Graph Transformer shows remarkable potential in brain network analysis due to its ability to model graph structures and complex node relationships. Most existing methods typically model the brain as a flat network, ignoring its modular structure, and their attention mechanisms treat all brain region connections equally, ignoring distance-related node connection patterns. However, brain information processing is a hierarchical process that involves local and long-range interactions between brain regions, interactions between regions and sub-functional modules, and interactions among functional modules themselves. This hierarchical interaction mechanism enables the brain to efficiently integrate local computations and global information flow, supporting the execution of complex cognitive functions. To address this issue, we propose BrainHGT, a hierarchical Graph Transformer that simulates the brain's natural information processing from local regions to global communities. Specifically, we design a novel long-short range attention encoder that utilizes parallel pathways to handle dense local interactions and sparse long-range connections, thereby effectively alleviating the over-globalizing issue. To further capture the brain's modular architecture, we designe a prior-guided clustering module that utilizes a cross-attention mechanism to group brain regions into functional communities and leverage neuroanatomical prior to guide the clustering process, thereby improving the biological plausibility and interpretability. Experimental results indicate that our proposed method significantly improves performance of disease identification, and can reliably capture the sub-functional modules of the brain, demonstrating its interpretability.

Deep learning has advanced fMRI analysis, yet it remains unclear which architectural inductive biases are most effective at capturing functional patterns in human brain activity. This issue is particularly important in small-sample settings, as most datasets fall into this category. We compare models with three major inductive biases in deep learning including convolutional neural networks (CNNs), long short-term memory networks (LSTMs), and Transformers for the task of biological sex classification. These models are evaluated within a unified pipeline using parcellated multivariate fMRI time series from the Human Connectome Project (HCP) 7-Tesla cohort, which includes four resting-state runs and four movie-watching task runs. We assess performance on Whole-brain, subcortex, and 12 functional networks. CNNs consistently achieved the highest discrimination for sex classification in both resting-state and movie-watching, while LSTM and Transformer models underperformed. Network-resolved analyses indicated that the Whole-brain, Default Mode, Cingulo-Opercular, Dorsal Attention, and Frontoparietal networks were the most discriminative. These results were largely similar between resting-state and movie-watching. Our findings indicate that, at this dataset size, discriminative information is carried by local spatial patterns and inter-regional dependencies, favoring convolutional inductive bias. Our study provides insights for selecting deep learning architectures for fMRI time series classification.

Structured pruning is one of the representative techniques for compressing large language models (LLMs) to reduce GPU memory consumption and accelerate inference speed. It offers significant practical value in improving the efficiency of LLMs in real-world applications. Current structured pruning methods typically rely on assessment of the importance of the structure units and pruning the units with less importance. Most of them overlooks the interaction and collaboration among artificial neurons that are crucial for the functionalities of LLMs, leading to a disruption in the macro functional architecture of LLMs and consequently a pruning performance degradation. Inspired by the inherent similarities between artificial neural networks and functional neural networks in the human brain, we alleviate this challenge and propose to prune LLMs by identifying and preserving functional networks within LLMs in this study. To achieve this, we treat an LLM as a digital brain and decompose the LLM into functional networks, analogous to identifying functional brain networks in neuroimaging data. Afterwards, an LLM is pruned by preserving the key neurons within these functional networks. Experimental results demonstrate that the proposed method can successfully identify and locate functional networks and key neurons in LLMs, enabling efficient model pruning. Our code is available at https://github.com/

In recent years, the rapid advancement of large language models (LLMs) in natural language processing has sparked significant interest among researchers to understand their mechanisms and functional characteristics. Although existing studies have attempted to explain LLM functionalities by identifying and interpreting specific neurons, these efforts mostly focus on individual neuron contributions, neglecting the fact that human brain functions are realized through intricate interaction networks. Inspired by cognitive neuroscience research on functional brain networks (FBNs), this study introduces a novel approach to investigate whether similar functional networks exist within LLMs. We use methods similar to those in the field of functional neuroimaging analysis to locate and identify functional networks in LLM. Experimental results show that, similar to the human brain, LLMs contain functional networks that frequently recur during operation. Further analysis shows that these functional networks are crucial for LLM performance. Masking key functional networks significantly impairs the model's performance, while retaining just a subset of these networks is adequate to maintain effective operation. This research provides novel insights into the interpretation of LLMs and the lightweighting of LLMs for certain downstream tasks. Code is available at https://github.com/WhatAboutMyStar/LLM_ACTIVATION.

Many existing methods that use functional magnetic resonance imaging (fMRI) classify brain disorders, such as autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD), often overlook the integration of spatial and temporal dependencies of the blood oxygen level-dependent (BOLD) signals, which may lead to inaccurate or imprecise classification results. To solve this problem, we propose a Spatio-Temporal Aggregation eorganization ransformer (STARFormer) that effectively captures both spatial and temporal features of BOLD signals by incorporating three key modules. The region of interest (ROI) spatial structure analysis module uses eigenvector centrality (EC) to reorganize brain regions based on effective connectivity, highlighting critical spatial relationships relevant to the brain disorder. The temporal feature reorganization module systematically segments the time series into equal-dimensional window tokens and captures multiscale features through variable window and cross-window attention. The spatio-temporal feature fusion module employs a parallel transformer architecture with dedicated temporal and spatial branches to extract integrated features. The proposed STARFormer has been rigorously evaluated on two publicly available datasets for the classification of ASD and ADHD. The experimental results confirm that the STARFormer achieves state-of-the-art performance across multiple evaluation metrics, providing a more accurate and reliable tool for the diagnosis of brain disorders and biomedical research. The codes will be available at: https://github.com/NZWANG/STARFormer.

Understanding the neurobiology of opioid use disorder (OUD) using resting-state functional magnetic resonance imaging (rs-fMRI) may help inform treatment strategies to improve patient outcomes. Recent literature suggests temporal characteristics of rs-fMRI blood oxygenation level-dependent (BOLD) signals may offer complementary information to functional connectivity analysis. However, existing studies of OUD analyze BOLD signals using measures computed across all time points. This study, for the first time in the literature, employs data-driven machine learning (ML) modeling of rs-fMRI BOLD features representing multiple time points to identify region(s) of interest that differentiate OUD subjects from healthy controls (HC). Following the triple network model, we obtain rs-fMRI BOLD features from the default mode network (DMN), salience network (SN), and executive control network (ECN) for 31 OUD and 45 HC subjects. Then, we use the Boruta ML algorithm to identify statistically significant BOLD features that differentiate OUD from HC, identifying the DMN as the most salient functional network for OUD. Furthermore, we conduct brain activity mapping, showing heightened neural activity within the DMN for OUD. We perform 5-fold cross-validation classification (OUD vs. HC) experiments to study the discriminative power of functional network features with and without fusing demographic features. The DMN shows the most discriminative power, achieving mean AUC and F1 scores of 80.91% and 73.97%, respectively, when fusing BOLD and demographic features. Follow-up Boruta analysis using BOLD features extracted from the medial prefrontal cortex, posterior cingulate cortex, and left and right temporoparietal junctions reveals significant features for all four functional hubs within the DMN.

Our brains have to do a lot of work when we watch a movie. There are plots to follow, dialogue to interpret, visuals to take in, and more. Now, scientists have created a detailed map of how the human brain functions during the process. Using data from functional magnetic resonance imaging (fMRI), a team from Massachusetts Institute of Technology mapped what different brain networks activate when subjects watch clips from a range of movies. They also saw how different executive networks in the brains are prioritized when watching easy versus difficult scenes.

Studying influential nodes (I-nodes) in brain networks is of great significance in the field of brain imaging. Most existing studies consider brain connectivity hubs as I-nodes. However, this approach relies heavily on prior knowledge from graph theory, which may overlook the intrinsic characteristics of the brain network, especially when its architecture is not fully understood. In contrast, self-supervised deep learning can learn meaningful representations directly from the data. This approach enables the exploration of I-nodes for brain networks, which is also lacking in current studies. This paper proposes a Self-Supervised Graph Reconstruction framework based on Graph-Transformer (SSGR-GT) to identify I-nodes, which has three main characteristics. First, as a self-supervised model, SSGR-GT extracts the importance of brain nodes to the reconstruction. Second, SSGR-GT uses Graph-Transformer, which is well-suited for extracting features from brain graphs, combining both local and global characteristics. Third, multimodal analysis of I-nodes uses graph-based fusion technology, combining functional and structural brain information. The I-nodes we obtained are distributed in critical areas such as the superior frontal lobe, lateral parietal lobe, and lateral occipital lobe, with a total of 56 identified across different experiments. These I-nodes are involved in more brain networks than other regions, have longer fiber connections, and occupy more central positions in structural connectivity. They also exhibit strong connectivity and high node efficiency in both functional and structural networks. Furthermore, there is a significant overlap between the I-nodes and both the structural and functional rich-club. These findings enhance our understanding of the I-nodes within the brain network, and provide new insights for future research in further understanding the brain working mechanisms.

Deep neural network (DNN) models have demonstrated impressive performance in various domains, yet their application in cognitive neuroscience is limited due to their lack of interpretability. In this study we employ two structurally different and complementary DNN-based models, a one-dimensional convolutional neural network (1D-CNN) and a bidirectional long short-term memory network (BiLSTM), to classify individual cognitive states from fMRI BOLD data, with a focus on understanding the cognitive underpinnings of the classification decisions. We show that despite the architectural differences, both models consistently produce a robust relationship between prediction accuracy and individual cognitive performance, such that low performance leads to poor prediction accuracy. To achieve model explainability, we used permutation techniques to calculate feature importance, allowing us to identify the most critical brain regions influencing model predictions. Across models, we found the dominance of visual networks, suggesting that task-driven state differences are primarily encoded in visual processing. Attention and control networks also showed relatively high importance, however, default mode and temporal-parietal networks demonstrated negligible contribution in differentiating cognitive states. Additionally, we observed individual trait-based effects and subtle model-specific differences, such that 1D-CNN showed slightly better overall performance, while BiLSTM showed better sensitivity for individual behavior; these initial findings require further research and robustness testing to be fully established. Our work underscores the importance of explainable DNN models in uncovering the neural mechanisms underlying cognitive state transitions, providing a foundation for future work in this domain.

Task-specific functional MRI (fMRI) images provide excellent modalities for studying the neuronal basis of cognitive processes. We use fMRI data to formulate and solve the problem of deconvolving task-specific aggregate neuronal networks into a set of basic building blocks called canonical networks, to use these networks for functional characterization, and to characterize the physiological basis of these responses by mapping them to regions of the brain. Our results show excellent task-specificity of canonical networks, i.e., the expression of a small number of canonical networks can be used to accurately predict tasks; generalizability across cohorts, i.e., canonical networks are conserved across diverse populations, studies, and acquisition protocols; and that canonical networks have strong anatomical and physiological basis. From a methods perspective, the problem of identifying these canonical networks poses challenges rooted in the high dimensionality, small sample size, acquisition variability, and noise. Our deconvolution technique is based on non-negative matrix factorization (NMF) that identifies canonical networks as factors of a suitably constructed matrix. We demonstrate that our method scales to large datasets, yields stable and accurate factors, and is robust to noise.