Table 1. Diagnosis of major depressive disorder from structural MRI studies.
| Study, year | MRI | Healthy controls | Patients | Diagnosis | Severity –HAMD: mean (s.d.) | Medication | Comparison | Classifier | Feature | Accuracy | Sensitivity | Specificity | P | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| N (males) | Age, years (s.d.) | N (males) | Age, years (s.d.) | ||||||||||||
| Costafreda10 2009 | 1.5 T | 37 (9) | 42.8 (6.7) | 37 (9) | 43.2 (8.8) | MDD | 20.6 (2.2) | Med free | MDD v. HV | SVM | GM | 67.6 | 64.9 | 70.3 | 0.027 |
| Gong11 | 3 T | 23 (12) | 38.2 | 23 (10) | 39.2 (12.9) | MDD | 24.2 (3.8) | Med naïve | MDD v. HV | SVM | GM | 76.1 | 69.6 | 82.6 | <0.001 |
| 2011 | WM | 84.7 | 73.9 | 95.7 | <0.001 | ||||||||||
| GM+WM | 76.1 | 73.9 | 78.3 | <0.001 | |||||||||||
| 23 (14) | 40.4 (12.6) | TRD | 23.5 (5.4) | Med naïve | TRD v. HV | GM | 67.4 | 65.2 | 69.6 | 0.01 | |||||
| WM | 58.7 | 60.9 | 56.5 | 0.13 | |||||||||||
| GM+WM | 65.2 | 65.2 | 65.2 | 0.02 | |||||||||||
| Liu12 2012 | 1.5 T | 17 (10) | 24.2 (4.4) | 17 (10) | 26.7 (7.7) | MDD | 25.6 (6.3) | Med naïve | MDD v. HV | Searchlight- | GM | 82.4 | – | – | – |
| PCA-SVM | WM | 91.2 | – | – | – | ||||||||||
| RFE-SVM | GM | 70.6 | – | – | – | ||||||||||
| WM | 76.5 | – | – | – | |||||||||||
| LLE- C Means | GM | 76.5 | – | – | – | ||||||||||
| WM | 88.2 | – | – | – | |||||||||||
| LLE-SVM | GM | 82.4 | – | – | – | ||||||||||
| WM | 88.2 | – | – | – | |||||||||||
| 18 (11) | 27.4 (7.7) | TRD | 23.9 (3.7) | On medsa | TRD v. HV | Searchlight- | GM | 85.7 | – | – | – | ||||
| PCA-SVM | WM | 85.7 | – | – | – | ||||||||||
| RFE-SVM | GM | 77.1 | – | – | – | ||||||||||
| WM | 85.7 | – | – | – | |||||||||||
| LLE-C Means | GM | 77.1 | – | – | – | ||||||||||
| WM | 85.7 | – | – | – | |||||||||||
| LLE-SVM | GM | 77.1 | – | – | – | ||||||||||
| WM | 85.7 | – | – | – | |||||||||||
| +Mwangi13 2012 | 1.5 T | 18 (7) | 40.6 (10.3) | 15 (6) | 46.1 (12.5) | TUD | 23.2 (4.3) | On meds | TUD v. HV | VBM-FBM-SVM | GM | 90.3 | 93.3 | 87.5 | 1×10–7b |
| 14 (7) | 43.0 (13.2) | 15 (5) | 44.7 (10.0) | TRD | 27.9 (5.8) | RVM | GM | 87.1 | 86.7 | 87.5 | 1×10–7b | ||||
| ++Kipli14 | – | – | – | – | – | – | – | – | – | SVM-EM | GM+WM+CSF | 85.3 | – | – | – |
| 2013 | Information gain-Rand Tree | 85.3 | – | – | – | ||||||||||
| SVM-K Means | 82.3 | – | – | – | |||||||||||
| Serpa15 2014 | 1.5 T | 38 (8) | 29.7 (7.9) | 19 (4) | 29.1 (8.3) | pMDD | 16.1c | On medsa | pMDD v. HV | SVM | GM+WM+ventricles | 59.6 | 31.6 | 73.7 | – |
| Qiu 201416 | 3 T | 32 (23) | 35.0 (11.2) | 32 (23) | 34.9 (11.1) | MDD | 24.3 (5.1) | Med naïve | MDD v. HV | SVM | Cortical thickness | 69 | 66 | 72 | 0.002 |
| Volume | 66 | 63 | 69 | 0.005 | |||||||||||
| Plial area | 69 | 69 | 69 | 0.001 | |||||||||||
| Curvature | 48 | 47 | 50 | 0.63 | |||||||||||
| Area | 59 | 66 | 53 | 0.10 | |||||||||||
| Sulcal depth | 58 | 56 | 59 | 0.12 | |||||||||||
| Jacobian Metric Distortion | 67 | 63 | 72 | 0.002 | |||||||||||
| +++Combination parametres | 69 | 69 | 69 | 0.002 | |||||||||||
Med, Medication; HAMD, 17-item Hamilton Depression Rating Scale; MDD, major depressive disorder; TRD, treatment-resistant depression; pMDD, psychotic MDD; HV, healthy volunteers; GM, grey matter; WM, white matter; CSF, cerebrospinal fluid; SVM, support vector machines; PCA, principle component analysis; RFE, recursive feature elimination; LLE, locally linear embedding; VBM, voxel based morphometry; RVM, relevance vector machine; FBM, feature based morphometry; EM, expectation-maximisation dustering algorithm; KMeans, simple K means classification via clustering; TUD, treatment unresponsive patients.
Depression status of MDD patients: first-episode – Liu (2012), Qiu (2014), Serpa (2014); first-episode and recurrent – Costafreda (2009); recurrent: Mwangi (2012); not stated: Gong (2011), Kipli (2013).
Some of the patients were medication free.
χ2P.
31-item HAMD.
Mwangi (2012): data were randomly divided into two sets (training set, testing set) of equal number of patients and controls (n=31). In patients, depression was considered to be treatment unresponsive. Minimum duration of illness was >3 months with antidepressant medication.
Kipli (2013), accuracy of 82.3% also obtained with other classifiers: information gain:-J48, information gain-RandomForest, SVM-K Means, SVM-RandomForest, ReliefF-RandomTree, all-naïve bayes.
Combined parameters: Qiu (2014) integrated all the morphometric parameters (i.e. cortical thickness, volume, plial area, curvature area, sulcal depth and Jacobian metric distortion) of the left and right hemispheres within a single model to investigate the discriminative power of the resulting combination.