Classification on Digital Pathological Images of Breast Cancer Based on Deep Features of Different Levels

Table 18

Literature comparison results on patient-level accuracy (/%).

Magnifications	40×	100×	200×	400×
Literature [12]	81.60 ± 3.00	79.90 ± 5.40	85.10 ± 3.10	82.30 ± 3.80
Literature [22]	90.60 ± 6.70	88.40 ± 4.80	84.60 ± 4.20	86.10 ± 6.20
Literature [23]	84.60 ± 2.90	84.80 ± 4.20	84.20 ± 1.70	81.60 ± 3.70
Literature [24]	92.10 ± 5.90	89.10 ± 5.20	87.20 ± 4.30	82.70 ± 3.00
Literature [27]	83.08 ± 2.10	83.17 ± 3.50	84.63 ± 2.70	82.10 ± 4.40
Our method	87.93 ± 3.91	87.41 ± 3.26	88.76 ± 2.50	85.55 ± 4.03