Table 4 Experimental results for the child datasets, trained exclusively on the children’s dental dataset (163 training images, 30 test images).

From: Children’s dental panoramic radiographs dataset for caries segmentation and dental disease detection

 

Recall

Specificity

ACC

IOU

Dice

mean

std

mean

std

mean

std

mean

std

mean

std

U-Net22

0.9200

0.0375

0.9803

0.0057

0.9710

0.0052

0.8387

0.0300

0.9120

0.0182

R2 U-Net23

0.8854

0.0679

0.9840

0.0055

0.9675

0.0081

0.8247

0.0563

0.9027

0.0375

PSPNet24

0.8875

0.0198

0.9856

0.0049

0.9683

0.0065

0.8324

0.0221

0.9083

0.0132

Deeplab V3+25

0.9486

0.0162

0.9701

0.0069

0.9670

0.0058

0.8121

0.0259

0.8961

0.0158

  1. Evaluation metrics include Recall, Specificity, Accuracy, Intersection over Union (IOU), and Dice index, with mean and standard deviation (std) values provided.
Back to article page