Table 4.

IOU score for ETT segmentation based on combined loss functions.

	Loss function	IOU score
		Fold1	Fold2	Fold3	Fold4	Fold5	Mean
1	BCE-Dice	0.8963	0.8839	0.7691	0.8603	0.8852	0.8590
2	BCE-Jaccard	0.8965	0.8498	0.7915	0.8671	0.8728	0.8555
3	BCE-Tversky	0.8697	0.8712	0.7781	0.8707	0.8728	0.8525
4	Focal-Dice	0.8963	0.8617	0.8283	0.8709	0.8689	0.8652
5	Focal-Jaccard	0.8928	0.8623	0.8175	0.8691	0.8800	0.8643
6	Focal-Tversky	0.8934	0.8802	0.8128	0.8706	0.8733	0.8661
7	MCC-Dice	0.9005	0.8550	0.7801	0.8687	0.8667	0.8542
8	MCC-Tversky	0.8984	0.8775	0.8247	0.8711	0.8699	0.8683
9	MCC-Jaccard	0.8812	0.8753	0.7932	0.8725	0.8594	0.8563
						Average =	0.8601

The bold values in each column indicates the highest IOU score.