. 2022 Mar 28;24(4):465. doi: 10.3390/e24040465

Table 3.

Evaluation criteria and performance measures for hepatic vessel skeletonization (analysis). True positives ( $T P$ ) are pixels classified correctly as positive, false positives ( $F P$ ) are pixels classified incorrectly as positive, true negatives ( $T N$ ) are pixels classified correctly as not positive and false negatives ( $F N$ ) are pixels classified incorrectly as not positive.

Metrics	Formula	Description
Dice [64]	$D i c e = \frac{2 * T P}{2 * T P + F P + F N}$	Similarity between two sample sets.
Accuracy [65]	$A c c u r a c y = \frac{T P + T N}{T P + T N + F P + F N}$	Proportion of detected true samples that are actually true.
Sensitivity; recall; true positive rate (TPR) [66]	$S e n s i t i v i t y = \frac{T P}{T P + T N}$	Proportion of positives that are correctly identified.
Specificity [66]	$S p e c i f i c i t y = \frac{T N}{F P + T N}$	Proportion of negatives that are correctly identified.
False positive rate ( $F P R$ ) [67]	$F P R = \frac{F P}{F P + T N}$	Ratio of the number of negative samples wrongly categorized as positive ( $F P$ ) to the total number of actual negative samples.
False negative rate ( $F N R$ ) [67]	$F N R = \frac{F N}{F N + T P}$	Ratio of the number of positive samples wrongly categorized as negative ( $F N$ ) to the total number of actual positive samples.
Root mean standard error ( $R M S E$ ) [68]	$R M S E = \sqrt{\frac{1}{\| R \|} (\sum_{i = 1}^{\| R \|} \| d_{R} \|)}$	Measure of the average squared difference between the result R and the actual value T (ground truth), where $d_{R}$ denotes the distances from points R to points T.
Hausdorff distance (HD) [60]	$d_{H} (A, B) = max {{sup}_{a \in A} {inf}_{b \in B} d (a, b),$ ${sup}_{b \in B} {inf}_{a \in A} d (a, b)}$	Overlapping index, which measures the largest Euclidean distance between two contours A and B and vice versa, computed over all pixels of each curve.