Skip to main content
. 2018 Apr 12;13(4):e0195756. doi: 10.1371/journal.pone.0195756

Table 2. Univariate Cox proportional hazard analysis of morphologic and kinetic features on MR imaging with survival outcomes.

Variables No. of patients (recurrence) Hazard ratio 95% CI P value
No (n = 269) Yes (n = 32)
Mass shape
    Oval 40 (14.9%) 5 (15.6%) Reference
    Round 27 (10.0%) 3 (9.4%) 0.926 0.221, 3.875 0.916
    Irregular 202 (75.1%) 24 (75.0%) 1.020 0.389, 2.673 0.968
Mass margin
    Circumscribed 25 (9.3%) 3 (9.4%) Reference
    Not circumscribed 244 (90.7%) 29 (90.6%) 1.072 0.327, 3.520 0.909
Mass internal enhancement
    Homogeneous 32 (11.9%) 2 (6.2%) Reference
    Heterogeneous 204 (75.8%) 22 (68.8%) 1.829 0.430, 7.779 0.414
    Rim enhancement 33 (123%) 8 (25.0%) 4.004 0.850, 18.858 0.079
Peak signal intensity (%)* 252.13 (77–986) 310.19(132–559) 1.004 1.002, 1.006 0.001
Early phase enhancement §
    Medium (%) 7.96 4.00 0.980 0.948, 1.013 0.234
    Rapid (%) 92.08 96.03 1.020 0.987, 1.054 0.236
Delayed phase enhancement §
    Persistent (%) 21.22 19.3 0.997 0.982, 1.013 0.732
    Plateau (%) 40.69 41.56 1.001 0.983, 1.021 0.886
    Washout (%) 38.08 39.19 1.001 0.989, 1.014 0.834
Kinetic curve presented by CAD
    Type 1 (delayed persistent enhancement pattern) 2 (0.7%) 0 (0%) Reference
    Type 2 (delayed plateau enhancement pattern) 13 (4.8%) 2 (6.3%) 4072785.2191 0, Inf 0.997
    Type 3 (delayed washout enhancement pattern) 254 (94.4%) 30 (93.7%) 3274787.6777 0, Inf 0.997

* Numbers represent the mean values with ranges

§ Numbers represent percentages

CI = confidence interval