Research Article
Comparison of Different Machine Learning Models in Prediction of Postirradiation Recurrence in Prostate Carcinoma Patients
Table 2
Comparison of characteristics among patients with and without disease progression in the training group.
| | Characteristic | The presence/absence of disease progression - training group | | Presence | Absence | Wilcoxon rank sum test |
| | Age (years) | | | | | Mean (SD) | 70.8 (6.3) | 71.2 (5.3) | | | Median (range) | 71.0 (55.0-84.0) | 72.0 (60.0-84.0) | | Baseline PSA | | | | | Mean (SD) | 53.2 (53.2) | 36.8 (40.1) | | | Median (range) | 35.8 (4.5-249.0) | 27.0 (5.6-190.0) | | T in clinical TNM | | | | | T1 and T2 | 8 (14.8%) | 0 (0%) | | | T3 and T4 | 46 (85.2%) | 25 (100%) | | N in clinical TNM | | | | | N0 | 50 (92.6%) | 25 (100%) | | | N1 | 4 (7.4%) | 0 (0%) | | UICC1 staging | | | | | I and II | 45 (83.3%) | 25 (100%) | | | III and IV | 9 (16.7%) | 0 (0%) | | Gleason’s score (category) | | | | | Low risk (GS2 6) | 4 (7.4%) | 4 (16.0%) | | | Intermediate risk (GS2 7) | 24 (44.4%) | 18 (72.0%) | | High risk (GS2 8-10) | 26 (48.2%) | 3 (12.0%) | | Nadir PSA3 | | | | | Mean (SD) | 2.7 (6.6) | 0.6 (1.0) | | | Median (range) | 0.7 (0-43.0) | 0.1 (0-4.2) | | Time to nadir PSA3 | | | | | Mean (SD) | 5.7 (3.4) | 10.6 (7.4) | | | Median (range) | 5.0 (1.0-15.0) | 8.0 (3.0-31.0) | | PSA3 doubling time | | | | | Mean (SD) | 3.6 (2.4) | 14.0 (11.8) | | | Median (range) | 3.1 (0.8-12.9) | 11.5 (2.1-50.0) | | PSA3 velocity | | | | | Mean (SD) | 3.1 (4.9) | 0.2 (0.2) | | | Median (range) | 1.4 (0.1-26.1) | 0.1 (0-0.8) | | | PSA3 in the moment of disease reevaluation | | | | | Mean (SD) | 32.1 (37.2) | 3.6 (3.1) | | | Median (range) | 19.1 (3.9-200.4) | 3.1 (0.3-15.5) | | | Total | 54 (68.4%) | 25 (31.6%) | — |
|
|
1UICC: Union for International Cancer Control; 2GS: Gleason’s score; 3PSA: prostate-specific antigen; #Fisher exact test.
|
