Table 3 Class-wise performance metrics of conventional CSHG-CervixNet using k-fold cross-validation.

From: A hybrid compound scaling hypergraph neural network for robust cervical cancer subtype classification using whole slide cytology images

Performance metrics- k-fold

Specificity

Precision

Recall

F1-score

Dyskeratotic

99.54

98.61

100

99.30

Koilocytotic

99.53

98.68

100

99.33

Metaplastic

100

100

98.70

99.34

Parabasal

100

100

100

100

Superficial_intermediate

100

100

96.96

98.46