.png){kind=icon}
.png){kind=icon}
Return
Integrating Structural and Metabolic Neuroimaging Biomarkers for Alzheimer’s Disease Diagnosis and Cognitive Score Estimation via Cross-Modal Gated Learning
.png){kind=icon}
OA
AI
C
L
S
DOI：10.3390/biology15131091.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Alzheimer’s disease is associated with structural brain atrophy and metabolic dysfunction, which can be observed using structural magnetic resonance imaging (sMRI) and fluorodeoxyglucose positron emission tomography (FDG-PET). However, effectively integrating these complementary neuroimaging biomarkers remains challenging. In this study, we propose CGMF-Net, a cross-modal gated learning framework for Alzheimer’s disease classification and cognitive score estimation. The model enhances consistent structural and metabolic responses before fusion and further models complementary interactions between the two modalities. Experiments on ADNI data show that CGMF-Net improves AD classification, generalizes favorably from ADNI-2 to ADNI-1, and estimates clinically relevant cognitive scores. These findings suggest that integrating structural and metabolic biomarkers through cross-modal gated learning may support computer-aided assessment of Alzheimer’s disease.
Keywords:
Alzheimer’s disease
structural MRI
FDG-PET
multimodal neuroimaging
cognitive score estimation
cross-modal gated learning
multi-task learning
HSIC
Journal
IF:
3.5
Papers:
7.1K
Citations:
2.3W
Organization
Cited Papers