4.2
230	73.9 ± 5.7	80.9 ± 5.1	13.9 ± 4.5	83.0 ± 4.9	87.4 ± 4.3