Table 4 Diagnostic Performance of Reviewers and DL-DSS for GB polyps larger than 10 mm.

From: Deep learning-based decision support system for the diagnosis of neoplastic gallbladder polyps on ultrasonography: Preliminary results

 

AUC

Comparison

Sensitivity

Specificity

Accuracy

F-1 Score

Step 1: Initial Image Analysis

Step 1 vs Step 2 (p value)

    

Reviewer A

0.92 [0.82–0.97]

0.96

90.9 (30/33) [75.7–98.1]

77.4 (24/31) [58.9–90.4]

89.1 (57/64) [67.6–94.4]

0.857 [0.769–0.895]

Reviewer B

0.68 [0.55–0.79]

<0.001

69.7 (23/33) [51.3–84.4]

54.8 (17/31) [36.0–72.7]

62.5 (40/64) [43.9–78.7]

0.657 [0.530–0.744]

Reviewer C

0.82 [0.70–0.90]

0.04

100 (33/33) [89.4–100.0]

41.9 (13/31) [24.5–60.9]

71.9 (46/64) [58.0–81.1]

0.786 [0.733–0.814]

Step 2: DL-DSS Alone

     

DL-DSS

0.92 [0.82–0.97]

 

78.8 (26/33) [61.1–91.0]

87.1 (27/31) [70.2–96.4]

82.8 (53/64) [65.5–93.6]

0.825 [0.705–0.896]

Step 3: DL-DSS Aided

Step 1 vs Step 3 (p value)

    

Reviewer A

0.94 [0.84–0.98]

0.16

87.9 (29/33) [71.8–96.6]

93.6 (29/31) [78.6–99.2]

90.7 (58/64) [75.1–97.9]

0.906 [0.807–0.953]

Reviewer B

0.89 [0.79–0.96]

<0.001

84.9 (28/33) [68.1–94.9]

87.1 (27/31) [70.2–96.4]

86.0 (55/64) [69.1–95.6]

0.862 [0.756–0.917]

Reviewer C

0.89 [0.79–0.96]

0.08

97.0 (32/33) [84.2–99.9]

54.8 (17/31) [36.0–72.7]

76.6 (49/31) [60.9–86.7]

0.810 [0.743–0.825]

  1. Note.- Data in brackets are 95% CI.