Table 2 Number of cell images used in the study.

From: Deep learning application of the discrimination of bone marrow aspiration cells in patients with myelodysplastic syndromes

 

Classification by algorithm

Total

Training set

Validation set

Test set

Total no. of cell images

6453

806

806

8065

Cell classification

Normal erythrocytes (EN)

1075 (16.7%)

138 (17.1%)

119 (14.8%)

1332 (16.5%)

Normal granulocytes (GN)

2323 (36.0%)

283 (35.1%)

288 (35.7%)

2894 (35.9%)

Normal megakaryocytes (MN)

354 (5.5%)

41 (5.1%)

50 (6.2%)

445 (5.5%)

Dysplastic erythrocytes (ED)

788 (12.2%)

125 (15.5%)

100 (12.4%)

1013 (12.6%)

Dysplastic granulocytes (GD)

930 (14.4%)

103 (12.8%)

130 (16.1%)

1163 (14.4%)

Dysplastic megakaryocytes (MD)

153 (2.4%)

18 (2.2%)

19 (2.4%)

190 (2.4%)

Blasts

529 (8.2%)

54 (6.7%)

59 (7.3%)

642 (8.0%)

Others

301 (4.7%)

44 (5.5%)

41 (5.1%)

386 (4.8%)