Table 2.

Segmentation metrics.

Metric	Equation	Range	Meaning
Sørensen-Dice Index	$\frac{2\times \left|\text{SM},\cap ,\text{GT}\right|}{|\text{SM}|+|\text{GT}|}$	0–1	Spatial overlap between masks
Jaccard Index	$\frac{|\text{SM}\cap \text{GT}|}{|\text{SM}|+|\text{GT}|-|\text{SM}\cap \text{GT}|}$	0–1	Spatial overlap between masks
Conformity Coefficient	$1-\frac{\text{FP}+\text{FN}}{\text{TP}}$	< 1	Ratio of incorrectly and correctly segmented voxels
True Positive Rate (TPR)	$\frac{\text{TP}}{\text{TP}+\text{FN}}$	0–1	Sensitivity
True Negative Rate (TNR)	$\frac{\text{TN}}{\text{TN}+\text{FP}}$	0–1	Specificity
Positive Predictive Value (PPV)	$\frac{\text{TP}}{\text{TP}+\text{FP}}$	0–1	Precision
Volume Ratio	$\frac{\text{SM}}{\text{GT}}$	≥ 0	Ratio of mask volumes

SM  segmentation mask, GT  ground truth mask, TP  true positive (i.e., voxels correctly segmented as muscle), TN  true negative (i.e., voxels correctly segmented as background), FP  false positive (i.e., voxels incorrectly segmented as muscle), FN  false negative (i.e., voxels incorrectly segmented as background).