Figure 4

Lipiodol coverage. Lipiodol coverage was measured on 24 h CT, averaged over all tumors as well as various subgroups. p values are derived from statistical tests (Mann–Whitney U test, Kruskal–Wallis test, Wilcoxon signed-rank test) of subgroup differences in percent tumor coverage. Bars represent 90% confidence intervals. (a) The percent Lipiodol coverage of the whole tumor for various subgroups. (b) Necrotic and viable tumor coverage with Lipiodol. Tumor enhancement on baseline MRI was significantly associated with Lipiodol deposition on 24 h CT (p < 0.0001), with 8.22% ± 14.59 more Lipiodol coverage in viable areas than necrotic areas of the tumor on BL MRI (c) Lipiodol density distribution in viable versus necrotic areas. The fraction of low density Lipiodol in necrotic areas was significantly higher than in viable areas (difference of 8.10% ± 15.69, p = 0.0002), and not significantly different for mid density Lipiodol (p = 0.0933), whereas the fraction of high density Lipiodol in viable areas was significantly higher than in necrotic areas (17.21% ± 22.19, p < 0.0001). For exact p values as well as mean ± standard deviation (SD) for (c) see Supplemental Materials.