Skip to main content

Advertisement

Table 1 Overview of the metrics implemented in this tool

From: Metrics for evaluating 3D medical image segmentation: analysis, selection, and tool

Metric Symb. Reference of use in medical images cat. Definition
Dice (=F1-Measure) DICE [1, 2, 15, 16, 5763] 1 (6)
Jaccard index JAC [15, 16, 2123, 59, 60, 62] 1 (7)
True positive rate (Sensitivity, Recall) TPR [10, 16, 60, 6264] 1 (10)
True negative rate (Specificity) TNR [10, 16, 60, 62] 1 (11)
False positive rate (=1-Specificity, Fallout) FPR → Specificity 1 (12)
False negative rate (=1-Sensitivity) FNR → Sensitivity 1 (13)
F-Measure (F1-Measure=Dice) FMS → Dice 1 (15), (16)
Global Consistency Error GCE [2123, 65, 66] 1 (17) to (19)
Volumetric Similarity VS [15, 2123, 59, 61, 67] 2 (21)
Rand Index RI [21, 22, 65, 66] 3 (30)
Adjusted Rand Index ARI [68, 69] 3 (32)
Mutual Information MI [2, 32, 57] 4 (33) to (38)
Variation of Information VOI [21, 22, 65, 66] 4 (39), (35)
Interclass correlation ICC [8, 70] 5 (41)
Probabilistic Distance PBD [8, 59] 5 (43)
Cohens kappa KAP [1, 62] 5 (44) to (46)
Area under ROC curve AUC [2, 64, 69] 5 (47)
Hausdorff distance HD [8, 15, 59, 6163, 71, 72] 6 (48), (49)
Average distance AVD [62, 63] 6 (50), (51)
Mahalanobis Distance MHD [15, 73] 6 (52) to (54)
  1. The symbols in the second column are used to denote the metrics throughout the paper. The column “reference of use” shows papers where the corresponding metric has been used in the evaluation of medical volume segmentation. The column “category” assigns each metric to one of the following categories: (1) Overlap based, (2) Volume based, (3) Pair counting based, (4) Information theoretic based, (5) Probabilistic based, and (6) Spatial distance based. The column “definition” shows the equation numbers where the metric is defined