Skip to main content

Table 4 Performance of several nuclei segmentation methods on the multiple organ H&E stained histopathology image dataset (MOSID)

From: An automatic nuclei segmentation method based on deep convolutional neural networks for histopathology images

Methods

Precision

Recall

F1-Score

ADC

AJI

CP [35]

N/A

N/A

0.405

0.597

0.123

Fiji [36]

N/A

N/A

0.665

0.649

0.273

CNN2 [13]

N/A

N/A

0.754

0.693

0.348

CNN3 [20]

N/A

N/A

0.827

0.762

0.508

NB [22]

0.836

0.894

0.852

0.809

N/A

NucSeg

0.913 ±0.003

0.821 ±0.004

0.861 ±0.001

0.812 ±0.001

0.669 ±0.001

NucSeg-N

0.897 ±0.004

0.813 ±0.004

0.849 ±0.002

0.805 ±0.002

0.649 ±0.004

NucSeg-P

0.909 ±0.002

0.777 ±0.004

0.835 ±0.002

0.809 ±0.001

0.664 ±0.002

NucSeg-NP

0.899 ±0.004

0.777 ±0.005

0.828 ±0.002

0.701 ±0.002

0.647 ±0.005