Graph neural networks for fMRI data often emphasize region-to-region similarities but neglect important factors like hemispheric asymmetry and functional grouping. We introduce a novel framework integrating Functional and Positional Context Embedding with Hemispheric Asymmetry Features to enhance early-stage Parkinson's Disease classification. Our approach refines graph representations of functional connectivity, yielding improved detection of subtle pathological patterns. These findings underscore the value of neuroscience-informed design in brain connectivity analysis.

Predicting disease states from functional brain connectivity is critical for the early diagnosis of severe neurodegenerative diseases such as Alzheimer's Disease and Parkinson's Disease. Existing studies commonly employ Graph Neural Networks (GNNs) to infer clinical diagnoses from node-based brain connectivity matrices generated through node-to-node similarities of regionally averaged fMRI signals. However, recent neuroscience studies found that such node-based connectivity does not accurately capture “functional connections" within the brain. This paper proposes a novel approach to brain network analysis that emphasizes edge functional connectivity (eFC), shifting the focus to inter-edge relationships. Additionally, we introduce a co-embedding technique to integrate edge functional connections effectively. Experimental results on the ADNI and PPMI datasets demonstrate that our method significantly outperforms state-of-the-art GNN methods in classifying functional brain networks.

The recent advances in neuroimaging technology allow us to understand how the human brain is wired in vivo and how functional activity is synchronized across multiple regions. Growing evidence shows that the complexity of the functional connectivity is far beyond the widely used mono-layer network. Indeed, the hierarchical processing information among distinct brain regions and across multiple channels requires using a more advanced multilayer model to understand the synchronization across the brain that underlies functional brain networks. However, the principled approach for characterizing network organization in the context of multilayer topologies is largely unexplored. In this work, we present a novel multi-variate hub identification method that takes both the intra- and inter-layer network topologies into account. Specifically, we put the spotlight on the multilayer graph embeddings that allow us to separate connector hubs (connecting across network modules) with their peripheral nodes. The removal of these hub nodes breaks down the entire multilayer brain network into a set of disconnected communities. We have evaluated our novel multilayer hub identification method in task-based and resting-state functional images. Complimenting ongoing findings using mono-layer brain networks, our multilayer network analysis provides a new understanding of brain network topology that links functional connectivities with brain states and disease progression.

Current diagnostic systems for Alzheimer's disease (AD) rely upon clinical signs and symptoms, despite the fact that the multiplicity of clinical symptoms renders various neuropsychological assessments inadequate to reflect the underlying pathophysiological mechanisms. Since putative neuroimaging biomarkers play a crucial role in understanding the etiology of AD, we sought to stratify the diverse relationships between AD biomarkers and cognitive decline in the aging population and uncover risk factors contributing to the diversities in AD. To do so, we capitalized on a large amount of neuroimaging data from the ADNI study to examine the inflection points along the dynamic relationship between cognitive decline trajectories and whole-brain neuroimaging biomarkers, using a state-of-the-art statistical model of change point detection. Our findings indicated that the temporal relationship between AD biomarkers and cognitive decline may differ depending on the synergistic effect of genetic risk and biological sex. Specifically, tauopathy-PET biomarkers exhibit a more dynamic and age-dependent association with Mini-Mental State Examination scores (p < 0.05), with inflection points at 72, 78, and 83 years old, compared with amyloid-PET and neurodegeneration (cortical thickness from MRI) biomarkers. In the landscape of health disparities in AD, our analysis indicated that biological sex moderates the rate of cognitive decline associated with APOE4 genotype. Meanwhile, we found that higher education levels may moderate the effect of APOE4, acting as a marker of cognitive reserve.

Characterizing a preclinical stage of Alzheimer’s Disease (AD) via single imaging is difficult as its early symptoms are quite subtle. Therefore, many neuroimaging studies are curated with various imaging modalities, e.g., MRI and PET, however, it is often challenging to acquire all of them from all subjects and missing data become inevitable. In this regards, in this paper, we propose a framework that generates unobserved imaging measures for specific subjects using their existing measures, thereby reducing the need for additional examinations. Our framework transfers modality-specific style while preserving AD-specific content. This is done by domain adversarial training that preserves modality-agnostic but AD-specific information, while a generative adversarial network adds an indistinguishable modality-specific style. Our proposed framework is evaluated on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) study and compared with other imputation methods in terms of generated data quality. Small average Cohen’s d < 0.19 between our generated measures and real ones suggests that the synthetic data are practically usable regardless of their modality type.

The purpose of this study is to explore learners' perceptions of peer assessment and evaluate its effectiveness based on the specificity of peer assessment criteria in teamᐨbased collaborative learning. The specificity of peer assessment criteria was examined through two independent variables: the use of detailed criteria and the presentation of evaluation grounds. Learners' perceptions were assessed via quantitative surveys and openᐨended questions focused on the usefulness and fairness of peer assessment, while the effectiveness was determined by the extent of equal participation in team activities, measured by actual team contribution scores. A total of 84 undergraduate students participated in the study. The findings revealed that the use of detailed criteria didn't show the statistical significance, but the provision of evaluation grounds significantly influenced learners' perceptions of the usefulness and fairness of peer assessment, though no interaction effect between the two was found. This suggests that regardless the use of specific criteria, evaluation grounds are provided enhances the perceived usefulness and fairness of peer assessment in evaluating team contributions. Moreover, the study found that group D (which used holistic criteria without providing evaluation grounds) demonstrated relatively unequal team contributions compared to other groups, highlighting the impact of specific peer assessment methods on team contribution within collaborative learning settings. To further clarify the quantitative results, openᐨended responses reflecting learners' perceptions of peer assessment were thoroughly reviewed to interpret the findings. The discussion was approached from the perspective of suggesting specific strategies for utilizing peer assessment to enhance learners' contributions in collaborative learning settings.

Mounting evidence shows that Alzheimer's disease (AD) is characterized by the propagation of tau aggregates throughout the brain in a prion-like manner. Since current pathology imaging technologies only provide a spatial mapping of tau accumulation, computational modeling becomes indispensable in analyzing the spatiotemporal propagation patterns of widespread tau aggregates from the longitudinal data. However, current state-of-the-art works focus on the longitudinal change of focal patterns, lacking a system-level understanding of the tau propagation mechanism that can explain and forecast the cascade of tau accumulation. To address this limitation, we conceptualize that the intercellular spreading of tau pathology forms a dynamic system where each node (brain region) is ubiquitously wired with other nodes while interacting with the build-up of pathological burdens. In this context, we formulate the biological process of tau spreading in a principled potential energy transport model (constrained by brain network topology), which allows us to develop an explainable neural network for uncovering the spatiotemporal dynamics of tau propagation from the longitudinal tau-PET scans. Specifically, we first translate the transport equation into a GNN (graph neural network) backbone, where the spreading flows are essentially driven by the potential energy of tau accumulation at each node. Conventional GNNs employ a l2-norm graph smoothness prior, resulting in nearly equal potential energies across nodes, leading to vanishing flows. Following this clue, we introduce the total variation (TV) into the graph transport model, where the nature of system's Euler-Lagrange equations is to maximize the spreading flow while minimizing the overall potential energy. On top of this min-max optimization scenario, we design a generative adversarial network (GAN-like) to characterize the TV-based spreading flow of tau aggregates, coined TauFlowNet. We evaluate our TauFlowNet on ADNI and OASIS datasets in terms of the prediction accuracy of future tau accumulation and explore the propagation mechanism of tau aggregates as the disease progresses. Compared to the current counterpart methods, our physics-informed deep model yields more accurate and interpretable results, demonstrating great potential in discovering novel neurobiological mechanisms through the lens of machine learning.

Human brain is a complex system composed of many components that interact with each other. A well-designed computational model, usually in the format of partial differential equations (PDEs), is vital to understand the working mechanisms that can explain dynamic and self-organized behaviors. However, the model formulation and parameters are often tuned empirically based on the predefined domain-specific knowledge, which lags behind the emerging paradigm of discovering novel mechanisms from the unprecedented amount of spatiotemporal data. To address this limitation, we sought to link the power of deep neural networks and physics principles of complex systems, which allows us to design explainable deep models for uncovering the mechanistic role of how human brain (the most sophisticated complex system) maintains controllable functions while interacting with external stimulations. In the spirit of optimal control, we present a unified framework to design an explainable deep model that describes the dynamic behaviors of underlying neurobiological processes, allowing us to understand the latent control mechanism at a system level. We have uncovered the pathophysiological mechanism of Alzheimer's disease to the extent of controllability of disease progression, where the dissected system-level understanding enables higher prediction accuracy for disease progression and better explainability for disease etiology than conventional (black box) deep models.

Whole-brain three dimensional (3D) nucleus instance segmentation is vital to quantify regional variations in many neuroscience studies using tissue clearing and microscopy imaging technologies. Due to the large size of whole-brain microscopy images (μm vs. mm resolution in MRI leads voxel volume increasing in 10 9 level), however, it is computationally challenging to train an end-to-end deep model that can recognize nuclei instances in the full 3D volume. Instead, it is common practice to first segment 2D instances in each slice and then assemble them into 3D instances. Moreover, the whole-brain segmentation often comprises a collection of nucleus instance segmentation results in pre-partitioned image stacks, each with a manageable size for applying deep models. Complex arrangements of nuclei in close proximity makes stitching together 2D nucleus segmentations across slices non-trivial, leading to inconsistent segmentation along inter-slice and cross-stack gaps, which undermines the nucleus instance segmentation accuracy of current state-of-the-art deep models. To address this challenge we present a flexible learning-based stitching component which can be either integrated into existing deep models or used as a post-processing step. The backbone is a contextual graph model which is trained to predict the one-to-one correspondence between 2D segmentations along the gap. We have evaluated the performance of our stitching model for 3D nucleus instance segmentation from light-sheet microscopy images of mouse brains. After integrating our stitching model into existing methods, a significant improvement in instance segmentation accuracy is achieved by alleviating the inconsistency issue across discontinuous slices and stacks.

BACKGROUND:Owing to antiretroviral therapy (ART), acquired immune deficiency syndrome (AIDS)-related mortality has significantly decreased. Retaining in care is an essential step for human immunodeficiency virus (HIV) care cascade. This study investigated the incidence of and risk factors for loss to follow-up (LTFU) in Korean people living with HIV (PLWH). MATERIALS AND METHODS:Data from the Korea HIV/AIDS cohort study (including prospective interval cohort and retrospective clinical cohort) were analyzed. LTFU was defined as not visiting the clinic for more than 1 year. Risk factors for LTFU were identified using the Cox regression hazard model. RESULTS:The study enrolled 3,172 adult HIV patients (median age, 36 years; male 92.97%). The median CD4 T cell count at enrollment was 234 cells/mm3 (interquartile range [IQR]: 85 - 373) and the median viral load at enrollment was 56,100 copies/mL (IQR: 15,000 - 203,992). The total follow-up duration was 16,487 person-years, and the overall incidence rate of LTFU was 85/1,000 person-years. In the multivariable Cox regression model, subjects on ART were less likely to have LTFU than subjects not on ART (hazard ratio [HR] = 0.253, 95% confidence interval [CI]: 0.220 - 0.291, P <0.0001). Among PLWH on ART, female sex (HR = 0.752, 95% CI: 0.582 - 0.971, P = 0.0291) and older age (>50: HR = 0.732, 95% CI: 0.602 - 0.890; 41 - 50: HR = 0.634, 95% CI: 0.530 - 0.750; 31 - 40: HR = 0.724, 95% CI: 0.618 - 0.847; ≤30: reference, P <0.0001) were associated with high rate of retention in care. The viral load at ART initiation ≥1,000,001 (HR = 1.545, 95% CI: 1.126 - 2.121, ≤10,000: reference) was associated with a higher rate of LTFU. CONCLUSION:Young and male PLWH may have a higher rate of LTFU, and an increased rate of LTFU may induce virologic failure.

Stacking excessive layers in DNN results in highly underdetermined system when training samples are limited, which is very common in medical applications. In this regard, we present a framework capable of deriving an efficient high-dimensional space with reasonable increase in model size. This is done by utilizing a transform (i.e., convolution) that leverages scale-space theory with covariance structure. The overall model trains on this transform together with a downstream classifier (i.e., Fully Connected layer) to capture the optimal multi-scale representation of the original data which corresponds to task-specific components in a dual space. Experiments on neuroimaging measures from Alzheimer’s Disease Neuroimaging Initiative (ADNI) study show that our model performs better and converges faster than conventional models even when the model size is significantly reduced. The trained model is made interpretable using gradient information over the multi-scale transform to delineate personalized AD-specific regions in the brain.

Functional neuroimaging technology offers a new window to snapshot the transient neural activity in-vivo. Although tremendous efforts have been made to characterize spontaneous functional fluctuations, little attention has been paid to the functional mechanisms of neural interactions. Inspired by the notion of holography, we propose an explainable machine learning approach to establishing a novel underpinning of self-organized cross-frequency coupling (CFC) through the lens of brain wave interference on top of the network topology. Specifically, we conceptualize that the interaction between ubiquitous neural activities and a collection of reference harmonic wavelets forms a region-adaptive interference pattern that captures cross-frequency coupling of remarkable neuronal oscillations. In this regard, assembling whole-brain CFC patterns under the constraint of brain wiring mechanisms constitutes a HoloBrain mapping that records a wide spectrum of spontaneous neural activities. Since each local interference pattern is a symmetric and positive-definite (SPD) matrix, we tailor a deep model of HoloBrain (coined DeepHoloBrain) to infer the latent feature representations on the Riemannian manifold of SPD matrices for predicting brain states and recognizing disease connectomes. We have applied DeepHoloBrain to the Human Connectome Project and several dementia-related datasets. Compared with current state-of-the-art deep models, our DeepHoloBrain not only improves the recognition/prediction accuracy but also sheds new light on understanding the neurobiological mechanisms of brain function and cognition.

Amyloid-beta (A β ) deposition and tau neurofibrillary tangles (tau) are important hallmarks of Alzheimer’s disease (AD). Although converging evidence shows that the interaction between A β and tau is the gateway to understanding the etiology of AD, these two AD hallmarks are often treated as independent variables in the current state-of-the-art early diagnostic model for AD, which might be partially responsible for the issue of lacking explainability. Inspired by recent progress in systems biology, we formulate the evolving biological process of A β cascade and tau propagation into a closed-loop feedback system where the system dynamics are constrained by region-to-region white matter fiber tracts in the brain. On top of this, we conceptualize that A β -tau interaction, following the principle of optimal control, underlines the pathophysiological mechanism of AD. In this context, we propose a deep reaction-diffusion model that leverages the capital of deep learning and insights into systems biology, which allows us to (1) enhance the prediction accuracy of developing AD and (2) uncover the latent control mechanism of A β -tau interactions. We have evaluated our novel explainable deep model on the neuroimaging data in Alzheimer’s Disease Neuroimaging Initiative (ADNI), where we achieve not only a higher prediction accuracy for disease progression but also a better understanding of disease etiology than conventional (“black-box”) deep models.

Background In recent decades, real-world evidence (RWE) in oncology has rapidly gained traction for its potential to answer clinical questions that cannot be directly addressed by randomized clinical trials. Integrating real-world data (RWD) into clinical research promises to contribute to more sustainable research designs, including extension, augmentation, enrichment, and pragmatic designs. Nevertheless, clinical research using RWD is still limited because of concerns regarding the shortage of best practices for extracting, harmonizing, and analyzing RWD. In particular, pragmatic screening methods to determine whether the content of a data source is sufficient to answer the research questions before conducting the research with RWD have not yet been established. Objective We examined the PAR (Preliminary Attainability Assessment of Real-World Data) framework and assessed its utility in breast cancer brain metastasis (BCBM), which has an unmet medical need for data attainability screening at the preliminary step of observational studies that use RWD. Methods The PAR framework was proposed to assess data attainability from a particular data source during the early research process. The PAR framework has four sequential stages, starting with clinical question clarification: (1) operational definition of variables, (2) data matching (structural/semantic), (3) data screening and extraction, and (4) data attainability diagramming. We identified 5 clinical questions to be used for PAR framework evaluation through interviews and validated them with a survey of breast cancer experts. We used the Samsung Medical Center Breast Cancer Registry, a hospital-based real-time registry implemented in March 2021, leveraging the institution’s anonymized and deidentified clinical data warehouse platform. The number of breast cancer patients in the registry was 45,129; it covered the period from June 1995 to December 2021. The registry consists of 24 base data marts that represent disease-specific breast cancer characteristics and care pathways. The outcomes included screening results of the clinical questions via the PAR framework and a procedural diagram of data attainability for each research question. Results Data attainability was tested for study feasibility according to the PAR framework with 5 clinical questions for BCBM. We obtained data sets that were sufficient to conduct studies with 4 of 5 clinical questions. The research questions stratified into 3 types when we developed data fields for clearly defined research variables. In the first, only 1 question could be answered using direct data variables. In the second, the other 3 questions required surrogate definitions that combined data variables. In the third, the question turned out to be not feasible for conducting further analysis. Conclusions The adoption of the PAR framework was associated with more efficient preliminary clinical research using RWD from BCBM. Furthermore, this framework helped accelerate RWE generation through clinical research by enhancing transparency and reproducibility and lowering the entry barrier for clinical researchers.

A brain network, viewed as a graph wiring different regions of interest (ROIs) in the brain, has been widely used to investigate brain dysfunction with various graph neural networks (GNNs). However, existing GNNs are built upon graph convolution that transforms measurements on the nodes, where ROI-wise features are not always guaranteed for brain networks. Therefore, the majority of existing graph analysis methods that rely on node features are inapplicable for network analysis unless a proxy such as node degree is provided. Moreover, the complex neurological interactions across different brain regions cannot be directly expressed in a simple node-to-node (i.e., 0-simplex) representation. In this paper, we propose a novel method, Hodge-Graph Neural Network (Hodge-GNN), that allows the GNN to directly derive desirable representations of graph edges and capture complex edge-wise topological features spatially via the Hodge Laplacian. Specifically, representing a graph as a simplicial complex holds a significant advantage over conventional methods that extract higher-order connectivity of a graph through hierarchical convolution in the spatial domain or graph transformation. The superiority of our method is validated in the Alzheimer's Disease Neuroimaging Initiative (ADNI) study, in comparison to benchmarking GNNs as well as state-of-the-art graph classification models.

A confluence of neuroscience and clinical evidence suggests that the disruption of structural connectivity (SC) and functional connectivity (FC) in the brain is an early sign of neurodegenerative diseases years before any clinical signs of the disease progression. Since the changes in SC-FC coupling may provide a potential putative biomarker that detects subtle brain network dysfunction more sensitively than does a single modality, tremendous efforts have been made to understand the relationship between SC and FC from the perspective of connectivity, sub-networks, and network topology. However, the methodology design of current analytic methods lacks the in-depth neuroscience underpinning of to what extent the altered SC-FC coupling mechanisms underline the cognitive decline. To address this challenge, we put the spotlight on a neural oscillation model that characterizes the system behavior of a set of (functional) neural oscillators coupled via (structural) nerve fibers throughout the brain. On top of this, we present a physics-guided graph neural network to understand the synchronization mechanism of system dynamics that is capable of predicting self-organized functional fluctuations. By doing so, we generate a novel SC-FC coupling biomarker that allows us to recognize the early sign of neurodegeneration through the lens of an altered SC-FC relationship. We have evaluated the statistical power and clinical value of new SC-FC biomarker in the early diagnosis of Alzheimer’s disease using the ADNI dataset. Compared to conventional SC-FC coupling methods, our physics-guided deep model not only yields higher prediction accuracy but also reveals the mechanistic role of SC-FC coupling alterations in disease progression.

Background:The AT[N] research framework focuses on three major biomarkers in Alzheimer's disease (AD): amyloid-β deposition (A), pathologic tau (T), and neurodegeneration [N]. Objective:We hypothesize that the diverse mechanisms such as A⟶T and A⟶[N] pathways from one brain region to others, may underlie the wide variation in clinical symptoms. We aim to uncover the causal-like effect of regional AT[N] biomarkers on cognitive decline as well as the interaction with non-modifiable risk factors such as age and APOE4. Methods:We apply multi-variate statistical inference to uncover all possible mechanistic spreading pathways through which the aggregation of an upstream biomarker (e.g., increased amyloid level) in a particular brain region indirectly impacts cognitive decline, via the cascade build-up of a downstream biomarker (e.g., reduced metabolism level) in another brain region. Furthermore, we investigate the survival time for each identified region-to-region pathological pathway toward the AD onset. Results:We have identified a collection of critical brain regions on which the amyloid burdens exert an indirect effect on the decline in memory and executive function (EF) domain, being mediated by the reduction of metabolism level at other brain regions. APOE4 status has been found not only involved in many A⟶N mechanistic pathways but also significantly contributes to the risk of developing AD. Conclusion:Our major findings include 1) the region-to-region A⟶N⟶MEM and A⟶N⟶MEM pathways exhibit distinct spatial patterns; 2) APOE4 is significantly associated with both direct and indirect effects on the cognitive decline while sex difference has not been identified in the mediation analysis.

Alzheimer’s disease (AD) is characterized by the propagation of tau aggregates throughout the brain in a prion-like manner. Tremendous efforts have been made to analyze the spatiotemporal propagation patterns of widespread tau aggregates. However, current works focus on the change of focal patterns in lieu of a system-level understanding of the tau propagation mechanism that can explain and forecast the cascade of tau accumulation. To fill this gap, we conceptualize that the intercellular spreading of tau pathology forms a dynamic system where brain region is ubiquitously wired with other nodes while interacting with the build-up of pathological burdens. In this context, we formulate the biological process of tau spreading in a principled potential energy transport model (constrained by brain network topology), which allows us to develop an explainable neural network for uncovering the spatiotemporal dynamics of tau propagation from the longitudinal tau-PET images. We first translate the transport equation into a backbone of graph neural network (GNN), where the spreading flows are essentially driven by the potential energy of tau accumulation at each node. Further, we introduce the total variation (TV) into the graph transport model to prevent the flow vanishing caused by the 𝓁_2 -norm regularization, where the nature of system’s Euler-Lagrange equations is to maximize the spreading flow while minimizing the overall potential energy. On top of this min-max optimization scenario, we design a generative adversarial network (GAN) to depict the TV-based spreading flow of tau aggregates, coined TauFlowNet. We evaluate TauFlowNet on ADNI dataset in terms of the prediction accuracy of future tau accumulation and explore the propagation mechanism of tau aggregates as the disease progresses. Compared to current methods, our physics-informed method yields more accurate and interpretable results, demonstrating great potential in discovering novel neurobiological mechanisms.

Graph neural networks (GNNs) are widely used in domains like social networks and biological systems. However, the locality assumption of GNNs, which limits information exchange to neighboring nodes, hampers their ability to capture long-range dependencies and global patterns in graphs. To address this, we propose a new inductive bias based on variational analysis, drawing inspiration from the Brachistochrone problem. Our framework establishes a mapping between discrete GNN models and continuous diffusion functionals. This enables the design of application-specific objective functions in the continuous domain and the construction of discrete deep models with mathematical guarantees. To tackle over-smoothing in GNNs, we analyze the existing layer-by-layer graph embedding models and identify that they are equivalent to l2-norm integral functionals of graph gradients, which cause over-smoothing. Similar to edge-preserving filters in image denoising, we introduce total variation (TV) to align the graph diffusion pattern with global community topologies. Additionally, we devise a selective mechanism to address the trade-off between model depth and over-smoothing, which can be easily integrated into existing GNNs. Furthermore, we propose a novel generative adversarial network (GAN) that predicts spreading flows in graphs through a neural transport equation. To mitigate vanishing flows, we customize the objective function to minimize transportation within each community while maximizing inter-community flows. Our GNN models achieve state-of-the-art (SOTA) performance on popular graph learning benchmarks such as Cora, Citeseer, and Pubmed.

Transcranial direct current stimulation (tDCS) is a safe and well-tolerated noninvasive method for stimulating the brain that is rapidly developing into a treatment method for various neurological and psychiatric conditions. In particular, there is growing evidence of a therapeutic role for tDCS in ameliorating or delaying the cognitive decline in Alzheimer's disease (AD). We provide a brief overview of the current development and application status of tDCS as a nonpharmacological therapeutic method for AD and mild cognitive impairment (MCI), summarize the levels of evidence, and identify the improvements needed for clinical applications. We also suggest future directions for large-scale controlled clinical trials of tDCS in AD and MCI, and emphasize the necessity of identifying the mechanistic targets to facilitate clinical applications.