Table 6.

Machine learning prediction for benign vs nonbenign histology

	Accuracy, % (± SD)	False-negative rate, % (± SD)	False-positive rate, % (± SD)	Area under receiver operator curve, % (± SD)
Historical practice	45.28	0.00	69.05	65.48
Random forest classifier	83.24 ± 4.33	29.17 ± 17.18	14.09 ± 8.79	78.37 ± 4.96
Rule set	77.47 ± 2.71	11.67 ± 1.32	45.83 ± 20.83	61.64 ± 10.28

Compares the results of historical practice to random forest classifier and the simplified rule set using 4 measures of performance.