Skip to main content

Advertisement

Springer Nature is making SARS-CoV-2 and COVID-19 research free. View research | View latest news | Sign up for updates

Table 3 Goodness of fit of eight logistic regression models fitted to the test set

From: High-throughput mammographic-density measurement: a tool for risk prediction of breast cancer

Model Null/Residual deviance P1 P2
Null 2,095.9 - -
Cumulus PD 2,057.6 6.3 × 10-10 -
ImageJ PD 2,062.8 8.8 × 10-9 -
Score 1 2,052.5 4.5 × 10-11 -
Cumulus PD + score 2 2,053.5 6.4 × 10-10 0.0424
Cumulus PD + score 3 2,043.8 4.9 × 10-10 0.0002
ImageJ PD + score 2 2,058.7 8.4 × 10-9 0.0427
ImageJ PD + score 3 2,049.3 7.7 × 10-11 0.0002
  1. Scores for individual images in the test set were derived by summing the products of the nonzero regression coefficients (estimated by using the training set) by the corresponding PC values of that individual image: (1) 123 principal components (PCs) as covariates; all regression coefficients included in the penalty term; (2) percentage density (PD) + 123 PCs as covariates; coefficients for the 123 PCs included in the penalty term, but not the coefficient for PD; and (3) PD + 123 PCs as covariates; all coefficients (123 PCs + PD) included in the penalty. P1, based on Likelihood Ratio Test, comparison with null model. P2, based on Likelihood Ratio Test, comparison with model including Cumulus or ImageJ PD.