Assessment of MRI-Based Radiomics in Preoperative T Staging of Rectal Cancer: Comparison between Minimum and Maximum Delineation Methods

Table 2

Selected radiomics features.

Model	No	Radiomics feature	Radiomics class	Filter
Method 1	1	Skewness	First order	Wavelet-HLL
	2	Maximum	First order	Wavelet-HLL
	3	High gray level zone emphasis	GLSZM	Wavelet-HLH
	4	Gray level nonuniformity	GLSZM	Wavelet-LHL
Method 2	1	Skewness	First order	Wavelet-HLL
	2	High gray level zone emphasis	GLSZM	Wavelet-LHL
	3	Skewness	First order	Wavelet-LHL
	4	High gray level run emphasis	GLRLM	Original
	5	High gray level run emphasis	GLRLM	Logarithm
	6	High gray level run emphasis	GLRLM	Square root
	7	High gray level run emphasis	GLRLM	Wavelet-LLL

GLSZM: gray level size zone matrix; GLRLM: gray level run length matrix. The wavelet transform decomposes the tumor area image into low-frequency components (L) or high-frequency components (H) in the , , and axes. Method 1: minimum delineation method; Method 2: maximum delineation method.