Table 1 Characteristics of 579,583 screening mammograms interpreted from September 2021 to December 2022
From: Equitable impact of an AI-driven breast cancer screening workflow in real-world US-wide deployment
| Ā | Overall | Cohort | |
|---|---|---|---|
Standard of care | Multistage AI-driven workflow | ||
Examinations, number (%) | |||
Whole population | 579,583 (100.0) | 370,692 (100.0) | 208,891 (100.0) |
Patient race and ethnicity, number (%) | |||
Asian | 48,552 (8.38) | 30,706 (8.28) | 17,846 (8.54) |
Black, non-Hispanic | 154,300 (26.62) | 100,559 (27.13) | 53,741 (25.73) |
Hispanic | 123,668 (21.34) | 78,915 (21.29) | 44,753 (21.42) |
White, non-Hispanic | 193,974 (33.47) | 123,654 (33.36) | 70,320 (33.66) |
Other race | 107,641 (18.57) | 67,564 (18.23) | 40,077 (19.19) |
Breast density, number (%) | |||
A | 40,095 (6.92) | 25,929 (6.99) | 14,166 (6.78) |
B | 275,476 (47.53) | 174,248 (47.01) | 101,228 (48.46) |
C | 241,446 (41.66) | 155,759 (42.02) | 85,687 (41.02) |
D | 22,566 (3.89) | 14,756 (3.98) | 7,810 (3.74) |
Age group, number (%) | |||
<40 | 5,414 (0.93) | 3,508 (0.95) | 1,906 (0.91) |
40ā49 | 131,739 (22.73) | 86,158 (23.24) | 45,581 (21.82) |
50ā59 | 171,447 (29.58) | 110,805 (29.89) | 60,642 (29.03) |
60ā69 | 164,688 (28.41) | 104,031 (28.06) | 60,657 (29.04) |
70ā79 | 90,745 (15.66) | 56,570 (15.26) | 34,175 (16.36) |
ā„80 | 15,550 (2.68) | 9,620 (2.60) | 5,930 (2.84) |
Age group additional breakdown, number (%) | |||
<55 | 219,157 (37.81) | 142,895 (38.55) | 76,262 (36.51) |
55ā64 | 177,367 (30.60) | 113,423 (30.60) | 63,944 (30.61) |
65+ | 183,059 (31.58) | 114,374 (30.85) | 68,685 (32.88) |
