Table 4.

Literature review of MDD diagnosis using ML and DL methods.

Reference	Model	Year	Modality	Subjects	Training Set/Test Set	Accuracy (%)
[102]	SVM	2015	rs-fMRI	21 BD, 25 MDD	leave-one-out	92.1
[103]	SVM	2018	T1-w	26 BD, 26 MDD	leave-two-out	75
[104]	SVM	2018	DTI	31 BD, 36 MDD	-	68.3
[105]	SVM	2017	rs-fMRI	19 cMDD, 19rMDD, 19 HC	leave-one-out	cMDD/HC-87 rMDD/HC-84 cMDD/rMDD-89
[106]	SVM	2021	T1-w	66 MDD	leave-one-out	78.59
[107]	SVM	2017	T1-w	19 GAD, 14 MDD, 24 HC	leave-one-out	GAD+MDD/HC -90.10, GAD/MDD-67.46
[108]	SVM	2017	rs-fMRI	38 MDD, 28 HC	-	97.54
[109]	SVM-based	2014	rs-fMRI	24 MDD, 29 HC	leave-one-out	92.5
[110]	GAN	2020	rs-fMRI	269 MDD, 286 HC	10-fold Cross-Validation	80.7
[111]	GNN	2020	rs-fMRI	29 MDD, 44 HC	10-fold Cross-Validation	74.1