Table 1 Recent studies utilizing ML and DL techniques for kidney disease diagnosis.
Author | Technique | Performance Metrics |
|---|---|---|
Black et al.14 | CNN | Accuracy: 83.2% |
Aziyus et al.15 | ResNet | Recognition Rate: 74% |
Kazemi et al.16 | Ensemble Learning Model | Accuracy: 97.1% |
Chaitanya et al.17 | Machine Learning Classifiers | Accuracy: 78% |
Convolutional Neural Network-Support Vector Machines | Accuracy: 96.59% | |
Hallscheidt et al.19 | - | Accuracy: 75%, Sensitivity: 85% |
Liu et al.20 | Machine Learning Techniques | Sensitivity: 95% |
Feng et al.21 | Support Vector Machines with Recursive Feature Elimination | Accuracy: 93.9%, Sensitivity: 87.8%, Specificity: 100% |
Kocak et al.22 | ANN with Adaptive Boosting | Accuracy: 84.6%, Sensitivity: 69.2%, Specificity: 100% |
Muhamed Ali et al.23 | Neighborhood Component Analysis and LSTM | Accuracy: 95%, Matthews Correlation Coefficient: 92% |
Zhang et al.24 | CNN | Accuracy: 97% |
Sun et al.25 | SVM | Sensitivity: 90.0%, Specificity: 89.1% |
Tabibu et al.26 | CNN | Accuracy: 93.39% |
Zabihollahy et al.27 | CNN | Accuracy: 83.8%, Precision: 90.32%, Recall: 84.21% |
Vendrami et al.28 | Random Forest, Gradient Boosting Machine, and Recursive Sectioning | Accuracy: 81.2% |
Skounakis et al.29 | Region Growing Semi-automatic Segmentation Algorithm | Accuracy: 97.2% |
Kahani et al.30 | Dual-Energy Kidney, Ureter and bladder | Overall accuracy of 92 ± 2% with acceptable noise robustness |
Liu et al.31 | CNN | Sensitivity: 96.4 |
Chittora et al.32 | SVM | Accuracy: 94.6% |
