Skip to main content
. 2020 Mar 2;8(3):e16334. doi: 10.2196/16334

Table 2.

Diagnostic performance of the readers in two reading modes with a Breast Imaging Reporting and Data System Category 4A threshold.

Lesions Novice Reader Expert Reader



Without CADa Combined with CAD P value Without CAD Combined with CAD P value



+ +
+ +
Lesions≤2 cm

Pathology







+ 44 22 64 23
52 9 69 33


27 157 7 156
19 170 2 146

Sensitivityb 61.97 90.14 <.001 73.24 97.18 <.001

Specificityb 87.71 87.15 .83 94.97 81.56 .004

PPVc 66.66 73.56 .22 85.25 67.65 .005

NPVd 85.33 95.71 .008 89.95 98.65 .005

Accuracy 80.40 88.00 .12 88.8 86 .52

AUCe (95% CI) 0.74 (0.67-0.82) 0.88 (0.83-0.93) <.001 0.84 (0.77-0.90) 0.90 (0.86-0.94) .002
Lesions>2 cm

Pathology











+ 61 11 79 28
67 4 79 23


18 58 0 41
12 65 0 46

Sensitivity 77.22 100 <.001 86.67 100 <.001

Specificity 84.06 59.42 <.001 96.72 66.66 <.001

PPV 84.72 73.83 .05 96.30 77.45 <.001

NPV 76.32 100 <.001 88.06 100 <.001

Accuracy 80.41 81.08 .86 91.74 84.46 .13

AUC (95% CI) 0.81 (0.73-0.88) 0.80 (0.72-0.87) .81 0.90 (0.84-0.95) 0.83 (0.76-0.91) .03
All lesions

Pathology







+ 105 33 143 51
119 13 148 52


45 215 7 197
31 235 2 96

Sensitivity 70 95.33 <.001 79.33 98.66 <.001

Specificity 86.69 79.43 .13 94.75 79.03 .001

PPV 76.08 73.71 .74 90.15 74.00 .003

NPV 82.69 96.57 .001 88.34 98.98 .002

Accuracy 80.40 85.43 .35 88.94 86.43 .52

AUC (95% CI) 0.78 (0.73-0.83) 0.87 (0.84-0.91) <.001 0.87 (0.83-0.91) 0.89 (0.85-0.92) .29

aCAD: computer-aided diagnosis.

bBreast Imaging Reporting and Data System assessment categories 4B, 4C, and 5 were considered positive for cancer for the calculation of sensitivity and specificity.

cPPV: positive predictive value.

dNPV: negative predictive value.

eAUC: area under the receiver operating characteristic curve.