Table 1 Use of different ML classifiers for AD diagnosis using MRI scans
From: Denouements of machine learning and multimodal diagnostic classification of Alzheimer’s disease
Technique | Classifier used | Modality | Data | Accuracy | References |
|---|---|---|---|---|---|
Region growing | Artificial neural network | MRI | KHMC | (100%) | [72] |
3D inception | CNN | MRI | ADNI | AD/NC (93.30%) AD/MCI (86.70%) MCI/NC (73.30%) | [73] |
Fractal analysis | KNN, SVM (linear), SVM (RBF), HLP (polynomial) | MRI | OASIS | SET-1: KNN (61.76%), SVM (linear) (59.41%), SVM (RBF) (64.71%), HLP (polynomial) (65.29%) SET-2: KNN (72.31%), SVM (linear) (75.38%), SVM (RBF) (76.15%), HLP (polynomial) (86.15%) SET-3: KNN (59.00%), SVM (linear) (64.00%), SVM (RBF) (68.00%), HLP (polynomial) (67.50%) | [74] |
Salient brain patterns | SVM, NN | MRI | OASIS | SVM (84.21%) NN (65.78%) | [75] |
K-OPLS, OPLS | Multivariate data analysis | MRI | ADNI | K-OPLS (88.70%) OPLS (88.40%) | [76] |
Hippocampal shape feature | SVM | MRI | ADNI | CASE 1 (90.40%) CASE 2 (89.40%) CASE 3 (90.40%) CASE 4 (93.60%) | [77] |
ROI | Naïve Baye, SVM, KNN | MRI | OASIS | Naive Baye (90.00%) SVM (95.00%) KNN (95.00%) | [78] |
Hippocampus volume, tensor-based morphometry, cortical thickness | LDA | MRI | ADNI | HC vs AD (89.00%) HC vs P-MCI (84.00%) S-MCI vs P-MCI (68.00%) | [79] |
Multivariate techniques | Logistic regression | MRI | Self | AD vs HC (83.00%) | [80] |
Gray-level co-occurrence matrix | Adaboost, KNN | MRI | OASIS | AD vs NC: Adaboost (100%), KNN (92.75%) AD vs MCI: Adaboost (100%), KNN (92.31%) MCI vs NC: Adaboost (90.28%), KNN (83.33%) | [81] |