Assisted Diagnosis of Alzheimer’s Disease Based on Deep Learning and Multimodal Feature Fusion

Table 5

Comparison results of three feature fusion methods and single-modal method (%).

AD versus NC	Acc	Sen	Spec	Precision	Recall	f1-score	AUC
sMRI	88.0	70.0	100.0	100.0	70.0	82.4	86.0
fMRI	88.0	80.0	93.3	88.9	80.0	84.2	88.7
CCA	84.0	80.0	86.7	80.0	80.0	80.0	88.0
Series fusion	88.0	80.0	93.3	88.9	80.0	84.2	90.0
KCCA	96.0	100.0	93.3	90.9	100.0	95.2	99.3
AD versus MCI	Acc	Sen	Spec	Precision	Recall	f1-score	AUC
sMRI	90.0	90.0	90.0	90.0	90.0	90.0	97.5
fMRI	85.0	90.0	80.0	81.8	90.0	85.7	87.0
CCA	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Series fusion	90.0	90.0	90.0	90.0	90.0	90.0	98.0
KCCA	100.0	100.0	100.0	100.0	100.0	100.0	100.0
MCI versus NC	Acc	Sen	Spec	Precision	Recall	f1-score	AUC
sMRI	76.0	50.0	93.3	83.3	50.0	62.5	63.3
fMRI	80.0	80.0	80.0	72.7	80.0	76.2	83.3
CCA	76.0	50.0	93.3	83.3	50.0	62.5	69.33
Series fusion	76.0	50.0	93.3	83.3	50.0	62.5	64.0
KCCA	100.0	100.0	100.0	100.0	100.0	100.0	100.0
LMCI versus EMCI	Acc	Sen	Spec	Precision	Recall	f1-score	AUC
sMRI	70.0	40.0	100.0	100.0	40.0	57.1	64.0
fMRI	100.0	100.0	100.0	100.0	100.0	100.0	100.0
CCA	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Series fusion	80.0	80.0	80.0	80.0	80.0	80.0	72.0
KCCA	100.0	100.0	100.0	100.0	100.0	100.0	100.0