Fig 1.
Patient flow diagram.
Fig 2.
(a) Manually outlining and filtering out the pixels with attenuation under -50 HU in locally advanced rectal cancer in 76-year-old man. (b) Corresponding images in the same patient applying LoG filters with fine, medium, and coarse filter values.
Table 1.
Patient characteristics (n = 95).
Table 2.
Results of texture feature analysis.
Table 3.
Texture features of non-responder versus responder group after CRT without filtration and for various filter scale values depicting fine, medium, and coarse textures.
Fig 3.
Kaplan–Meier curves according to texture features.
Kaplan-Meier curves without filtration showed a significant difference in DFS for (a) entropy, (b) uniformity, and (c) standard deviation.
Table 4.
Spearman rank correlation for texture features without filtration.
Table 5.
Multivariable Cox proportional hazards regression analysis of texture features with CT stage and age as dependent covariate.