*2.5. Statistical Analysis*

Tumors that underwent histopathology examination were used as the "gold standard" and compared to the tumor sizes derived from CESM images. The normality of the distribution was assessed using the Shapiro–Wilk test and the continuous variables were summarized using the arithmetic mean with standard deviation for data following a normal distribution or a median with quartiles 1 and 3 for data demonstrating anonnormal distribution. Comparison of four respective aspects referring to the maximal tumor dimension (defined as the maximum of three dimensions measured in the CESM and histopathology) included the Pearson's correlation coefficient (R-value) in order to measure the strength of the relationship between low-energy and subtraction images of CESM measurements. Paired *t*-tests were used to assess mean differences between each analyzed study participant. The correlations of data are illustrated by plotting the actual measurements while all paired measurements for each patient are summarized using paired

linear plots. A *p*-value of <0.05 was considered statistically significant. The diagnostic performance indexes for low-energy and subtraction images for complete response and non-complete response were tested using the Clopper–Pearson test with 95% confidence intervals. Statistical analyses were carried out using Statistica 10 (StatSoft Inc., Tulsa, OK, USA) and MedCalc Statistical Software 16.4.3 (MedCalc Software by, Ostend, Belgium) software.
