Unsupervised Cardiac Image Segmentation via Multiswarm Active Contours with a Shape Prior

Table 1

Similarity measure with the Jaccard and Dice indices, Hausdorff distance, and the Maximum cardinality similarity metric among the regions segmented by the Graph Cut method, traditional ACM, interactive Tseng method, our proposed method, and the regions outlined by experts of the CT dataset.

Test	Graph Cut versus experts				ACM versus experts				Tseng versus experts				Our method versus experts
Image	(J)	(D)	(H)	(MCSM)	(J)	(D)	(H)	(MCSM)	(J)	(D)	(H)	(MCSM)	(J)	(D)	(H)	(MCSM)
1	0.551	0.711	4.000	0.343	0.698	0.822	10.049	0.681	0.636	0.777	10.198	0.781	0.836	0.911	1.036	0.759
5	0.698	0.822	5.385	0.497	0.636	0.777	5.099	0.396	0.666	0.800	10.000	0.644	0.800	0.888	3.989	0.833
10	0.607	0.755	3.605	0.374	0.607	0.755	4.472	0.452	0.578	0.733	5.000	0.684	0.764	0.866	1.000	0.813
15	0.475	0.644	10.000	0.512	0.836	0.911	6.580	0.695	0.800	0.888	7.211	0.422	0.875	0.933	2.719	0.793
20	0.428	0.600	2.828	0.449	0.875	0.933	7.214	0.660	0.836	0.911	10.000	0.552	0.914	0.955	3.105	0.764
25	0.800	0.888	4.900	0.475	0.914	0.955	5.000	0.723	0.698	0.822	7.000	0.511	0.764	0.866	1.381	0.729
30	0.730	0.844	2.236	0.386	0.607	0.755	2.828	0.621	0.636	0.777	5.385	0.748	0.730	0.844	2.828	0.790
35	0.636	0.777	5.000	0.550	0.551	0.711	1.000	0.720	0.875	0.933	2.828	0.781	0.914	0.955	1.082	0.920
40	0.607	0.755	4.500	0.493	0.500	0.666	4.123	0.740	0.800	0.888	2.000	0.755	0.875	0.933	6.000	0.797
45	0.525	0.688	10.414	0.500	0.525	0.688	5.000	0.617	0.607	0.755	3.000	0.835	0.636	0.777	1.082	0.863
50	0.451	0.622	9.798	0.469	0.698	0.822	18.384	0.601	0.730	0.844	1.414	0.519	0.698	0.822	2.828	0.749
55	0.428	0.600	6.082	0.434	0.764	0.866	12.529	0.419	0.800	0.888	1.000	0.430	0.836	0.911	5.099	0.645
60	0.764	0.866	9.848	0.439	0.666	0.800	2.236	0.692	0.730	0.844	1.000	0.509	0.764	0.866	1.000	0.778
65	0.875	0.933	11.313	0.394	0.914	0.955	8.000	0.616	0.875	0.933	8.000	0.615	0.800	0.888	8.000	0.635
70	0.451	0.622	16.278	0.467	0.525	0.688	1.000	0.688	0.607	0.755	3.000	0.505	0.578	0.733	5.236	0.863
75	0.500	0.666	19.798	0.484	0.551	0.711	2.236	0.561	0.525	0.688	2.828	0.683	0.607	0.755	2.828	0.655
80	0.551	0.711	14.866	0.398	0.607	0.755	5.000	0.504	0.578	0.733	2.000	0.879	0.666	0.800	5.000	0.843
85	0.578	0.733	12.236	0.394	0.666	0.800	3.162	0.542	0.698	0.822	3.083	0.827	0.730	0.844	4.885	0.737
90	0.698	0.822	6.403	0.468	0.764	0.866	4.123	0.511	0.800	0.888	9.433	0.534	0.764	0.866	6.336	0.832
95	0.764	0.866	13.605	0.502	0.836	0.911	11.401	0.567	0.836	0.911	5.099	0.687	0.800	0.888	8.000	0.904
100	0.525	0.688	14.123	0.467	0.578	0.733	1.000	0.576	0.551	0.711	1.000	0.579	0.636	0.777	2.236	0.878
105	0.406	0.577	1.414	0.523	0.475	0.644	13.601	0.604	0.607	0.755	5.385	0.632	0.607	0.755	1.414	0.838
110	0.384	0.555	6.0	0.602	0.428	0.600	5.000	0.691	0.500	0.666	7.000	0.575	0.551	0.711	8.000	0.776
115	0.836	0.911	8.944	0.586	0.800	0.888	13.038	0.461	0.836	0.911	4.472	0.523	0.875	0.933	4.242	0.869
120	0.764	0.866	9.848	0.514	0.836	0.911	15.231	0.695	0.764	0.866	6.000	0.663	0.836	0.911	6.000	0.817
125	0.666	0.800	13.601	0.611	0.875	0.933	14.142	0.609	0.914	0.955	6.708	0.718	0.956	0.977	3.000	0.948
130	0.578	0.733	10.770	0.598	0.607	0.755	12.649	0.677	0.875	0.933	7.280	0.618	0.914	0.955	4.123	0.843
135	0.698	0.822	11.401	0.487	0.698	0.822	17.720	0.741	0.956	0.977	5.000	0.681	0.956	0.977	2.236	0.705
140	0.636	0.777	8.540	0.534	0.764	0.866	11.704	0.588	0.836	0.911	1.000	0.653	0.875	0.933	2.000	0.834
144	0.525	0.688	6.827	0.568	0.578	0.733	7.280	0.619	0.730	0.844	10.000	0.843	0.914	0.955	1.000	0.935
Average	0.607	0.755	8.485	0.546	0.666	0.800	7.182	0.692	0.764	0.866	5.716	0.778	0.875	0.933	5.228	0.856