Skip to main content

Table 1 Composition of training data set and test data set of two data sets

From: Machine learning prediction of HER2-low expression in breast cancers based on hematoxylin–eosin-stained slides

Dataset type

Her2 scores

Training data set

Test data set

USTC-BC

HER2 0

41 slides

18 slides

HER2 1 + 

56 slides

24 slides

HER2 2 +

74 slides

31 slides

HER2 3 + 

74 slides

32 slides

Total

245 slides

105 slides

 

350 slides

 

TCGA-BRCA

HER2 0

40 slides

17 slides

HER2 1 + 

148 slides

63 slides

HER2 2 + 

102 slides

44 slides

HER2 3 + 

62 slides

26 slides

Total

352 slides

150 slides

 

502 slides