Table 2 Proportions of accurately classified samples for different predicted cell numbers by trained CNN.

From: Construction of a system using a deep learning algorithm to count cell numbers in nanoliter wells for viable single-cell experiments

Predicted number of cells per well

Overall

0

1

2

>2

Number of wells according to trained CNN

7,920

6,467

1,203

225

25

Number of concordant wells

7,787

6,420

1,150

193

24

Proportion

0.98321

0.99273

0.95594

0.85778

0.96

Back to article page