Table 3 Diagnostic ability of cyst fluid analysis, cytology, and AI in differentiating malignant and benign lesions.

From: Diagnostic ability of artificial intelligence using deep learning analysis of cyst fluid in differentiating malignant from benign pancreatic cystic lesions

 

AI

AI using only CEA

CEA

Cytology

p*

p†

p**

Sensitivity

95.7% (22/23)

91.3% (21/23)

60.9% (14/23)

47.8% (11/23)

1.000

0.021

0.001

Specificity

91.9% (57/62)

88.7% (55/62)

75.8% (47/62)

100.0% (62/62)

0.625

0.031

<0.001

PPV

81.5% (22/27)

75.0% (21/28)

48.3% (14/29)

100.0% (11/11)

—

—

—

NPV

98.3% (57/58)

96.5% (55/57)

83.9% (47/56)

83.8% (62/74)

—

—

—

Accuracy

92.9% (79/85)

89.4% (76/85)

71.8% (61/85)

85.9% (73/85)

0.375

<0.001

0.210

AUC

0.966

0.956

0.719

0.739

   
  1. AI, artificial intelligence; AUC, area under the receiver-operating characteristics curve; CEA, carcinoembryonic antigen; CA19-9, carbohydrate antigen 19-9; CA125, carbohydrate antigen 125; NPV, negative predictive value; PPV, positive predictive value.
  2. p* McNemar test; AI vs AI using only CEA.
  3. p† McNemar test; AI vs CEA.
  4. p** McNemar test; AI vs cytology.
Back to article page