Skip to main content

Table 2 The performance of the proposed method for nucleus segmentation compared to U-Net, CNN, support-vector machine (SVM) based on the data set two presented using average in each measure of similarity

From: Threshold estimation based on local minima for nucleus and cytoplasm segmentation

 

JI

DSC

Specificity

Sensitivity

Precision

Proposed method

0.902925

0.948079

0.993110

0.952951

0.948903

U-Net [36]

0.917556

0.956456

0.995364

0.950505

0.964065

CNN [21]

0.910000

0.979200

0.960800

0.876300

SVM [21]

0.860000

0.818600

0.989700

0.934300