Table 4.

Consistency analysis of models and radiologists compared with pathological or formal report.

	Accuracy	Kappa	Z score	P-Value
Model A	79.1 (72.8, 84.5)	0.744	22.9	<0.01
Model B	71.1 (64.4, 77.3)	0.649	20.0	<0.01
Model C	93.5 (89,2, 96.5)	0.861	12.2	<0.01
Model D	93.0 (88.6, 96.1)	0.851	12.1	<0.01
Model E	72.7 (62.9, 81.2)	0.541	7.31	<0.01
Model F	93.9 (87.3, 97.7)	0.901	11.8	<0.01
Model G	91.9 (84.7, 96.5)	0.867	11.4	<0.01
Reader 1	88.1 (82.8, 92.2)	0.853	25.7	<0.01
Reader 2	77.1 (70.7, 82.7)	0.723	22.8	<0.01
Reader 3	84.6 (78.8, 89.3)	0.810	24.4	<0.01
Reader consensus	86.1 (80.5, 90.5)	0.829	25.1	<0.01