ColoRectalCADx: Expeditious Recognition of Colorectal Cancer with Integrated Convolutional Neural Networks and Visual Explanations Using Mixed Dataset Evidence

Table 15

Comparison of the results of previous state-of-the-art methods of Kvasir-SEG dataset.

Method	Dice coefficient
ResU-Net++ [65]	0.8133
ResU-Net-mod	0.7909
ResU-Net	0.5144
U-Net	0.7147
U-Net (ours)	0.8129