Multi-View Separable Pyramid Network for AD Prediction at MCI Stage by 18F-FDG Brain PET Imaging

• Published in: IEEE transactions on medical imaging Vol. 40; no. 1; pp. 81 - 92
• Main Authors: Pan, Xiaoxi, Phan, Trong-Le, Adel, Mouloud, Fossati, Caroline, Gaidon, Thierry, Wojak, Julien, Guedj, Eric
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.01.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Alzheimer's disease; Cognitive ability; Convolution; Decision making; Deep learning; Dementia; Dementia disorders; F-FDG PET; Feature extraction; Fluorine isotopes; Impairment; Machine learning; Magnetic resonance imaging; Medical imaging; mild cognitive impairment; Neurodegenerative diseases; Neuroimaging; Older people; Positron emission; Positron emission tomography; Separable convolution; slice-wise CNN; Spatial data; spatial-wise CNN; Three-dimensional displays; Tomography; Two dimensional displays
• DOI: 10.1109/TMI.2020.3022591
• ISSN: 0278-0062; 1558-254X

Alzheimer's Disease (AD), one of the main causes of death in elderly people, is characterized by Mild Cognitive Impairment (MCI) at prodromal stage. Nevertheless, only part of MCI subjects could progress to AD. The main objective of this paper is thus to identify those who will develop a dementia of AD type among MCI patients. 18 F-FluoroDeoxyGlucose Positron Emission Tomography ( 18 F-FDG PET) serves as a neuroimaging modality for early diagnosis as it can reflect neural activity via measuring glucose uptake at resting-state. In this paper, we design a deep network on 18 F-FDG PET modality to address the problem of AD identification at early MCI stage. To this end, a Multi-view Separable Pyramid Network (MiSePyNet) is proposed, in which representations are learned from axial, coronal and sagittal views of PET scans so as to offer complementary information and then combined to make a decision jointly. Different from the widely and naturally used 3D convolution operations for 3D images, the proposed architecture is deployed with separable convolution from slice-wise to spatial-wise successively, which can retain the spatial information and reduce training parameters compared to 2D and 3D networks, respectively. Experiments on ADNI dataset show that the proposed method can yield better performance than both traditional and deep learning-based algorithms for predicting the progression of Mild Cognitive Impairment, with a classification accuracy of 83.05%.