Table 3.

AI over-detection score indicating the impact of false positives in AI prediction on surgeons’ recognition.

Score	Type of false positivity	Number (N = 500)	Proportion	Cumulative proportion
5	No false positives	237	47.4%	47.4%
4	False positives but negligible impact	155	31.0%	78.4%
3	False positives but no impact on surgical judgment or procedure	90	18.0%	96.4%
2	False positives with minor impact on surgical judgment and procedure	14	2.8%	99.2%
1	False positives with an unacceptable negative impact on surgery	4	0.8%	100%