Table 3.

Application of AI in cervical cell segmentation.

Reference	Year	Number of subjects	Methods	Datasets	Results
Wang et al. (44)	2014	362 cervical cell images (3722 cells)	Mean-Shift clustering algorithm	Private	Sensitivity: 94.25% Specificity: 93.45%
Song et al. (50)	2019	8 cervical cell images 22 cervical images	CNN	ISBI2015 Private	DSC: 0.84 DSC: 0.83
Zhao et al. (45)	2016	917 single-cell images	Superpixel-based Markov random field	Herlev The real-word Datasets	Herlev ZSI of nuclei: 0.93 ZSI of cytoplasm: 0.82
Gautam et al. (46)	2018	917 single-cell images	Patch-based CNN	Herlev	DSC: 0.90 Precision: 89%

CNN, convolutional neural network; DSC, dice similarity coefficient; ZSI, zijdenbos similarity index.