Medical Imaging Lesion Detection Based on Unified Gravitational Fuzzy Clustering

Table 3

Performance criteria of the proposed method applied on high-grade/low-grade tumor images from INNN: Dice, Jaccard, sensitivity, and specificity score.

ID	Dice	Jaccard	Sensitivity	Specificity
INNN001	0.942	0.851	0.943	1
INNN002	0.899	0.785	0.926	1
INNN003	0.907	0.839	0.938	1
INNN004	0.860	0.763	0.918	1
INNN005	0.968	0.882	0.978	1
INNN006	0.801	0.701	0.880	1
INNN007	0.795	0.699	0.834	1
INNN008	0.807	0.731	0.890	1
INNN009	0.861	0.725	0.914	1
INNN010	0.842	0.711	0.906	1
INNN011	0.828	0.701	0.886	1
INNN012	0.837	0.707	0.901	1
INNN013	0.871	0.765	0.933	1
INNN014	0.933	0.831	0.940	1
INNN015	0.911	0.876	0.939	1
INNN016	0.907	0.855	0.927	1
INNN017	0.732	0.658	0.811	1
INNN018	0.790	0.684	0.839	1
INNN019	0.801	0.713	0.846	1
INNN020	0.951	0.873	0.967	1
INNN021	0.511	0.440	0.816	0.901
INNN022	0.887	0.778	0.949	1
INNN023	0.819	0.718	0.850	1
INNN024	0.918	0.877	0.942	1
INNN025	0.851	0.728	0.944	1
Mean	0.850	0.756	0.905	0.996
Standard deviation	0.091	0.097	0.048	0.019