Table 1 Clinical characteristics of patient cohorts
From: Artificial intelligence-assisted prostate cancer diagnosis for reduced use of immunohistochemistry
Cohorts | SUH | SFR | SCH |
|---|---|---|---|
Dataset type | Internal | External | External |
Pathologists represented | 12 | N/A | 5 |
Scanner model | Hamamatsu NanoZoomer S60 | Philips IntelliSite UFS | Philips IntelliSite UFS |
Number of patients | n = 99 | n = 49 | n = 75 |
|---|---|---|---|
Age (years) | |||
≤49 years | 1 (1.0%) | 0 (0.0%) | 0 (0.0%) |
50–54 years | 3 (3.0%) | 2 (4.1%) | 1 (1.3%) |
55–59 years | 9 (9.1%) | 8 (16.3%) | 5 (6.7%) |
60–64 years | 25 (25.3%) | 4 (8.2%) | 10 (13.3%) |
65–69 years | 23 (23.2%) | 14 (28.6%) | 16 (21.3%) |
≥70 years | 38 (38.4%) | 21 (42.8%) | 43 (57.4%) |
Prostate-specific antigen (ng/ml) | |||
0–3 ng/mL | 7 (7.1%) | 1 (2.0%) | 1 (1.3%) |
>3–5 ng/mL | 15 (15.1%) | 4 (8.2%) | 6 (8.0%) |
>5–10 ng/mL | 50 (50.5%) | 29 (59.2%) | 18 (24.0%) |
≥10 ng/mL | 26 (26.3%) | 8 (16.3%) | 13 (17.3%) |
«Elevated» | 0 (0.0%) | 0 (0.0%) | 8 (10.7%) |
Missing | 1 (1.0%) | 7 (14.3%) | 29 (38.7%) |
Number of WSIs | n = 234 | n = 112 | n = 164 |
|---|---|---|---|
ISUP grades (Gleason scores) | |||
Benign | 129 (55.1%) | 66 (58.9%) | 65 (39.6%) |
ISUP 1 (3 + 3) | 60 (25.6%) | 41 (36.6%) | 50* (30.5 %) |
ISUP 2 (3 + 4) | 15 (6.4%) | 3 (2.7%) | 17* (10.4%) |
ISUP 3 (4 + 3) | 10 (4.3%) | 1 (0.9%) | 24* (14.6%) |
ISUP 4 (4 + 4, 3 + 5, 5 + 3) | 9 (3.9%) | 1 (0.9%) | 3* (1.8%) |
ISUP 5 (4 + 5, 5 + 4, 5 + 5) | 11 (4.7%) | 0 (0.0%) | 5* (3.1%) |
Cancer length (mm) | |||
No cancer | 129 (55.1%) | 66 (58.9%) | 65 (39.6%) |
>0–1 mm | 23 (9.8%) | 1 (0.9%) | 3* (1.8%) |
>1–5 mm | 42 (18.0%) | 21 (18.8%) | 30* (18.3%) |
>5–10 mm | 15 (6.4%) | 6 (5.4%) | 8* (4.9%) |
>10 mm | 25 (10.7%) | 13 (11.6%) | 57* (34.8%) |
Missing | 0 (0.0%) | 5 (4.4%) | 1 (0.6%) |
