Background
Methods
Datasets
Normal | minor changes | major changes | Total | |
---|---|---|---|---|
case number | 221 | 40 | 8 | 269 |
Height (m) | 0.94 ± 0.18 | 0.89 ± 0.20 | 1.10 ± 0.34 | 0.94 ± 0.19 |
Weight (kg) | 14.26 ± 6.39 | 13.28 ± 8.16 | 26.8 ± 26.5 | 14.49 ± 8.14 |
CTR | 0.54 ± 0.05 | 0.55 ± 0.05 | 0.54 ± 0.03 | 0.54 ± 0.05 |
ASD size (mm) | 10.21 ± 3.29 | 10.44 ± 3.00 | 11.31 ± 5.52 | 10.28 ± 3.32 |
LVEF (%) | 66.19 ± 2.29 | 65.98 ± 2.47 | 64.25 ± 1.98 | 66.10 ± 2.33 |
BMI (kg/m2) | 15.59 ± 1.83 | 15.88 ± 2.27 | 17.63 ± 5.04 | 15.70 ± 2.07 |
BSA (m2) | 0.59 ± 0.19 | 0.55 ± 0.22 | 0.87 ± 0.60 | 0.59 ± 0.22 |
Operating age (year) | 3.06 ± 2.02 | 2.66 ± 2.60 | 5.50 ± 4.54 | 3.07 ± 2.25 |
Input variable | Ranges | Input variable | Ranges |
---|---|---|---|
Sexa | 0–1 | Potassium (mmol/L) | 3-5.7 |
Height (m) | 0.64–1.30 | Sodium (mmol/L) | 134–146 |
Weight (kg) | 7–35 | CK (U/L) | 13.83-276.87 |
Lung blooda | 0–1 | CK-MB (U/L) | 8–43 |
Precardiac spacea | 0–1 | PT (S) | 9.8–13.7 |
CTR | 0.4–0.65 | INR | 0.8–1.19 |
Right heart enlargementa | 0–1 | APTT (S) | 20.1–31.8 |
ASD size (mm) | 5–22 | TT (S) | 14-119.6 |
LVEF (%) | 60–73 | Fibrinogen (mg/dl) | 0.97–3.5 |
Leukocyte (×109/L) | 4-13.88 | FDP (µg/ml) | 2.5-87.94 |
Erythrocyte (×109/L) | 3.5–5.33 | D dimer (mg/L) | 0-3.02 |
Hemoglobin (g/L) | 86–152 | BMI (kg/m2) | 0.19–14.78 |
Platelets (×109/L) | 100–443 | BSA (m2) | 0.345-1.124 |
Albumin (g/L) | 40-54.98 | Operating age (year) | 0.50-14.78 |
ALT (U/L) | 5.98–109.6 | Creatinine (µmol/l) | 4.78-318.69 |
AST (U/L) | 10.21–95.17 | Urea (mmol/L) | 0.99–7.34 |
Synthetic minority oversampling technique algorithm
Random forest
Support vector machine
K-Nearest neighbor algorithm
Logistic regression
AdaBoost
Decision tree
Model evaluation
Result
Comparison of dataset imbalance processing methods
Accuracy (%) | Sensitivity (%) | Specificity (%) | MCC | AUC | |
---|---|---|---|---|---|
No-Synthetic Minority Oversampling Technique | 81.78 | 2.50 | 95.63 | -0.0335 | 0.5665 |
Synthetic Minority Oversampling Technique | 94.65 | 92.50 | 94.98 | 0.7980 | 0.8956 |
Classifier | Accuracy (%) | Sensitivity (%) | Specificity (%) | MCC | AUC |
---|---|---|---|---|---|
Logistic Regression | 78.60 | 77.50 | 78.76 | 0.4231 | 0.7286 |
K-Nearest Neighbor | 78.93 | 97.50 | 76.06 | 0.5295 | 0.8219 |
Decision Tree | 82.61 | 80.00 | 83.01 | 0.4927 | 0.8035 |
AdaBoost | 84.95 | 87.50 | 84.56 | 0.5658 | 0.7489 |
Support Vector Machine | 89.30 | 95.00 | 88.42 | 0.6774 | 0.8744 |
Random Forest | 94.65 | 92.50 | 94.98 | 0.7980 | 0.8956 |