Breast cancer Ki-67 expression prediction by digital breast tomosynthesis radiomics features

European Radiology Experimental

Table 2 Values of feature parameters differetiating between low and high Ki-67 expression on digital breast tomosynthesis

Feature	Feature class	AUC	p value	Lower 95% CI	Upper 95% CI
Sphericity	Shape	0.613	0.007	0.462	0.764
Autocorrelation	Grey level co-occurrence matrix	0.625	0.002	0.474	0.776
Interquartile range	First order	0.633	0.003	0.481	0.784
Robust mean absolute deviation	First order	0.641	0.009	0.487	0.795
Short-run high grey-level emphasis	Grey level run length matrix	0.676	0.001	0.534	0.818

Features were included if the area under the curve (AUC) was > 0.6 with a p < 0.05. CI Confidence interval