Skip to main content
. 2022 Jul 5;13:100116. doi: 10.1016/j.jpi.2022.100116

Table 2.

Difference in agreement rates between manual reads vs. ground truth, and image analysis vs. ground truth.

MR status
GT status IA status Amplified Non-amplified Total
Amplified Amplified 309 3 312
Non-amplified 2 7 9
Total 311 10 321
MR PPA n/N (%) (95% CI) 311/321(96.9) (94.2, 99.0)
IA PPA n/N (%) (95% CI) 312/321(97.2) (95.0, 99.3)
Difference of PPA (IA - MR) n/N (%) (95% CI) 1/321 (0.3) (-0.9, 1.8)
Non-amplified Amplified 8 11 19
Non-amplified 0 312 312
Total 8 323 331
MR NPA n/N (%) (95% CI) 323/331 (97.6) (95.8, 99.4)
IA NPA n/N (%) (95% CI) 312/331 (94.3) (90.8, 97.3)
Difference of NPA (IA - MR) n/N (%) (95% CI) -11/331 (-3.3) (-6.2, -1.2)
Overall Amplified 317 14 331
Non-amplified 2 319 321
Total 319 333 652
MR OPA n/N (%) (95% CI) 634/652 (97.2) (95.7, 98.6)
IA OPA n/N (%) (95% CI) 624/652 (95.7) (93.8, 97.5)
Difference of OPA (IA - MR) n/N (%) (95% CI) -10/652 (-1.5) (-3.1, -0.2)

GT ground truth, IA image analysis, MR manual read; PPA positive percent agreement, NPA negative percent agreement, OPA overall percent agreement, CI confidence interval.

Note: 2-sided 95% confidence intervals were calculated using the percentile bootstrap method with cases stratified by the 4 screening bins.

Note: This analysis includes all 3 study Reader Pathologists' results.