Skip to main content
. 2022 Jun 15;32(12):8238–8246. doi: 10.1007/s00330-022-08909-x

Table 1.

Definition of different scenarios artificial intelligence (AI) and radiologists can be combined and used in screen-reading. Percentage of screen-readings performed by AI, by the radiologists (R1 and R2) and the corresponding reduction in screen-reading volume are also presented. The volume reduction relates to the screen-reading of mammograms prior to consensus and further assessments. An AI score of 1 indicates the lowest probability of malignancy by the AI system, and 10 indicates the highest probability

Scenarios Selection rate of AI and radiologists/setup based on AI score AI, % of screen-readings R1, % of screen-readings R2, % of screen-readings Reduced screening volume, %
Standard independent double reading 0 0 100 100 0
1 AI as one of two readers AI and R1 selects 5.8% each* 100 100 0 50
2 AI as one of two readers AI selects 10.1% (AI score = 10) and R1 selects 5.8%* 100 100 0 50
3 AI selects cases to be double read

AI score 1–5: negative

AI score 6–10: R1+R2

100 50 50 50
4 AI selects cases to be double read

AI score 1–7: negative

AI score 8–10: R1+R2

100 30 30 70
5 AI selects cases to be double read

AI score 1–9: negative

AI score 10: R1+R2

100 10 10 90
6 AI selects cases to be single read

AI score 1–5: negative

AI score 6–10: R1

100 50 0 75
7 AI selects cases to be single and double read

AI score 1–5: R1

AI score 6–10: R1+R2

100 100 50 25
8 AI selects cases to be single and double read

AI score 1–7: R1

AI score 8–10: R1+R2

100 100 30 35
9 AI selects cases to be single and double read

AI score 1–5: negative

AI score 6–7.5: R1

AI score 7.6–10: R1+R2

100 50 25 63
10 AI selects cases to be single and double read

AI score 1–5: negative

AI score 6–7.5: R1+R2

AI score 7.6–10: R1

100 50 25 63
11 AI selects cases to be recalled 3.2%* 100 0 0 100

*5.8% mimics the average selection rate/positive interpretations of the individual radiologists in the study sample, and 3.2% mimics the recall rate in the study sample after independent double reading