Table 2 Participant characteristics per phase
From: Prospective implementation of AI-assisted screen reading to improve early detection of breast cancer
Variable | Initial pilot (n = 3,746) | Extended pilot (n = 9,112) | Live use (n = 15,953) |
|---|---|---|---|
Age (continuous, years), mean (s.d.) | 58.2 (11.0) | 58.2 (10.7) | 58.6 (10.5) |
Age group, n (%) | |||
≤35 years | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) |
36–45 years | 518 (13.8%) | 1,149 (12.6%) | 1,583 (9.9%) |
46–55 years | 1,218 (32.5%) | 2,998 (32.9%) | 5,420 (34.0%) |
56–65 years | 940 (25.1%) | 2,493 (27.4%) | 4,699 (29.5%) |
66–75 years | 806 (21.5%) | 1,902 (20.9%) | 3,196 (20.0%) |
>75 years | 264 (7.0%) | 570 (6.3%) | 1,055 (6.6%) |
Family historya, n (%) | |||
No | 3,620 (96.6%) | 8,838 (97.0%) | 15,338 (96.1%) |
Yes | 126 (3.4%) | 274 (3.0%) | 615 (3.9%) |
Tábar classification of parenchymal patternsb, n (%) | |||
1 | 1,506 (40.2%) | 3,950 (43.3%) | 7,468 (46.8%) |
2 | 729 (19.5%) | 1,697 (18.6%) | 2,921 (18.3%) |
3 | 336 (9.0%) | 679 (7.5%) | 465 (2.9%) |
4 | 365 (9.7%) | 848 (9.3%) | 1,423 (8.9%) |
5 | 114 (3.0%) | 246 (2.7%) | 310 (1.9%) |
Missing | 696 (18.6%) | 1,692 (18.6%) | 3,366 (21.1%) |
