Table 2.

Results on the training and validation sets for models evaluated.

Model	Number of Parameters (Trainable)	Number of Epochs Trained	Training Accuracy	Training Loss	Validation Accuracy	Validation Loss
Baseline	4,772,220	50	95.25%	0.1247	81.93%	0.5657
Baseline + Augmentation	4,772,220	50	93.82%	0.1640	82.13%	0.6210
VGG16	35,663,873	50	99.75%	0.0081	79.00%	2.9023
Baseline + Nucleus U-net	4,772,261	50	95.85%	0.1923	83.36%	0.5808
Baseline + Mitosis U-net	4,772,261	50	95.28%	0.1236	83.97%	0.4597
Baseline + Epithelium U-net	4,772,261	50	95.23%	0.1261	85.07%	0.4045
Baseline + Tubule U-net	4,772,261	50	96.02%	0.1048	82.93%	0.6176
ConcatNet (Baseline + all U-nets)	4,772,384	50	95.90%	0.1082	86.23%	0.4357