Texture analysis of T2-weighted cardiovascular magnetic resonance imaging to discriminate between cardiac amyloidosis and hypertrophic cardiomyopathy

Table 2 Differences of selected texture features between CA and HCM

Image type	Feature class	Features names	CA	HCM	P value
Wavelet-LHL	GLRLM	Long run emphasis	6.639 (5.796, 7.310)	7.360 (6.721, 8.266)	< 0.0001
Wavelet-LHL	GLRLM	Short run emphasis	0.535 (0.510, 0.567)	0.504 (0.478, 0.530)	< 0.0001
Wavelet-LLH	GLRLM	Long run emphasis	143.8 (104.5, 192.4)	193.9 (148.7, 252.2)	< 0.0001
Wavelet-HLH	GLDM	Small dependence high gray level emphasis	0.143 ± 0.019	0.128 ± 0.016	< 0.0001
Original	First order	Skewness	0.323 ± 0.436	0.107 ± 0.447	< 0.0001
Original	GLDM	Large dependence emphasis	48.3 (38.8, 56.2)	55.6 (50.2, 60.0)	< 0.0001
Original	GLDM	Correlation	0.911 (0.858, 0.951)	0.819 (0.712, 0.908)	< 0.0001

GLRLM: gray level run length matrix, GLDM: gray level dependence matrix, GLCM: gray level co-occurrence matrix; H: high wavelet filter; L: low wavelet filter

ISSN: 1471-2261