A Markov Random Field and Adaptive Regularization Embedded Level Set Segmentation Model Solving by Graph Cuts

Table 1

Region overlap metrics of different algorithms corresponding to Figure 4 numbered in order from up to down.

Image no.	Algorithms	JS	DSC	FPR	FNR
Image #1	GAC	1.0000	1.0000	0.0000	0.0000
	LSR	0.9573	0.9782	0.0168	0.0268
	CV	1.0000	1.0000	0.0000	0.0000
	FTC	1.0000	1.0000	0.0000	0.0000
	LBF	0.9946	0.9973	0.0005	0.0048
	LAC	0.9800	0.9899	0.0200	0.0060
	LPFE	0.9517	0.9753	0.0184	0.0310
	NFT	0.9763	0.9880	0.0200	0.0039
	Our method	1.0000	1.0000	0.0000	0.0000
Image #2	GAC	0.8476	0.9175	0.0671	0.0974
	LSR	0.7789	0.8757	0.1499	0.0971
	CV	0.8807	0.9366	0.1169	0.0031
	FTC	0.8544	0.9215	0.0127	0.1436
	LBF	0.9741	0.9770	0.0116	0.0043
	LAC	0.9200	0.9583	0.0725	0.0087
	LPFE	0.9535	0.9762	0.0244	0.0233
	NFT	0.7703	0.8703	0.0308	0.2103
	Our method	0.9833	0.9916	0.0114	0.0030
Image #3	GAC	0.7715	0.8710	0.1851	0.0646
	LSR	0.4215	0.5931	0.1396	0.5475
	CV	0.4475	0.6183	0.2917	0.4514
	FTC	0.5055	0.6716	0.0119	0.4914
	LBF	0.6990	0.8228	0.0087	0.2967
	LAC	0.7712	0.8709	0.2045	0.0380
	LPFE	0.9023	0.9487	0.0152	0.0850
	NFT	0.4719	0.6412	0.0189	0.5238
	Our method	0.9711	0.9853	0.0113	0.0120
Image #4	GAC	0.6547	0.7913	0.3362	0.0205
	LSR	0.3446	0.5125	0.0818	0.6445
	CV	0.2670	0.4214	0.3271	0.6932
	FTC	0.3232	0.4886	0.0463	0.6754
	LBF	0.2622	0.4154	0.0464	0.7369
	LAC	0.7049	0.8269	0.2582	0.0659
	LPFE	0.6453	0.7844	0.0471	0.3474
	NFT	0.3689	0.5390	0.0484	0.6279
	Our method	0.9429	0.9706	0.0460	0.0122