Skip to main content

Advertisement

Table 3 Univariate modelling results from the training dataset

From: A novel and fully automated mammographic texture analysis for risk prediction: results from two case-control studies

Parameter Standardized odds ratio 95% CI for odds ratio χ2 P value mC 95% CI for mC
Coarseness 1.22 (1.06–1.41) 7.28 6.98E-03 0.58 (0.53– 0.62)
Contrast 0.73 (0.62–0.85) 16.75 4.27E-05 0.40 (0.36– 0.45)
SD 1.32 (1.13–1.54) 13.11 2.94E-04 0.57 (0.52– 0.61)
Sum average 16 1.52 (1.31–1.77) 31.26 2.25E-08 0.61 (0.56– 0.65)
Sum average 32 1.52 (1.31–1.77) 31.75 1.75E-08 0.61 (0.56– 0.66)
Sum average 64 1.48 (1.28–1.71) 29.07 6.98E-08 0.58 (0.53– 0.63)
Volumetric PD 1.36 (1.18–1.57) 18.05 2.16E-05 0.59 (0.55– 0.64)
  1. Total number of observations (N) = 1051, including 264 cases and 787 controls. Standardized odds ratio is the change in odds for a standard deviation (in controls) increase in predictors, adjusted for age and body mass index. Sum average 16, 32 and 64 are the texture feature sum average using images downsized by a factor of 16, 32 and 64
  2. PD percent density, CI confidence interval, mC matched concordance index, SD standard deviation