Skip to main content

Table 2 Different types of AI models

From: Accuracy of artificial intelligence algorithms in predicting acute respiratory distress syndrome: a systematic review and meta-analysis

Model

Study

Sensitivity

Sen-I2

Specificity

Spe-I2

AUC

XGB

7

0.84 (0.75–0.89)

98.11 (97.44–98.78)

0.89 (0.83–0.93)

99.01 (98.73–99.30)

0.93 (0.90–0.95)

RF

5

0.80 (0.73–0.86)

75.58 (53.66–97.50)

0.90 (0.80–0.96)

94.03 (90.34–97.72)

0.89 (0.86–0.91)

LR

17

0.80 (0.71–0.86)

97.91 (97.44–98.39)

0.80 (0.74–0.84)

96.98 (96.22–97.75)

0.86 (0.83–0.89)

CNN

7

0.80 (0.72–0.86)

95.30 (93.07–97.52)

0.91 (0.83–0.96)

99.47 (99.35–99.59)

0.91 (0.88–0.93)

SVM

12

0.72 (0.54–0.85)

94.43 (92.36–96.49)

0.89 (0.82–0.94)

98.80 (98.53–99.07)

0.90 (0.87–0.92)

ANN

5

0.74 (0.64–0.83)

95.72 (93.32–98.12)

0.88 (0.77–0.94)

98.92 (98.53–99.31)

0.86 (0.83–0.89)

  1. XGB Xtreme Gradient Boosting, RF Random forest, LR Logistic regression, CNN Convolutional neural network, SVM Support vector machine, ANN Artificial neural network