Table 1 Distribution of cases and WSIs to training, validation and test set, as well as respective distributions of stains, IHC antibodies and scanner models.

From: A Multi-Stain Breast Cancer Histological Whole-Slide-Image Data Set from Routine Diagnostics

 

Training

Validation

Test

Total

Cases

750 (65.0%)

100 (8.7%)

303 (26.3%)

1153

WSIs

3406 (80.9%)

200 (4.7%)

606 (14.8%)

4212

Stain and antibodies

H&E

750 (22.0%)

100 (50.0%)

303 (50.0%)

1153 (27.4%)

ER

732 (21.5%)

29 (14.5%)

84 (13.9%)

845 (20.1%)

KI67

732 (21.5%)

29 (14.5%)

82 (13.5%)

843 (20.0%)

PGR

728 (21.4%)

28 (14.0%)

81 (13.4%)

837 (19.9%)

HER2

464 (13.6%)

14 (7.0%)

56 (9.2%)

534 (12.7%)

Scanners

NanoZoomer S360

559 (16.4%)

38 (19.0%)

205 (33.8%)

802 (19.0%)

NanoZoomer XR (1)

884 (26.0%)

46 (23.0%)

203 (33.5%)

1133 (26.9%)

NanoZoomer XR (2)

1963 (57.6%)

116 (58.0%)

198 (32.7%)

2277 (54.1%)

  1. Three Hamamatsu NanoZoomer scanners were used for WSI digitisation, one NanoZoomer S360 and two NanoZoomer XR.