MTANS: Multi-Scale Mean Teacher Combined Adversarial Network with Shape-Aware Embedding for Semi-Supervised Brain Lesion Segmentation.

The annotation of brain lesion images is a key step in clinical diagnosis and treatment of a wide spectrum of brain diseases. In recent years, segmentation methods based on deep learning have gained unprecedented popularity, leveraging a large amount of data with high-quality voxel-level annotations. However, due to the limited time clinicians can provide for the cumbersome task of manual image segmentation, semi-supervised medical image segmentation methods present an alternative solution as they require only a few labeled samples for training. In this paper, we propose a novel semi-supervised segmentation framework that combines improved mean teacher and adversarial network. Specifically, our framework consists of (i) a student model and a teacher model for segmenting the target and generating the signed distance maps of object surfaces, and (ii) a discriminator network for extracting hierarchical features and distinguishing the signed distance maps of labeled and unlabeled data. Besides, based on two different adversarial learning processes, a multi-scale feature consistency loss derived from the student and teacher models is proposed, and a shape-aware embedding scheme is integrated into our framework. We evaluated the proposed method on the public brain lesion datasets from ISBI 2015, ISLES 2015, and BRATS 2018 for the multiple sclerosis lesion, ischemic stroke lesion, and brain tumor segmentation respectively. Experiments demonstrate that our method can effectively leverage unlabeled data while outperforming the supervised baseline and other state-of-the-art semi-supervised methods trained with the same labeled data. The proposed framework is suitable for joint training of limited labeled data and additional unlabeled data, which is expected to reduce the effort of obtaining annotated images.

RFDCR: Automated brain lesion segmentation using cascaded random forests with dense conditional random fields.

Segmentation of brain lesions from magnetic resonance images (MRI) is an important step for disease diagnosis, surgical planning, radiotherapy and chemotherapy. However, due to noise, motion, and partial volume effects, automated segmentation of lesions from MRI is still a challenging task. In this paper, we propose a two-stage supervised learning framework for automatic brain lesion segmentation. Specifically, in the first stage, intensity-based statistical features, template-based asymmetric features, and GMM-based tissue probability maps are used to train the initial random forest classifier. Next, the dense conditional random field optimizes the probability maps from the initial random forest classifier and derives the whole tumor regions referred as the region of interest (ROI). In the second stage, the optimized probability maps are further intergraded with features from the intensity-based statistical features and template-based asymmetric features to train subsequent random forest, focusing on classifying voxels within the ROI. The output probability maps will be also optimized by the dense conditional random fields, and further used to iteratively train a cascade of random forests. Through hierarchical learning of the cascaded random forests and dense conditional random fields, the multimodal local and global appearance information is integrated with the contextual information, and the output probability maps are improved layer by layer to finally obtain optimal segmentation results. We evaluated the proposed method on the publicly available brain tumor datasets BRATS 2015 & BRATS 2018, as well as the ischemic stroke dataset ISLES 2015. The results have shown that our framework achieves competitive performance compared to the state-of-the-art brain lesion segmentation methods. In addition, contralateral difference and skewness were identified as the important features in the brain tumor and ischemic stroke segmentation tasks, which conforms to the knowledge and experience of medical experts, further reflecting the reliability and interpretability of our framework.

Individual identification and individual variability analysis based on cortical folding features in developing infant singletons and twins.

Studying the early dynamic development of cortical folding with remarkable individual variability is critical for understanding normal early brain development and related neurodevelopmental disorders. This study focuses on the fingerprinting capability and the individual variability of cortical folding during early brain development. Specifically, we aim to explore (a) whether the developing neonatal cortical folding is unique enough to be considered as a "fingerprint" that can reliably identify an individual within a cohort of infants; (b) which cortical regions manifest more individual variability and thus contribute more for infant identification; (c) whether the infant twins can be distinguished by cortical folding. Hence, for the first time, we conduct infant individual identification and individual variability analysis involving twins based on the developing cortical folding features (mean curvature, average convexity, and sulcal depth) in 472 neonates with 1,141 longitudinal MRI scans. Experimental results show that the infant individual identification achieves 100% accuracy when using the neonatal cortical folding features to predict the identities of 1- and 2-year-olds. Besides, we observe high identification capability in the high-order association cortices (i.e., prefrontal, lateral temporal, and inferior parietal regions) and two unimodal cortices (i.e., precentral gyrus and lateral occipital cortex), which largely overlap with the regions encoding remarkable individual variability in cortical folding during the first 2 years. For twins study, we show that even for monozygotic twins with identical genes and similar developmental environments, their cortical folding features are unique enough for accurate individual identification; and in some high-order association cortices, the differences between monozygotic twin pairs are significantly lower than those between dizygotic twins. This study thus provides important insights into individual identification and individual variability based on cortical folding during infancy.

Exploring folding patterns of infant cerebral cortex based on multi-view curvature features: Methods and applications.

The highly convoluted cortical folding of the human brain is intriguingly complex and variable across individuals. Exploring the underlying representative patterns of cortical folding is of great importance for many neuroimaging studies. At term birth, all major cortical folds are established and are minimally affected by the complicated postnatal environments; hence, neonates are the ideal candidates for exploring early postnatal cortical folding patterns, which yet remain largely unexplored. In this paper, we propose a novel method for exploring the representative regional folding patterns of infant brains. Specifically, first, multi-view curvature features are constructed to comprehensively characterize the complex characteristics of cortical folding. Second, for each view of curvature features, a similarity matrix is computed to measure the similarity of cortical folding in a specific region between any pair of subjects. Next, a similarity network fusion method is adopted to nonlinearly and adaptively fuse all the similarity matrices into a single one for retaining both shared and complementary similarity information of the multiple characteristics of cortical folding. Finally, based on the fused similarity matrix and a hierarchical affinity propagation clustering approach, all subjects are automatically grouped into several clusters to obtain the representative folding patterns. To show the applications, we have applied the proposed method to a large-scale dataset with 595 normal neonates and discovered representative folding patterns in several cortical regions, i.e., the superior temporal gyrus (STG), inferior frontal gyrus (IFG), precuneus, and cingulate cortex. Meanwhile, we have revealed sex difference in STG, IFG, and cingulate cortex, as well as hemispheric asymmetries in STG and cingulate cortex in terms of cortical folding patterns. Moreover, we have also validated the proposed method on a public adult dataset, i.e., the Human Connectome Project (HCP), and revealed that certain major cortical folding patterns of adults are largely established at term birth.

Computational neuroanatomy of baby brains: A review.

The first postnatal years are an exceptionally dynamic and critical period of structural, functional and connectivity development of the human brain. The increasing availability of non-invasive infant brain MR images provides unprecedented opportunities for accurate and reliable charting of dynamic early brain developmental trajectories in understanding normative and aberrant growth. However, infant brain MR images typically exhibit reduced tissue contrast (especially around 6 months of age), large within-tissue intensity variations, and regionally-heterogeneous, dynamic changes, in comparison with adult brain MR images. Consequently, the existing computational tools developed typically for adult brains are not suitable for infant brain MR image processing. To address these challenges, many infant-tailored computational methods have been proposed for computational neuroanatomy of infant brains. In this review paper, we provide a comprehensive review of the state-of-the-art computational methods for infant brain MRI processing and analysis, which have advanced our understanding of early postnatal brain development. We also summarize publically available infant-dedicated resources, including MRI datasets, computational tools, grand challenges, and brain atlases. Finally, we discuss the limitations in current research and suggest potential future research directions.

Joint prediction of longitudinal development of cortical surfaces and white matter fibers from neonatal MRI.

The human brain can be modeled as multiple interrelated shapes (or a multishape), each for characterizing one aspect of the brain, such as the cortex and white matter pathways. Predicting the developing multishape is a very challenging task due to the contrasting nature of the developmental trajectories of the constituent shapes: smooth for the cortical surface and non-smooth for white matter tracts due to changes such as bifurcation. We recently addressed this problem and proposed an approach for predicting the multishape developmental spatiotemporal trajectories of infant brains based only on neonatal MRI data using a set of geometric, dynamic, and fiber-to-surface connectivity features. In this paper, we propose two key innovations to further improve the prediction of multishape evolution. First, for a more accurate cortical surface prediction, instead of simply relying on one neonatal atlas to guide the prediction of the multishape, we propose to use multiple neonatal atlases to build a spatially heterogeneous atlas using the multidirectional varifold representation. This individualizes the atlas by locally maximizing its similarity to the testing baseline cortical shape for each cortical region, thereby better representing the baseline testing cortical surface, which founds the multishape prediction process. Second, for temporally consistent fiber prediction, we propose to reliably estimate spatiotemporal connectivity features using low-rank tensor completion, thereby capturing the variability and richness of the temporal development of fibers. Experimental results confirm that the proposed variants significantly improve the prediction performance of our original multishape prediction framework for both cortical surfaces and fiber tracts shape at 3, 6, and 9 months of age. Our pioneering model will pave the way for learning how to predict the evolution of anatomical shapes with abnormal changes. Ultimately, devising accurate shape evolution prediction models that can help quantify and predict the severity of a brain disorder as it progresses will be of great aid in individualized treatment planning.

Can we predict subject-specific dynamic cortical thickness maps during infancy from birth?

Understanding the early dynamic development of the human cerebral cortex remains a challenging problem. Cortical thickness, as one of the most important morphological attributes of the cerebral cortex, is a sensitive indicator for both normal neurodevelopment and neuropsychiatric disorders, but its early postnatal development remains largely unexplored. In this study, we investigate a key question in neurodevelopmental science: can we predict the future dynamic development of cortical thickness map in an individual infant based on its available MRI data at birth? If this is possible, we might be able to better model and understand the early brain development and also early detect abnormal brain development during infancy. To this end, we develop a novel learning-based method, called Dynamically-Assembled Regression Forest (DARF), to predict the development of the cortical thickness map during the first postnatal year, based on neonatal MRI features. We applied our method to 15 healthy infants and predicted their cortical thickness maps at 3, 6, 9, and 12 months of age, with respectively mean absolute errors of 0.209 mm, 0.332 mm, 0.340 mm, and 0.321 mm. Moreover, we found that the prediction precision is region-specific, with high precision in the unimodal cortex and relatively low precision in the high-order association cortex, which may be associated with their differential developmental patterns. Additional experiments also suggest that using more early time points for prediction can further significantly improve the prediction accuracy. Hum Brain Mapp 38:2865-2874, 2017. © 2017 Wiley Periodicals, Inc.

Disrupted topological organization of structural networks revealed by probabilistic diffusion tractography in Tourette syndrome children.

Tourette syndrome (TS) is a childhood-onset neurobehavioral disorder. Although previous TS studies revealed structural abnormalities in distinct corticobasal ganglia circuits, the topological alterations of the whole-brain white matter (WM) structural networks remain poorly understood. Here, we used diffusion MRI probabilistic tractography and graph theoretical analysis to investigate the topological organization of WM networks in 44 drug-naive TS children and 41 age- and gender-matched healthy children. The WM networks were constructed by estimating inter-regional connectivity probability and the topological properties were characterized using graph theory. We found that both TS and control groups showed an efficient small-world organization in WM networks. However, compared to controls, TS children exhibited decreased global and local efficiency, increased shortest path length and small worldness, indicating a disrupted balance between local specialization and global integration in structural networks. Although both TS and control groups showed highly similar hub distributions, TS children exhibited significant decreased nodal efficiency, mainly distributed in the default mode, language, visual, and sensorimotor systems. Furthermore, two separate networks showing significantly decreased connectivity in TS group were identified using network-based statistical (NBS) analysis, primarily composed of the parieto-occipital cortex, precuneus, and paracentral lobule. Importantly, we combined support vector machine and multiple kernel learning frameworks to fuse multiple levels of network topological features for classification of individuals, achieving high accuracy of 86.47%. Together, our study revealed the disrupted topological organization of structural networks related to pathophysiology of TS, and the discriminative topological features for classification are potential quantitative neuroimaging biomarkers for clinical TS diagnosis. Hum Brain Mapp 38:3988-4008, 2017. © 2017 Wiley Periodicals, Inc.

Multidirectional and Topography-based Dynamic-scale Varifold Representations with Application to Matching Developing Cortical Surfaces.

The human cerebral cortex is marked by great complexity as well as substantial dynamic changes during early postnatal development. To obtain a fairly comprehensive picture of its age-induced and/or disorder-related cortical changes, one needs to match cortical surfaces to one another, while maximizing their anatomical alignment. Methods that geodesically shoot surfaces into one another as currents (a distribution of oriented normals) and varifolds (a distribution of non-oriented normals) provide an elegant Riemannian framework for generic surface matching and reliable statistical analysis. However, both conventional current and varifold matching methods have two key limitations. First, they only use the normals of the surface to measure its geometry and guide the warping process, which overlooks the importance of the orientations of the inherently convoluted cortical sulcal and gyral folds. Second, the 'conversion' of a surface into a current or a varifold operates at a fixed scale under which geometric surface details will be neglected, which ignores the dynamic scales of cortical foldings. To overcome these limitations and improve varifold-based cortical surface registration, we propose two different strategies. The first strategy decomposes each cortical surface into its normal and tangent varifold representations, by integrating principal curvature direction field into the varifold matching framework, thus providing rich information of the orientation of cortical folding and better characterization of the complex cortical geometry. The second strategy explores the informative cortical geometric features to perform a dynamic-scale measurement of the cortical surface that depends on the local surface topography (e.g., principal curvature), thereby we introduce the concept of a topography-based dynamic-scale varifold. We tested the proposed varifold variants for registering 12 pairs of dynamically developing cortical surfaces from 0 to 6 months of age. Both variants improved the matching accuracy in terms of closeness to the target surface and the goodness of alignment with regional anatomical boundaries, when compared with three state-of-the-art methods: (1) diffeomorphic spectral matching, (2) conventional current-based surface matching, and (3) conventional varifold-based surface matching.

Combining tract- and atlas-based analysis reveals microstructural abnormalities in early Tourette syndrome children.

Tourette syndrome (TS) is a neurological disorder that causes uncontrolled repetitive motor and vocal tics in children. Examining the neural basis of TS churned out different research studies that advanced our understanding of the brain pathways involved in its development. Particularly, growing evidence points to abnormalities within the fronto-striato-thalamic pathways. In this study, we combined Tract-Based Spatial Statistics (TBSS) and Atlas-based regions of interest (ROI) analysis approach, to investigate the microstructural diffusion changes in both deep and superficial white matter (SWM) in TS children. We then characterized the altered microstructure of white matter in 27 TS children in comparison with 27 age- and gender-matched healthy controls. We found that fractional anisotropy (FA) decreases and radial diffusivity (RD) increases in deep white matter (DWM) tracts in cortico-striato-thalamo-cortical (CSTC) circuit as well as SWM. Furthermore, we found that lower FA values and higher RD values in white matter regions are correlated with more severe tics, but not tics duration. Besides, we also found both axial diffusivity and mean diffusivity increase using Atlas-based ROI analysis. Our work may suggest that microstructural diffusion changes in white matter is not only restricted to the gray matter of CSTC circuit but also affects SWM within the primary motor and somatosensory cortex, commissural and association fibers. Hum Brain Mapp 37:1903-1919, 2016. © 2016 Wiley Periodicals, Inc.

Cortical morphological networks for profiling autism spectrum disorder using tensor component analysis.

Atypical neurodevelopmental disorders such as Autism Spectrum Disorder (ASD) can alter the cortex morphology at different levels: (i) a low-order level where cortical regions are examined individually, (ii) a high-order level where the relationship between two cortical regions is considered, and (iii) a multi-view high-order level where the relationship between regions is examined across multiple brain views. In this study, we propose to use the emerging multi-view cortical morphological network (CMN), which is derived from T1-w magnetic resonance imaging (MRI), to profile autistic and typical brains and pursue new ways of fingerprinting 'cortical morphology' at the intersection of 'network neuroscience'. Each CMN view models the pairwise morphological dissimilarity at the connection level using a specific cortical attribute (e.g., thickness). Specifically, we set out to identify the inherently most representative morphological connectivities shared across different views of the cortex in both autistic and normal control (NC) populations using tensor component analysis. We thus discover the connectional profiles of both populations shared across different CMNs of the left and right hemispheres, respectively. One of the most representative morphological cortical attributes for assessing the abnormal brain structures in patients with ASD is cortical thickness. The most representative morphological connectivities in multi-view CMN population of normal control and ASD subjects, respectively, and in both left and right hemispheres within the temporal, frontal, and insular lobes of individuals with ASD. These representative connectivities are corresponded to specific clinical features observed in individuals with ASD.

Delineating Transdiagnostic Subtypes in Neurodevelopmental Disorders via Contrastive Graph Machine Learning of Brain Connectivity Patterns.

Neurodevelopmental disorders, such as Attention Deficit/Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD), are characterized by comorbidity and heterogeneity. Identifying distinct subtypes within these disorders can illuminate the underlying neurobiological and clinical characteristics, paving the way for more tailored treatments. We adopted a novel transdiagnostic approach across ADHD and ASD, using cutting-edge contrastive graph machine learning to determine subtypes based on brain network connectivity as revealed by resting-state functional magnetic resonance imaging. Our approach identified two generalizable subtypes characterized by robust and distinct functional connectivity patterns, prominently within the frontoparietal control network and the somatomotor network. These subtypes exhibited pronounced differences in major cognitive and behavioural measures. We further demonstrated the generalizability of these subtypes using data collected from independent study sites. Our data-driven approach provides a novel solution for parsing biological heterogeneity in neurodevelopmental disorders.

Predicting the evolution trajectory of population-driven connectional brain templates using recurrent multigraph neural networks.

The mapping of the time-dependent evolution of the human brain connectivity using longitudinal and multimodal neuroimaging datasets provides insights into the development of neurological disorders and the way they alter the brain morphology, structure and function over time. Recently, the connectional brain template (CBT) was introduced as a compact representation integrating a population of brain multigraphs, where two brain regions can have multiple connections, into a single graph. Given a population of brain multigraphs observed at a baseline timepoint t1, we aim to learn how to predict the evolution of the population CBT at follow-up timepoints t>t1. Such model will allow us to foresee the evolution of the connectivity patterns of healthy and disordered individuals at the population level. Here we present recurrent multigraph integrator network (ReMI-Net⋆) to forecast population templates at consecutive timepoints from a given single timepoint. In particular, we unprecedentedly design a graph neural network architecture to model the changes in the brain multigraph and identify the biomarkers that differentiate between the typical and atypical populations. Addressing such issues is of paramount importance in diagnosing neurodegenerative disorders at early stages and promoting new clinical studies based on the pinned-down biomarker brain regions or connectivities. In this paper, we demonstrate the design and use of the ReMI-Net⋆ model, which learns both the multigraph node level and time level dependencies concurrently. Thanks to its novel graph convolutional design and normalization layers, ReMI-Net⋆ predicts well-centered, discriminative, and topologically sound connectional templates over time. Additionally, the results show that our model outperforms all benchmarks and state-of-the-art methods by comparing and discovering the atypical connectivity alterations over time. Our ReMI-Net⋆ code is available on GitHub at https://github.com/basiralab/ReMI-Net-Star.

Federated Brain Graph Evolution Prediction Using Decentralized Connectivity Datasets With Temporally-Varying Acquisitions.

Foreseeing the evolution of brain connectivity between anatomical regions from a baseline observation can propel early disease diagnosis and clinical decision making. Such task becomes challenging when learning from multiple decentralized datasets with missing timepoints (e.g., datasets collected from different hospitals with a varying sequence of acquisitions). Federated learning (FL) is an emerging paradigm that enables collaborative learning among multiple clients (i.e., hospitals) in a fully privacy-preserving fashion. However, to the best of our knowledge, there is no FL work that foresees the time-dependent brain connectivity evolution from a single timepoint-let alone learning from non-iid decentralized longitudinal datasets with varying acquisition timepoints. In this paper, we propose the first FL framework to significantly boost the predictive performance of local hospitals with missing acquisition timepoints while benefiting from other hospitals with available data at those timepoints without sharing data. Specifically, we introduce 4D-FED-GNN+, a novel longitudinal federated GNN framework that works in (i) a uni-mode, where it acts as a graph self-encoder if the next timepoint is locally missing or (ii) in a dual-mode, where it concurrently acts as a graph generator and a self-encoder if the local follow-up data is available. Further, we propose a dual federation strategy, where (i) GNN layer-wise weight aggregation and (ii) pairwise GNN weight exchange between hospitals in a random order. To improve the performance of the poorly-conditioned hospitals (e.g., consecutive missing timepoints, intermediate missing timepoint), we further propose a second variant, namely 4D-FED-GNN++, which federates based on an ordering of the local hospitals computed using their incomplete sequential patterns. Our comprehensive experiments on real longitudinal datasets show that overall 4D-FED-GNN+ and 4D-FED-GNN++ significantly outperform benchmark methods. Our source code is available at https://github.com/basiralab/4D-FedGNN-Plus.

Graph neural network based unsupervised influential sample selection for brain multigraph population fusion.

Graph neural networks (GNNs) have witnessed remarkable proliferation due to the increasing number of applications where data is represented as graphs. GNN-based multigraph population fusion methods for estimating population representative connectional brain templates (CBT) have recently led to improvements, especially in network neuroscience. However, prior studies do not consider how an individual training brain multigraph influences the quality of GNN training for brain multigraph population fusion. To address this issue, we propose two major sample selection methods to quantify the influence of a training brain multigraph on the brain multigraph population fusion task using GNNs, in a fully unsupervised manner: (1) GraphGradIn, in which we use gradients w.r.t GNN weights to trace changes in the centeredness loss of connectional brain template during the training phase; (2) GraphTestIn, in which we exclude a training brain multigraph of interest during the refinement process in the test phase to infer its influence on the CBT centeredness loss. Next, we select the most influential multigraphs to build the training set for brain multigraph population fusion into a CBT. We conducted extensive experiments on brain multigraph datasets to show that using a dataset of influential training samples improves the learned connectional brain template in terms of centeredness, discriminativeness, and topological soundness. Finally, we demonstrate the use of our methods by discovering the connectional fingerprints of healthy and neurologically disordered brain multigraph populations including Alzheimer's disease and Autism spectrum disorder patients. Our source code is available at https://github.com/basiralab/GraphGradIn.

Graph Neural Networks in Network Neuroscience.

Noninvasive medical neuroimaging has yielded many discoveries about the brain connectivity. Several substantial techniques mapping morphological, structural and functional brain connectivities were developed to create a comprehensive road map of neuronal activities in the human brain -namely brain graph. Relying on its non-euclidean data type, graph neural network (GNN) provides a clever way of learning the deep graph structure and it is rapidly becoming the state-of-the-art leading to enhanced performance in various network neuroscience tasks. Here we review current GNN-based methods, highlighting the ways that they have been used in several applications related to brain graphs such as missing brain graph synthesis and disease classification. We conclude by charting a path toward a better application of GNN models in network neuroscience field for neurological disorder diagnosis and population graph integration. The list of papers cited in our work is available at https://github.com/basiralab/GNNs-in-Network-Neuroscience.

Comparative survey of multigraph integration methods for holistic brain connectivity mapping.

One of the greatest scientific challenges in network neuroscience is to create a representative map of a population of heterogeneous brain networks, which acts as a connectional fingerprint. The connectional brain template (CBT), also named network atlas, presents a powerful tool for capturing the most representative and discriminative traits of a given population while preserving its topological patterns. The idea of a CBT is to integrate a population of heterogeneous brain connectivity networks, derived from different neuroimaging modalities or brain views (e.g., structural and functional), into a unified holistic representation. Here we review current state-of-the-art methods designed to estimate well-centered and representative CBT for populations of single-view and multi-view brain networks. We start by reviewing each CBT learning method, then we introduce the evaluation measures to compare CBT representativeness of populations generated by single-view and multigraph integration methods, separately, based on the following criteria: Centeredness, biomarker-reproducibility, node-level similarity, global-level similarity, and distance-based similarity. We demonstrate that the deep graph normalizer (DGN) method significantly outperforms other multi-graph and all single-view integration methods for estimating CBTs using a variety of healthy and disordered datasets in terms of centeredness, reproducibility (i.e., graph-derived biomarkers reproducibility that disentangle the typical from the atypical connectivity variability), and preserving the topological traits at both local and global graph-levels.

Template-based graph registration network for boosting the diagnosis of brain connectivity disorders.

Brain graphs are powerful representations to explore the biological roadmaps of the human brain in its healthy and disordered states. Recently, a few graph neural networks (GNNs) have been designed for brain connectivity synthesis and diagnosis. However, such non-Euclidean deep learning architectures might fail to capture the neural interactions between different brain regions as they are trained without guidance from any prior biological template-i.e., template-free learning. Here we assume that using a population-driven brain connectional template (CBT) that captures well the connectivity patterns fingerprinting a given brain state (e.g., healthy) can better guide the GNN training in its downstream learning task such as classification or regression. To this aim we design a plug-in graph registration network (GRN) that can be coupled with any conventional graph neural network (GNN) so as to boost its learning accuracy and generalizability to unseen samples. Our GRN is a graph generative adversarial network (gGAN), which registers brain graphs to a prior CBT. Next, the registered brain graphs are used to train typical GNN models. Our GRN can be integrated into any GNN working in an end-to-end fashion to boost its prediction accuracy. Our experiments showed that GRN remarkably boosted the prediction accuracy of four conventional GNN models across four neurological datasets.

Quantifying the reproducibility of graph neural networks using multigraph data representation.

Graph neural networks (GNNs) have witnessed an unprecedented proliferation in tackling several problems in computer vision, computer-aided diagnosis and related fields. While prior studies have focused on boosting the model accuracy, quantifying the reproducibility of the most discriminative features identified by GNNs is still an intact problem that yields concerns about their reliability in clinical applications in particular. Specifically, the reproducibility of biological markers across clinical datasets and distribution shifts across classes (e.g., healthy and disordered brains) is of paramount importance in revealing the underpinning mechanisms of diseases as well as propelling the development of personalized treatment. Motivated by these issues, we propose, for the first time, reproducibility-based GNN selection (RG-Select), a framework for GNN reproducibility assessment via the quantification of the most discriminative features (i.e., biomarkers) shared between different models. To ascertain the soundness of our framework, the reproducibility assessment embraces variations of different factors such as training strategies and data perturbations. Despite these challenges, our framework successfully yielded replicable conclusions across different training strategies and various clinical datasets. Our findings could thus pave the way for the development of biomarker trustworthiness and reliability assessment methods for computer-aided diagnosis and prognosis tasks. RG-Select code is available on GitHub at https://github.com/basiralab/RG-Select.

Multigraph classification using learnable integration network with application to gender fingerprinting.

Multigraphs with heterogeneous views present one of the most challenging obstacles to classification tasks due to their complexity. Several works based on feature selection have been recently proposed to disentangle the problem of multigraph heterogeneity. However, such techniques have major drawbacks. First, the bulk of such works lies in the vectorization and the flattening operations, failing to preserve and exploit the rich topological properties of the multigraph. Second, they learn the classification process in a dichotomized manner where the cascaded learning steps are pieced in together independently. Hence, such architectures are inherently agnostic to the cumulative estimation error from step to step. To overcome these drawbacks, we introduce MICNet (multigraph integration and classifier network), the first end-to-end graph neural network based model for multigraph classification. First, we learn a single-view graph representation of a heterogeneous multigraph using a GNN based integration model. The integration process in our model helps tease apart the heterogeneity across the different views of the multigraph by generating a subject-specific graph template while preserving its geometrical and topological properties conserving the node-wise information while reducing the size of the graph (i.e., number of views). Second, we classify each integrated template using a geometric deep learning block which enables us to grasp the salient graph features. We train, in end-to-end fashion, these two blocks using a single objective function to optimize the classification performance. We evaluate our MICNet in gender classification using brain multigraphs derived from different cortical measures. We demonstrate that our MICNet significantly outperformed its variants thereby showing its great potential in multigraph classification.