Cortical graph neural network for AD and MCI diagnosis and transfer learning across populations

Abstract

Combining machine learning with neuroimaging data has a great potential for early diagnosis of mild cognitive impairment (MCI) and Alzheimer's disease (AD). However, it remains unclear how well the classifiers built on one population can predict MCI/AD diagnosis of other populations. This study aimed to employ a spectral graph convolutional neural network (graph-CNN), that incorporated cortical thickness and geometry, to identify MCI and AD based on 3089 T 1 -weighted MRI data of the ADNI-2 cohort, and to evaluate its feasibility to predict AD in the ADNI-1 cohort (n = 3602) and an Asian cohort (n = 347). For the ADNI-2 cohort, the graph-CNN showed classification accuracy of controls (CN) vs. AD at 85.8% and early MCI (EMCI) vs. AD at 79.2%, followed by CN vs. late MCI (LMCI) (69.3%), LMCI vs. AD (65.2%), EMCI vs. LMCI (60.9%), and CN vs. EMCI (51.8%). We demonstrated the robustness of the graph-CNN among the existing deep learning approaches, such as Euclidean-domain-based multilayer network and 1D CNN on cortical thickness, and 2D and 3D CNNs on T 1 -weighted MR images of the ADNI-2 cohort. The graph-CNN also achieved the prediction on the conversion of EMCI to AD at 75% and that of LMCI to AD at 92%. The find-tuned graph-CNN further provided a promising CN vs. AD classification accuracy of 89.4% on the ADNI-1 cohort and >90% on the Asian cohort. Our study demonstrated the feasibility to transfer AD/MCI classifiers learned from one population to the other. Notably, incorporating cortical geometry in CNN has the potential to improve classification performance.

Atrophy-centered subtyping of mild cognitive impairment

“…Our choice to utilize a deep learning framework was further motivated by the assumption that complex and non-linear relationships exist between whole brain structure and progression of MCI/AD. Similar to other machine and deep learning models where "transfer learning" (33) is applied, we propose a classification framework which is trained on a domain different than the one being tested (34)(35)(36). However, rather than evaluating the performance of the model against clinically-defined labels (e.g., progressive and stable MCI, or AD convertors and non-convertors), our approach was to re-label data from individuals with MCI based on its proximity to the model's trained labels, that is, AD and CN.…”

Graph of Hippocampal Subfields Grading for Alzheimer’s Disease Prediction

“…This method extracts inter-subject variability from different features (for instance, voxel-based and cortical thickness) and various MRI modalities [20]. Recently, convolutional neural networks (CNN) have been used to capture relationships between anatomical structures volumes [60], and cortical thickness [68]. It is interesting to note that methods based on inter-subject similarities and intra-subject variability have performed similarly for AD prediction.…”

Identifying Early Mild Cognitive Impairment by Multi-Modality MRI-Based Deep Learning

“…In the past few years, several pioneering studies only focused on sMRI data to detect EMCI (Raeper et al, 2018 ; Yue et al, 2018 ; Taheri and Naima, 2019 ; Wee et al, 2019 ), and they have utilized morphological features and demographic factors to perform feature selection after the sMRI image is divided into 45 subcortical regions or 68 cortical regions. Interestingly, Wee et al ( 2019 ) constructed cortical thickness graphs using sMRI data and input them into the popular graph CNN.…”

“…In the past few years, several pioneering studies only focused on sMRI data to detect EMCI (Raeper et al, 2018 ; Yue et al, 2018 ; Taheri and Naima, 2019 ; Wee et al, 2019 ), and they have utilized morphological features and demographic factors to perform feature selection after the sMRI image is divided into 45 subcortical regions or 68 cortical regions. Interestingly, Wee et al ( 2019 ) constructed cortical thickness graphs using sMRI data and input them into the popular graph CNN. sMRI is one of the common neuroimaging tool for disease diagnosis; however, there are many studies illustrating that multi-modality data are more effective than single-modality data for EMCI classification (Amoroso et al, 2018 ; Cheng et al, 2019 ; Forouzannezhad et al, 2020 ; Hao et al, 2020 ; Lei et al, 2020 ), and these studies have shown that different neuroimaging data may provide complementary information that is beneficial to diagnose EMCI.…”