Skip to main content
. Author manuscript; available in PMC: 2024 Jul 28.
Published in final edited form as: J Comput Assist Tomogr. 2023 Jul 28;47(6):845–849. doi: 10.1097/RCT.0000000000001503

Table 3.

Perceived Usefulness of Artificial Intelligence Program given Hypothetical System Performance

Hypothetical System Performance Rated as Useful (%)
Percentage of pancreatic cancer detected by AI
 100%
 95%
 90%
 85%
 80%
91/101 (90.1)
84/100 (84.0)
57/100 (57.0)
28/97 (28.9)
18/96 (18.8)
Number of false positives provided by AI for second look
 0 false positives/100 normal pancreata
 5 false positives/100 normal pancreata
 10 false positives/100 normal pancreata
 15 false positives/100 normal pancreata
 20 false positives/100 normal pancreata

66/90 (73.3)
86/100 (86.0)
54/96 (56.3)
17/91 (18.7)
15/90 (16.7)