Learning lifespan brain anatomical correspondence via cortical developmental continuity transfer.

Identifying anatomical correspondences in the human brain throughout the lifespan is an essential prerequisite for studying brain development and aging. But given the tremendous individual variability in cortical folding patterns, the heterogeneity of different neurodevelopmental stages, and the scarce of neuroimaging data, it is difficult to infer reliable lifespan anatomical correspondence at finer scales. To solve this problem, in this work, we take the advantage of the developmental continuity of the cerebral cortex and propose a novel transfer learning strategy: the model is trained from scratch using the age group with the largest sample size, and then is transferred and adapted to the other groups following the cortical developmental trajectory. A novel loss function is designed to ensure that during the transfer process the common patterns will be extracted and preserved, while the group-specific new patterns will be captured. The proposed framework was evaluated using multiple datasets covering four lifespan age groups with 1,000+ brains (from 34 gestational weeks to young adult). Our experimental results show that: 1) the proposed transfer strategy can dramatically improve the model performance on populations (e.g., early neurodevelopment) with very limited number of training samples; and 2) with the transfer learning we are able to robustly infer the complicated many-to-many anatomical correspondences among different brains at different neurodevelopmental stages. (Code will be released soon: https://github.com/qidianzl/CDC-transfer).

Role of Data-driven Regional Growth Model in Shaping Brain Folding Patterns.

The surface morphology of the developing mammalian brain is crucial for understanding brain function and dysfunction. Computational modeling offers valuable insights into the underlying mechanisms for early brain folding. Recent findings indicate significant regional variations in brain tissue growth, while the role of these variations in cortical development remains unclear. In this study, we unprecedently explored how regional cortical growth affects brain folding patterns using computational simulation. We first developed growth models for typical cortical regions using machine learning (ML)-assisted symbolic regression, based on longitudinal real surface expansion and cortical thickness data from prenatal and infant brains derived from over 1,000 MRI scans of 735 pediatric subjects with ages ranging from 29 post-menstrual weeks to 24 months. These models were subsequently integrated into computational software to simulate cortical development with anatomically realistic geometric models. We comprehensively quantified the resulting folding patterns using multiple metrics such as mean curvature, sulcal depth, and gyrification index. Our results demonstrate that regional growth models generate complex brain folding patterns that more closely match actual brains structures, both quantitatively and qualitatively, compared to conventional uniform growth models. Growth magnitude plays a dominant role in shaping folding patterns, while growth trajectory has a minor influence. Moreover, multi-region models better capture the intricacies of brain folding than single-region models. Our results underscore the necessity and importance of incorporating regional growth heterogeneity into brain folding simulations, which could enhance early diagnosis and treatment of cortical malformations and neurodevelopmental disorders such as cerebral palsy and autism.

Mask-Guided Vision Transformer for Few-Shot Learning.

Learning with little data is challenging but often inevitable in various application scenarios where the labeled data are limited and costly. Recently, few-shot learning (FSL) gained increasing attention because of its generalizability of prior knowledge to new tasks that contain only a few samples. However, for data-intensive models such as vision transformer (ViT), current fine-tuning-based FSL approaches are inefficient in knowledge generalization and, thus, degenerate the downstream task performances. In this article, we propose a novel mask-guided ViT (MG-ViT) to achieve an effective and efficient FSL on the ViT model. The key idea is to apply a mask on image patches to screen out the task-irrelevant ones and to guide the ViT focusing on task-relevant and discriminative patches during FSL. Particularly, MG-ViT only introduces an additional mask operation and a residual connection, enabling the inheritance of parameters from pretrained ViT without any other cost. To optimally select representative few-shot samples, we also include an active learning-based sample selection method to further improve the generalizability of MG-ViT-based FSL. We evaluate the proposed MG-ViT on classification, object detection, and segmentation tasks using gradient-weighted class activation mapping (Grad-CAM) to generate masks. The experimental results show that the MG-ViT model significantly improves the performance and efficiency compared with general fine-tuning-based ViT and ResNet models, providing novel insights and a concrete approach toward generalizing data-intensive and large-scale deep learning models for FSL.

Advancing microbial production through artificial intelligence-aided biology.

Microbial cell factories (MCFs) have been leveraged to construct sustainable platforms for value-added compound production. To optimize metabolism and reach optimal productivity, synthetic biology has developed various genetic devices to engineer microbial systems by gene editing, high-throughput protein engineering, and dynamic regulation. However, current synthetic biology methodologies still rely heavily on manual design, laborious testing, and exhaustive analysis. The emerging interdisciplinary field of artificial intelligence (AI) and biology has become pivotal in addressing the remaining challenges. AI-aided microbial production harnesses the power of processing, learning, and predicting vast amounts of biological data within seconds, providing outputs with high probability. With well-trained AI models, the conventional Design-Build-Test (DBT) cycle has been transformed into a multidimensional Design-Build-Test-Learn-Predict (DBTLP) workflow, leading to significantly improved operational efficiency and reduced labor consumption. Here, we comprehensively review the main components and recent advances in AI-aided microbial production, focusing on genome annotation, AI-aided protein engineering, artificial functional protein design, and AI-enabled pathway prediction. Finally, we discuss the challenges of integrating novel AI techniques into biology and propose the potential of large language models (LLMs) in advancing microbial production.

Hierarchical functional differences between gyri and sulci at different scales.

Gyri and sulci are 2 fundamental cortical folding patterns of the human brain. Recent studies have suggested that gyri and sulci may play different functional roles given their structural and functional heterogeneity. However, our understanding of the functional differences between gyri and sulci remains limited due to several factors. Firstly, previous studies have typically focused on either the spatial or temporal domain, neglecting the inherently spatiotemporal nature of brain functions. Secondly, analyses have often been restricted to either local or global scales, leaving the question of hierarchical functional differences unresolved. Lastly, there has been a lack of appropriate analytical tools for interpreting the hierarchical spatiotemporal features that could provide insights into these differences. To overcome these limitations, in this paper, we proposed a novel hierarchical interpretable autoencoder (HIAE) to explore the hierarchical functional difference between gyri and sulci. Central to our approach is its capability to extract hierarchical features via a deep convolutional autoencoder and then to map these features into an embedding vector using a carefully designed feature interpreter. This process transforms the features into interpretable spatiotemporal patterns, which are pivotal in investigating the functional disparities between gyri and sulci. We evaluate the proposed framework on Human Connectome Project task functional magnetic resonance imaging dataset. The experiments demonstrate that the HIAE model can effectively extract and interpret hierarchical spatiotemporal features that are neuroscientifically meaningful. The analyses based on the interpreted features suggest that gyri are more globally activated, whereas sulci are more locally activated, demonstrating a distinct transition in activation patterns as the scale shifts from local to global. Overall, our study provides novel insights into the brain's anatomy-function relationship.

Exploring Alzheimer's disease: a comprehensive brain connectome-based survey.

Dementia is an escalating global health challenge, with Alzheimer's disease (AD) at its forefront. Substantial evidence highlights the accumulation of AD-related pathological proteins in specific brain regions and their subsequent dissemination throughout the broader area along the brain network, leading to disruptions in both individual brain regions and their interconnections. Although a comprehensive understanding of the neurodegeneration-brain network link is lacking, it is undeniable that brain networks play a pivotal role in the development and progression of AD. To thoroughly elucidate the intricate network of elements and connections constituting the human brain, the concept of the brain connectome was introduced. Research based on the connectome holds immense potential for revealing the mechanisms underlying disease development, and it has become a prominent topic that has attracted the attention of numerous researchers. In this review, we aim to systematically summarize studies on brain networks within the context of AD, critically analyze the strengths and weaknesses of existing methodologies, and offer novel perspectives and insights, intending to serve as inspiration for future research.

Mapping dynamic spatial patterns of brain function with spatial-wise attention.

Objective: Using functional magnetic resonance imaging (fMRI) and deep learning to discover the spatial pattern of brain function, or functional brain networks (FBNs) has been attracted many reseachers. Most existing works focus on static FBNs or dynamic functional connectivity among fixed spatial network nodes, but ignore the potential dynamic/time-varying characteristics of the spatial networks themselves. And most of works based on the assumption of linearity and independence, that oversimplify the relationship between blood-oxygen level dependence signal changes and the heterogeneity of neuronal activity within voxels.Approach: To overcome these problems, we proposed a novel spatial-wise attention (SA) based method called Spatial and Channel-wise Attention Autoencoder (SCAAE) to discover the dynamic FBNs without the assumptions of linearity or independence. The core idea of SCAAE is to apply the SA to generate FBNs directly, relying solely on the spatial information present in fMRI volumes. Specifically, we trained the SCAAE in a self-supervised manner, using the autoencoder to guide the SA to focus on the activation regions. Experimental results show that the SA can generate multiple meaningful FBNs at each fMRI time point, which spatial similarity are close to the FBNs derived by known classical methods, such as independent component analysis.Main results: To validate the generalization of the method, we evaluate the approach on HCP-rest, HCP-task and ADHD-200 dataset. The results demonstrate that SA mechanism can be used to discover time-varying FBNs, and the identified dynamic FBNs over time clearly show the process of time-varying spatial patterns fading in and out.Significance: Thus we provide a novel method to understand human brain better. Code is available athttps://github.com/WhatAboutMyStar/SCAAE.

Technical note: Generalizable and promptable artificial intelligence model to augment clinical delineation in radiation oncology.

BACKGROUND: Efficient and accurate delineation of organs at risk (OARs) is a critical procedure for treatment planning and dose evaluation. Deep learning-based auto-segmentation of OARs has shown promising results and is increasingly being used in radiation therapy. However, existing deep learning-based auto-segmentation approaches face two challenges in clinical practice: generalizability and human-AI interaction. A generalizable and promptable auto-segmentation model, which segments OARs of multiple disease sites simultaneously and supports on-the-fly human-AI interaction, can significantly enhance the efficiency of radiation therapy treatment planning. PURPOSE: Meta's segment anything model (SAM) was proposed as a generalizable and promptable model for next-generation natural image segmentation. We further evaluated the performance of SAM in radiotherapy segmentation. METHODS: Computed tomography (CT) images of clinical cases from four disease sites at our institute were collected: prostate, lung, gastrointestinal, and head & neck. For each case, we selected the OARs important in radiotherapy treatment planning. We then compared both the Dice coefficients and Jaccard indices derived from three distinct methods: manual delineation (ground truth), automatic segmentation using SAM's 'segment anything' mode, and automatic segmentation using SAM's 'box prompt' mode that implements manual interaction via live prompts during segmentation. RESULTS: Our results indicate that SAM's segment anything mode can achieve clinically acceptable segmentation results in most OARs with Dice scores higher than 0.7. SAM's box prompt mode further improves Dice scores by 0.1∼0.5. Similar results were observed for Jaccard indices. The results show that SAM performs better for prostate and lung, but worse for gastrointestinal and head & neck. When considering the size of organs and the distinctiveness of their boundaries, SAM shows better performance for large organs with distinct boundaries, such as lung and liver, and worse for smaller organs with less distinct boundaries, like parotid and cochlea. CONCLUSIONS: Our results demonstrate SAM's robust generalizability with consistent accuracy in automatic segmentation for radiotherapy. Furthermore, the advanced box-prompt method enables the users to augment auto-segmentation interactively and dynamically, leading to patient-specific auto-segmentation in radiation therapy. SAM's generalizability across different disease sites and different modalities makes it feasible to develop a generic auto-segmentation model in radiotherapy.

A Unified and Biologically Plausible Relational Graph Representation of Vision Transformers.

Vision transformer (ViT) and its variants have achieved remarkable success in various tasks. The key characteristic of these ViT models is to adopt different aggregation strategies of spatial patch information within the artificial neural networks (ANNs). However, there is still a key lack of unified representation of different ViT architectures for systematic understanding and assessment of model representation performance. Moreover, how those well-performing ViT ANNs are similar to real biological neural networks (BNNs) is largely unexplored. To answer these fundamental questions, we, for the first time, propose a unified and biologically plausible relational graph representation of ViT models. Specifically, the proposed relational graph representation consists of two key subgraphs: an aggregation graph and an affine graph. The former considers ViT tokens as nodes and describes their spatial interaction, while the latter regards network channels as nodes and reflects the information communication between channels. Using this unified relational graph representation, we found that: 1) model performance was closely related to graph measures; 2) the proposed relational graph representation of ViT has high similarity with real BNNs; and 3) there was a further improvement in model performance when training with a superior model to constrain the aggregation graph.

ARHGAP6 Suppresses Breast Cancer Tumor Growth by Promoting Ferroptosis via RhoA-ROCK1-p38 MAPK Signaling.

BACKGROUND: Ferroptosis, a distinct iron-dependent form of regulated cell death, is induced by severe lipid peroxidation due to reactive oxygen species (ROS) generation. Breast cancer patient survival is correlated with the tumor-suppressing properties of Rho guanosine triphosphatase hydrolase enzyme (GTPase)-activating protein 6 (ARHGAP6). This study investigates the impact and mechanisms of ARHGAP6 on ferroptosis in breast cancer. METHODS: Using quantitative RT-PCR, Western blotting, and immunofluorescence staining, ARHGAP6 expression was detected in a gene expression dataset, cancer tissue samples, and cells. ARHGAP6 was overexpressed or silenced in breast cancer cell lines. Cell proliferation was measured using 5-ethynyl-2-deoxyuridine (EdU) assay, and cell death rate was determined using LDH cytotoxicity assay. As indicators of ferroptosis, Fe2+ ion content, lipid ROS, glutathione peroxidase 4 (GPX4), ChaC glutathione specific gamma-glutamylcyclotransferase 1 (CHAC1), prostaglandin-endoperoxide synthase 2 (PTGS2), solute carrier family 7 member 11 (SLC7A11), and acyl-CoA synthetase long chain family member 4 (ACSL4) levels were evaluated. RESULTS: ARHGAP6 was obviously downregulated in cancer tissues and cells. ARHGAP6 overexpression decreased cell proliferation, elevated cell death and lipid ROS, decreased GPX4 and SLC7A11, increased PTGS2, ACSL4, and CHAC1, and inhibited RhoA/ROCK1 and p38 MAPK signaling in cancer cells. ARHGAP6 knockdown exerted opposite effects to those of ARHGAP6 overexpression. p38 signaling suppression reversed the effect of ARHGAP6 knockdown on ferroptosis, while RhoA/ROCK1 signaling inhibition compromised the effect of ARHGAP6 on p38 MAPK signaling. In mice models, ARHGAP6 together with the ferroptosis inducer RSL3 cooperatively enhanced ferroptosis and inhibited tumor growth of cancer cells. ARHGAP6 mRNA level was positively correlated with that of ferroptosis indicators in tumor tissues. CONCLUSIONS: This study revealed that ARHGAP6 inhibited tumor growth of breast cancer by inducing ferroptosis via RhoA/ROCK1/p38 MAPK signaling. Integrating ARHGAP6 with ferroptosis-inducing agents may be a promising therapeutic strategy for breast cancer treatment.

Causalized convergent cross-mapping and its approximate equivalence with directed information in causality analysis.

Convergent cross-mapping (CCM) has attracted increased attention recently due to its capability to detect causality in nonseparable systems under deterministic settings, which may not be covered by the traditional Granger causality. From an information-theoretic perspective, causality is often characterized as the directed information (DI) flowing from one side to the other. As information is essentially nondeterministic, a natural question is: does CCM measure DI flow? Here, we first causalize CCM so that it aligns with the presumption in causality analysis-the future values of one process cannot influence the past of the other, and then establish and validate the approximate equivalence of causalized CCM (cCCM) and DI under Gaussian variables through both theoretical derivations and fMRI-based brain network causality analysis. Our simulation result indicates that, in general, cCCM tends to be more robust than DI in causality detection. The underlying argument is that DI relies heavily on probability estimation, which is sensitive to data size as well as digitization procedures; cCCM, on the other hand, gets around this problem through geometric cross-mapping between the manifolds involved. Overall, our analysis demonstrates that cross-mapping provides an alternative way to evaluate DI and is potentially an effective technique for identifying both linear and nonlinear causal coupling in brain neural networks and other settings, either random or deterministic, or both.

Disease2Vec: Encoding Alzheimer's progression via disease embedding tree.

For decades, a variety of predictive approaches have been proposed and evaluated in terms of their prediction capability for Alzheimer's Disease (AD) and its precursor - mild cognitive impairment (MCI). Most of them focused on prediction or identification of statistical differences among different clinical groups or phases, especially in the context of binary or multi-class classification. The continuous nature of AD development and transition states between successive AD related stages have been typically overlooked. Though a few progression models of AD have been studied recently, they were mainly designed to determine and compare the order of specific biomarkers. How to effectively predict the individual patient's status within a wide spectrum of continuous AD progression has been largely understudied. In this work, we developed a novel learning-based embedding framework to encode the intrinsic relations among AD related clinical stages by a set of meaningful embedding vectors in the latent space (Disease2Vec). We named this process as disease embedding. By Disease2Vec, our framework generates a disease embedding tree (DETree) which effectively represents different clinical stages as a tree trajectory reflecting AD progression and thus can be used to predict clinical status by projecting individuals onto this continuous trajectory. Through this model, DETree can not only perform efficient and accurate prediction for patients at any stages of AD development (across five fine-grained clinical groups instead of typical two groups), but also provide richer status information by examining the projecting locations within a wide and continuous AD progression process. (Code will be available: https://github.com/qidianzl/Disease2Vec.).

MEET: A Multi-Band EEG Transformer for Brain States Decoding.

OBJECTIVE: Electroencephalography (EEG) is among the most widely used and inexpensive neuroimaging techniques. Compared to the CNN or RNN based models, Transformer can better capture the temporal information in EEG signals and focus more on global features of the brain's functional activities. Importantly, according to the multiscale nature of EEG signals, it is crucial to consider the multi-band concept into the design of EEG Transformer architecture. METHODS: We propose a novel Multi-band EEG Transformer (MEET) to represent and analyze the multiscale temporal time series of human brain EEG signals. MEET mainly includes three parts: 1) transform the EEG signals into multi-band images, and preserve the 3D spatial information between electrodes; 2) design a Band Attention Block to compute the attention maps of the stacked multi-band images and infer the fused feature maps; 3) apply the Temporal Self-Attention and Spatial Self-Attention modules to extract the spatiotemporal features for the characterization and differentiation of multi-frame dynamic brain states. RESULTS: The experimental results show that: 1) MEET outperforms state-of-the-art methods on multiple open EEG datasets (SEED, SEED-IV, WM) for brain states classification; 2) MEET demonstrates that 5-bands fusion is the best integration strategy; and 3) MEET identifies interpretable brain attention regions. SIGNIFICANCE: MEET is an interpretable and universal model based on the multiband-multiscale characteristics of EEG. CONCLUSION: The innovative combination of band attention and temporal/spatial self-attention mechanisms in MEET achieves promising data-driven learning of the temporal dependencies and spatial relationships of EEG signals across the entire brain in a holistic and comprehensive fashion.

Differentiating ChatGPT-Generated and Human-Written Medical Texts: Quantitative Study.

BACKGROUND: Large language models, such as ChatGPT, are capable of generating grammatically perfect and human-like text content, and a large number of ChatGPT-generated texts have appeared on the internet. However, medical texts, such as clinical notes and diagnoses, require rigorous validation, and erroneous medical content generated by ChatGPT could potentially lead to disinformation that poses significant harm to health care and the general public. OBJECTIVE: This study is among the first on responsible artificial intelligence-generated content in medicine. We focus on analyzing the differences between medical texts written by human experts and those generated by ChatGPT and designing machine learning workflows to effectively detect and differentiate medical texts generated by ChatGPT. METHODS: We first constructed a suite of data sets containing medical texts written by human experts and generated by ChatGPT. We analyzed the linguistic features of these 2 types of content and uncovered differences in vocabulary, parts-of-speech, dependency, sentiment, perplexity, and other aspects. Finally, we designed and implemented machine learning methods to detect medical text generated by ChatGPT. The data and code used in this paper are published on GitHub. RESULTS: Medical texts written by humans were more concrete, more diverse, and typically contained more useful information, while medical texts generated by ChatGPT paid more attention to fluency and logic and usually expressed general terminologies rather than effective information specific to the context of the problem. A bidirectional encoder representations from transformers-based model effectively detected medical texts generated by ChatGPT, and the F1 score exceeded 95%. CONCLUSIONS: Although text generated by ChatGPT is grammatically perfect and human-like, the linguistic characteristics of generated medical texts were different from those written by human experts. Medical text generated by ChatGPT could be effectively detected by the proposed machine learning algorithms. This study provides a pathway toward trustworthy and accountable use of large language models in medicine.

Genetic risks of Alzheimer's by APOE and MAPT on cortical morphology in young healthy adults.

Genetic risk factors such as APOE ε4 and MAPT (rs242557) A allele are associated with amyloid and tau pathways and grey matter changes at both early and established stages of Alzheimer's disease, but their effects on cortical morphology in young healthy adults remain unclear. A total of 144 participants aged from 18 to 24 underwent 3T MRI and genotyping for APOE and MAPT to investigate unique impacts of these genetic risk factors in a cohort without significant comorbid conditions such as metabolic and cardiovascular diseases. We segmented the cerebral cortex into 68 regions and calculated the cortical area, thickness, curvature and folding index for each region. Then, we trained machine learning models to classify APOE and MAPT genotypes using these morphological features. In addition, we applied a growing hierarchical self-organizing maps algorithm, which clustered the 68 regions into 4 subgroups representing different morphological patterns. Then, we performed general linear model analyses to estimate the interaction between APOE and MAPT on cortical patterns. We found that the classifiers using all cortical features could accurately classify individuals carrying genetic risks of dementia outperforming each individual feature alone. APOE ε4 carriers had a more convoluted and thinner cortex across the cerebral cortex. A similar pattern was found in MAPT A allele carriers only in the regions that are vulnerable for early tau pathology. With the clustering analysis, we found a synergetic effect between APOE ε4 and MAPT A allele, i.e. carriers of both risk factors showed the most deviation of cortical pattern from the typical pattern of that cluster. Genetic risk factors of dementia by APOE ε4 and MAPT (rs242557) A allele were associated with variations of cortical morphology, which can be observed in young healthy adults more than 30 years before Alzheimer's pathology is likely to occur and 50 years before dementia symptoms may begin.

Genetic Influence on Gyral Peaks.

Genetic mechanisms have been hypothesized to be a major determinant in the formation of cortical folding. Although there is an increasing number of studies examining the heritability of cortical folding, most of them focus on sulcal pits rather than gyral peaks. Gyral peaks, which reflect the highest local foci on gyri and are consistent across individuals, remain unstudied in terms of heritability. To address this knowledge gap, we used high-resolution data from the Human Connectome Project (HCP) to perform classical twin analysis and estimate the heritability of gyral peaks across various brain regions. Our results showed that the heritability of gyral peaks was heterogeneous across different cortical regions, but relatively symmetric between hemispheres. We also found that pits and peaks are different in a variety of anatomic and functional measures. Further, we explored the relationship between the levels of heritability and the formation of cortical folding by utilizing the evolutionary timeline of gyrification. Our findings indicate that the heritability estimates of both gyral peaks and sulcal pits decrease linearly with the evolution timeline of gyrification. This suggests that the cortical folds which formed earlier during gyrification are subject to stronger genetic influences than the later ones. Moreover, the pits and peaks coupled by their time of appearance are also positively correlated in respect of their heritability estimates. These results fill the knowledge gap regarding genetic influences on gyral peaks and significantly advance our understanding of how genetic factors shape the formation of cortical folding. The comparison between peaks and pits suggests that peaks are not a simple morphological mirror of pits but could help complete the understanding of folding patterns.

A generic framework for embedding human brain function with temporally correlated autoencoder.

Learning an effective and compact representation of human brain function from high-dimensional fMRI data is crucial for studying the brain's functional organization. Traditional representation methods such as independent component analysis (ICA) and sparse dictionary learning (SDL) mainly rely on matrix decomposition which represents the brain function as spatial brain networks and the corresponding temporal patterns. The correspondence of those brain networks across individuals are built by viewing them as one-hot vectors and then performing the matching. However, those one-hot vectors do not encode the regularity and/or variability of different brains very well, and thus are limited in effectively representing the functional brain activities across individuals and among different time points. To address this problem, in this paper, we formulate the human brain functional representation as an embedding problem, and propose a novel embedding framework based on the Transformer model to encode the brain function in a compact, stereotyped and comparable latent space where the brain activities are represented as dense embedding vectors. We evaluate the proposed embedding framework on the publicly available Human Connectome Project (HCP) task fMRI dataset. The experiments on brain state prediction task indicate the effectiveness and generalizability of the learned embedding. We also explore the interpretability of the learned embedding from both spatial and temporal perspective. In general, our approach provides novel insights on representing the regularity and variability of human brain function in a general, comparable, and stereotyped latent space.

An integrated finite element method and machine learning algorithm for brain morphology prediction.

The human brain development experiences a complex evolving cortical folding from a smooth surface to a convoluted ensemble of folds. Computational modeling of brain development has played an essential role in better understanding the process of cortical folding, but still leaves many questions to be answered. A major challenge faced by computational models is how to create massive brain developmental simulations with affordable computational sources to complement neuroimaging data and provide reliable predictions for brain folding. In this study, we leveraged the power of machine learning in data augmentation and prediction to develop a machine-learning-based finite element surrogate model to expedite brain computational simulations, predict brain folding morphology, and explore the underlying folding mechanism. To do so, massive finite element method (FEM) mechanical models were run to simulate brain development using the predefined brain patch growth models with adjustable surface curvature. Then, a GAN-based machine learning model was trained and validated with these produced computational data to predict brain folding morphology given a predefined initial configuration. The results indicate that the machine learning models can predict the complex morphology of folding patterns, including 3-hinge gyral folds. The close agreement between the folding patterns observed in FEM results and those predicted by machine learning models validate the feasibility of the proposed approach, offering a promising avenue to predict the brain development with given fetal brain configurations.

Core-Periphery Principle Guided Redesign of Self-Attention in Transformers.

Designing more efficient, reliable, and explainable neural network architectures is critical to studies that are based on artificial intelligence (AI) techniques. Previous studies, by post-hoc analysis, have found that the best-performing ANNs surprisingly resemble biological neural networks (BNN), which indicates that ANNs and BNNs may share some common principles to achieve optimal performance in either machine learning or cognitive/behavior tasks. Inspired by this phenomenon, we proactively instill organizational principles of BNNs to guide the redesign of ANNs. We leverage the Core-Periphery (CP) organization, which is widely found in human brain networks, to guide the information communication mechanism in the self-attention of vision transformer (ViT) and name this novel framework as CP-ViT. In CP-ViT, the attention operation between nodes is defined by a sparse graph with a Core-Periphery structure (CP graph), where the core nodes are redesigned and reorganized to play an integrative role and serve as a center for other periphery nodes to exchange information. We evaluated the proposed CP-ViT on multiple public datasets, including medical image datasets (INbreast) and natural image datasets. Interestingly, by incorporating the BNN-derived principle (CP structure) into the redesign of ViT, our CP-ViT outperforms other state-of-the-art ANNs. In general, our work advances the state of the art in three aspects: 1) This work provides novel insights for brain-inspired AI: we can utilize the principles found in BNNs to guide and improve our ANN architecture design; 2) We show that there exist sweet spots of CP graphs that lead to CP-ViTs with significantly improved performance; and 3) The core nodes in CP-ViT correspond to task-related meaningful and important image patches, which can significantly enhance the interpretability of the trained deep model.