Table 5.

Morphological features used in the classification experiment.

No.	Feature	Description	NEN1	Lumen	EN2
1	Area	Area of selection in square pixels	*	*	*
2	Perimeter	The length of the outside boundary of the selection	*	*	*
3	Width	Width of the smallest rectangle enclosing the selection	*	*	*
4	Height	Height of the smallest rectangle enclosing the selection	*	*	*
5	MajorAxis	Major (primary) axis length of the best fitting ellipse	*	*	*
6	MinorAxis	Minor (secondary) axis length of the best fitting ellipse	*	*	*
7	Circularity	4π × (Area/Perimeter2). A value of 1.0 indicates a perfect circle	*	*	*
8	Feret's diameter	The longest distance between any two points along the selection boundary	*	*	*
9	AspectRatio	MajorAxis/MinorAxis	*	*	*
10	Skewness	The third order moment about the mean	*	*	*
11	Roundness	4 × Area/(π × MajorAxis2). The measure of the sharpness of particle's edge and corners	*	*	*
12	Solidity	Area/ConvexArea	*	*	*
13	RMSAA	Root-mean-squared atypia-amplitude		*
14	TSAV	Total sum of atypia volatilities for PIPs		*
15	AtypiaRatio	The ratio of atypia region		*
16	#AtypiaRegions	The number of atypia regions for identifying papillary		*
17	CytoplasmLength	The cytoplasm length of epithelial nucleus [32]			*
18	CytoplasmLengthSD	The standard deviation of CytoplasmLength			*

NEN¹: nonepithelial nuclei; EN²: epithelial nuclei.