Table 2. Segmentation Metrics.

Segmentation Metric	Formula
Mean/Dice Similarity Coefficient (MO)	$\text{MO }=2\frac{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{S}}_{\text{r}}{{{\displaystyle \cap }}^{\text{}}}^{\text{}}{\text{T}}_{\text{r}}\right|}{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left(\left|{\text{S}}_{\text{r}}\right|+\left|{\text{T}}_{\text{r}}\right|\right)}$
Union Overlap/Jaccard Coefficient (UO)	$\text{UO }=\frac{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{S}}_{\text{r}}{{{\displaystyle \cap }}^{\text{}}}^{\text{}}{\text{T}}_{\text{r}}\right|}{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{S}}_{\text{r}}{{{\displaystyle \cup }}^{\text{}}}^{\text{}}{\text{T}}_{\text{r}}\right|}$
False Negative (FN)	$\text{FN }=\frac{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{T}}_{\text{r}}\backslash {\text{S}}_{\text{r}}\right|}{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{T}}_{\text{r}}\right|}$
False Positive (FP)	$\text{FP }=\frac{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{S}}_{\text{r}}\backslash {\text{T}}_{\text{r}}\right|}{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left|{\text{S}}_{\text{r}}\right|}$
Volume Similarity/Relative Volume Difference (VS)	$\text{VS }=2\frac{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left(\left|{\text{S}}_{\text{r}}\right|-\left|{\text{T}}_{\text{r}}\right|\right)}{{{{\displaystyle \sum }}^{\text{}}}_{\text{r}}\left(\left|{\text{S}}_{\text{r}}\right|+\left|{\text{T}}_{\text{r}}\right|\right)}$
Distance Error (DEr)	${\text{DE}}_{\text{r}}=\frac{1}{\text{P}}{\displaystyle {\sum }_{\text{p}=1}^{\text{p}}\text{mindist }\left({\text{S}}_{\text{r}}{\text{B}}_{\text{p}},{\text{T}}_{\text{r}}\text{B}\right)}$ where mindist = minimum distance from source region boundary point SrBp to entire target region boundary TrB.
Spatial Extent (SE)	$\text{SE }=\text{ vol}\left(\text{P}>\text{t}\right)0\le \text{SE}\le \text{V}$ where $\text{P}\left(\text{x},\text{y},\text{z}\right)=\frac{1}{\text{N}}{\displaystyle {\sum }_{\text{i}=1}^{\text{N}}{\text{L}}_{\text{i}}\left(\text{x},\text{y},\text{z}\right)}$

Sr = Source Region (Image 1), Tr = Target Region (Image 2). There are several metrics for measuring overlap of two 3-dimensional images. Most studies focus on the Dice Similarity Coefficient or Mean Overlap. Due to its ubiquitous use this study will also use it as the main metric of comparison.