Table 4. Evaluation results for the proposed E-CNN, its individuals (modified TL models) when number of epochs = 10, and the standard TL models on colon histopathlogical images dataset based on the average accuracy, sensitivity, specificity, and average standard deviation (STD) in 10 runs, best results in bold.

Pretrained models	Accuracy	Sensitivity	Specificity
Standard DenseNet121	90.41 ± 3.1	91.25 ± 2.9	100 ± 0
Standard MobileNetV2	90.27 ± 2.9	88.25 ± 1.9	99.23 ± 2.3
Standard InceptionV3	87.12 ± 2.0	92.75 ± 2.0	100 ± 0
Standard VGG16	62.19 ± 7.0	63.21 ± 7.3	100 ± 9.9
Modified DenseNet121	92.32 ± 2.8	92.99 ± 2.8	100 ± 0.0
Modified MobileNetV2	92.19 ± 3.8	90.75 ± 2.0	100 ± 0.0
Modified InceptionV3	89.86 ± 2.2	95.0 ± 1.5	100 ± 0.0
Modified VGG16	72.73 ± 3.9	73.0 ± 3.6	87.43 ± 12.4
Proposed E-CNN (product)	95.20 ± 1.64	95.62 ± 1.50	100 ± 0.0
Proposed E-CNN (Majority voting)	94.52 ± 1.73	95.0 ± 1.58	100 ± 0.0