Joint Correlational and Discriminative Ensemble Classifier Learning for Dementia Stratification Using Shallow Brain Multiplexes

Abstract

The demented brain wiring undergoes several changes with dementia progression. However, in early dementia stages, particularly early mild cognitive impairment (eMCI), these remain challenging to spot. Hence, developing accurate diagnostic techniques for eMCI identification is critical for early intervention to prevent the onset of Alzheimer's Disease (AD). There is a large body of machine-learning based research developed for classifying different brain states (e.g., AD vs MCI). These works can be fundamentally grouped into two categories. The first uses correlational methods, such as canonical correlation analysis (CCA) and its variants, with the aim to identify most correlated features for diagnosis. The second includes discriminative methods, such as feature selection methods and linear discriminative analysis (LDA) and its variants to identify brain features that distinguish between two brain states. However, existing methods examine these correlational and discriminative brain data independently, which overlooks the complementary information provided by both techniques, which could prove to be useful in the classification of patients with dementia. On the other hand, how early dementia affects cortical brain connections in morphology remains largely unexplored. To address these limitations, we propose a joint correlational and discriminative ensemble learning framework for eMCI diagnosis that leverages a novel brain network representation, derived from the cortex. Specifically, we devise 'the shallow convolutional brain multiplex' (SCBM), which not only measures the similarity in morphology between pairs of brain regions, but also encodes the relationship between two morphological brain networks. Then, we represent each individual brain using a set of SCBMs, which are used to train joint ensemble CCA-SVM and LDA-based classifier. Our framework outperformed several state-of-the-art methods by 3-7% including independent correlational and discriminative methods.

A Review on Image- and Network-based Brain Data Analysis Techniques for Alzheimer's Disease Diagnosis Reveals a Gap in Developing Predictive Methods for Prognosis

“…Fourth, all network-based analysis methods overlooked how dementia states affect the relationship between cortical regions in morphology in both stability, conversion, or reversal MCI evolution scenarios. To fill this gap and noting that several studies [27,46] reported that morphological features of the brain, such as cortical thickness, can be affected in neurological disorders, one can use the recently proposed morphological brain networks for dementia diagnosis [47,48,49,50]. Last, none of these works proposed a technique for predicting the full trajectory of brain shape changes as MCI progresses towards AD, remains stable, or reverses to normal.…”

“…Besides, the absence of network-based predictive models is remarkable (Table 1). As such, the use of advanced network and shape analysis methods, using machine learning, could prove fruitful for both classification [47,48,49,50] and prediction tasks.…”

“…This gap needs to be filled, considering there are studies that indicate morphological features of the brain, such as cortical thickness, can be affected in neurological disorders, including AD [46,27]. As such, the use of advanced network and shape analysis methods, using machine learning, could prove fruitful for both classification [47,48,49,50] and prediction tasks.…”

Graph Neural Networks in Network Neuroscience

“…We collected 30 papers from the Medical Image Computing and Computer Assisted Intervention conference (MICCAI), the Information Processing in Medical Imaging (IPMI) conference, Journal of Neuroscience Methods, IEEE Transactions on Medical Imaging journal, Frontiers in Neuroscience, Neuroimage, Cerebral Cortex journals and bioRxiv. Non-GNN based papers related to brain graphs were excluded from our review such as those proposing CNN-based architectures for a connectomic related task or ML-based works such as (33,34,29). We refer the reader to our GitHub link where all papers cited in our work are available at https://github.com/basiralab/GNNs-in-Network-Neuroscience.…”

Adversarial Brain Multiplex Prediction From a Single Network for High-Order Connectional Gender-Specific Brain Mapping

“…To address the limitation of conventional low-order brain network representation, recent pioneering works [5,6] introduced the concept of a brain multiplex, which in its shallow form, is composed of a source network intra-layer, a target intra-layer, and a convolutional inter-layer capturing the high-order relationship between both intra-layers. Basically, a brain multiplex can be viewed as a tensor stacking two brain networks (also called intra-layers) and one inter-layer that encodes the similarity between these two intra-layers.…”