Skip to main content

Table 3 Results of univariate cox proportional hazards done on 12 commonly reported texture parameters

From: The effects of segmentation algorithms on the measurement of 18F-FDG PET texture parameters in non-small cell lung cancer

Texture parameter FH FLAB 40P
HR (95% CI) AIC (95% CI) HR (95% CI) AIC (95% CI) HR 40P (95% CI) AIC (95% CI)
MATV 1.1 (0.55–2.02) 220.81 (178.62–261.66) 1.06 (0.51–2.16) 220.8(176.18–259.29) 0.8 (0.39–1.61) 220.46 (178.15–259.25)
TLG 1.0 (0.6–1.7) 220.84 (177.82–261.74) 1.0(0.57–1.73) 220.84 (173.1–257.46) 0.88 (0.53–1.47) 220.55 (176.2–264.39)
SUVmean 1 (1–1) 220.41(176.44–261.64) 1 (1–1) 220.24(178.26–261.34) 1 (1–1) 220.5(172.19–262.68)
SUVmax 1 (1–1) 220.77 (174.54–259.51) 1 (1–1) 220.77 (178.24–258.84) 1 (1–1) 220.69 (175.34–261.62)
SUV standard deviation 1 (1–1) 220.23 (176.06–260.46) 1 (1–1) 220.33 (177.47–262.87) 1 (1–1) 220.26 (177.06–264.65)
first-order entropy 0.16 (0.04–0.74)* 216.12 (174.54–255.93) 0.2 (0.06–0.73)* 215.91 (175.69–259.35) 0.04 (0.005–0.328 212.47 (169.02–256.07)
GLCM entropy 0.56 (0.22–1.42) 219.53 (178.49–261.04) 0.45 (0.2–1.01) 217.8 (173.74–258.62) 0.21 (0.04–0.89)* 216.75 (173.21–256.01)
GLCM homogeneity 14.89 (0.12–1799.35) 219.68 (175.77–258.64) 32.51 (0.26–3999.05) 219.12 (173.93–256.75) 193.94 (0.02–2,314,503.14) 219.62 (176.21–259.9)
GLCM dissimilarity 0.836 (0.54–1.28) 220.18 (176.78–257.79) 0.75 (0.49–1.13) 219.09 (173.93–257.71) 0.83 (0.5–1.37) 220.32 (177.06–257.92)
GLSZM intensity variability 0.46 (0.18–1.18) 218.24 (169.74–256.67) 0.65 (0.32–1.34) 219.55 (174.97–260.52) 0.72 (0.35–1.47) 220.07 (173.75–258.65)
NGTDM coarseness 0.88 (0.74–1.05) 218.64 (170.64–257.63) 0.9 (0.74–1.08) 219.6 (175.71–260.62) 0.82 (0.67–1.02) 217.62 (168.78–252.86)
NGTDM contrast 0.24 (0.04–1.35) 218.45 (178.17–260.37) 0.18 (0.03–0.85) 216.85 (173.33–256) 0.46 (0.04–4.6) 220.42 (173.91–257.85)
  1. AIC Akaike information criterion, CI confidence interval, HR hazard ratio
  2. *p value <0.05
  3. Variable log-transformed