Machine-Learning Classifier for Patients with Major Depressive Disorder: Multifeature Approach Based on a High-Order Minimum Spanning Tree Functional Brain Network

Table 6

Comparison of classification results from different methods.

Method	Research	Disease	Accuracy	Sensitivity	Specificity	AUC
Partial FC	Guo et al., 2013 [36]	MDD	86.01%	-	-	-
	Qiao et al., 2016 [37]	MCI	89.01%	86.67%	91.30%	-
	This study	MDD	63.06%	50.56%	87.37%	71.02%
Pearson FC	Wong et al., 2012 [38]	MDD	63.00%	40.00%	83.00%	-
	Liu et al., 2015 [39]	SAD	82.50%	85.00%	80.00%	-
	This study	MDD	66.67%	46.43%	81.58%	74.46%
High-order FC	Chen et al., 2016 [14]	MCI	88.14%	86.21%	90.00%	92.99%
	This study	MDD	92.51%	88.51%	93.19%	92.83%
Frequent subgraph	Du et al., 2016 [26]	ADHD	94.91%	93.22%	96.94%	96.90%
	Fei et al., 2014 [25]	MCI	97.30%	-	-	95.83%
Frequent and local cluster coefficient	Wang et al., 2014 [27]	MCI	97.27%	-	-	92.00%
High-order MST FC	Subgraph features	MDD	73.32%	80.36%	67.58%	75.67%
	Minimum spanning tree features	MDD	94.04%	98.26%	92.50%	97.84%
	Proposed	MDD	97.54%	100.00%	96.67%	99.06%

FC, functional connectivity; MST, minimum spanning tree; SAD, social anxiety disorder; MCI, mild cognitive impairment; ADHD, attention deficit hyperactivity disorder; MDD, major depressive disorder; AUC, area under receiver operating characteristic (ROC) curve.