Existing foundation models for neuroimaging are often prohibitively large and data-intensive. We introduce BrainSymphony, a lightweight, parameter-efficient foundation model that achieves state-of-the-art performance while being pre-trained on significantly smaller public datasets. BrainSymphony's strong multimodal architecture processes functional MRI data through parallel spatial and temporal transformer streams, which are then efficiently distilled into a unified representation by a Perceiver module. Concurrently, it models structural connectivity from diffusion MRI using a novel signed graph transformer to encode the brain's anatomical structure. These powerful, modality-specific representations are then integrated via an adaptive fusion gate. Despite its compact design, our model consistently outperforms larger models on a diverse range of downstream benchmarks, including classification, prediction, and unsupervised network identification tasks. Furthermore, our model revealed novel insights into brain dynamics using attention maps on a unique external psilocybin neuroimaging dataset (pre- and post-administration). BrainSymphony establishes that architecturally-aware, multimodal models can surpass their larger counterparts, paving the way for more accessible and powerful research in computational neuroscience.
In this work, we develop a method to identify continuous-time nonlinear networked dynamics via the Koopman operator framework. The proposed technique consists of two steps: the first step identifies the neighbors of each node, and the second step identifies the local dynamics at each node from a predefined set of dictionary functions. The technique can be used to either identify the Boolean network of interactions (first step) or to solve the complete network identification problem that amounts to estimating the local node dynamics and the nature of the node interactions (first and second steps). Under a sparsity assumption, the data required to identify the complete network dynamics is significantly less than the total number of dictionary functions describing the dynamics. This makes the proposed approach attractive for identifying large dimensional networks with sparse interconnections. The accuracy and performance of the proposed identification technique are demonstrated with several examples.
Longwave infrared (LWIR) hyperspectral imaging can be used for many tasks in remote sensing, including detecting and identifying effluent gases by LWIR sensors on airborne platforms. Once a potential plume has been detected, it needs to be identified to determine exactly what gas or gases are present in the plume. During identification, the background underneath the plume needs to be estimated and removed to reveal the spectral characteristics of the gas of interest. Current standard practice is to use ``global" background estimation, where the average of all non-plume pixels is used to estimate the background for each pixel in the plume. However, if this global background estimate does not model the true background under the plume well, then the resulting signal can be difficult to identify correctly. The importance of proper background estimation increases when dealing with weak signals, large libraries of gases of interest, and with uncommon or heterogeneous backgrounds. In this paper, we propose two methods of background estimation, in addition to three existing methods, and compare each against global background estimation to determine which perform best at estimating the true background radiance under a plume, and for increasing identification confidence using a neural network classification model. We compare the different methods using 640 simulated plumes. We find that PCA is best at estimating the true background under a plume, with a median of 18,000 times less MSE compared to global background estimation. Our proposed K-Nearest Segments algorithm improves median neural network identification confidence by 53.2%.
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 time-frequency characteristics of rs-fMRI blood oxygenation level-dependent (BOLD) signals may offer complementary information to traditional analysis techniques. 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) for time-frequency analysis of local neural activity within key functional networks to differentiate OUD subjects from healthy controls (HC). We obtain time-frequency features based on rs-fMRI BOLD signals from the default mode network (DMN), salience network (SN), and executive control network (ECN) for 31 OUD and 45 HC subjects. Then, we perform 5-fold cross-validation classification (OUD vs. HC) experiments to study the discriminative power of functional network features while taking into consideration significant demographic features. The DMN and SN show the most discriminative power, significantly (p < 0.05) outperforming chance baselines with mean F1 scores of 0.7097 and 0.7018, respectively, and mean AUCs of 0.8378 and 0.8755, respectively. Follow-up Boruta ML analysis of selected time-frequency (wavelet) features reveals significant (p < 0.05) detail coefficients for all three functional networks, underscoring the need for ML and time-frequency analysis of rs-fMRI BOLD signals in the study of OUD.
As is true of many complex tasks, the work of discovering, describing, and understanding the diversity of life on Earth (viz., biological systematics and taxonomy) requires many tools. Some of this work can be accomplished as it has been done in the past, but some aspects present us with challenges which traditional knowledge and tools cannot adequately resolve. One such challenge is presented by species complexes in which the morphological similarities among the group members make it difficult to reliably identify known species and detect new ones. We address this challenge by developing new tools using the principles of machine learning to resolve two specific questions related to species complexes. The first question is formulated as a classification problem in statistics and machine learning and the second question is an out-of-distribution (OOD) detection problem. We apply these tools to a species complex comprising Southeast Asian stream frogs (Limnonectes kuhlii complex) and employ a morphological character (hind limb skin texture) traditionally treated qualitatively in a quantitative and objective manner. We demonstrate that deep neural networks can successfully automate the classification of an image into a known species group for which it has been trained. We further demonstrate that the algorithm can successfully classify an image into a new class if the image does not belong to the existing classes. Additionally, we use the larger MNIST dataset to test the performance of our OOD detection algorithm. We finish our paper with some concluding remarks regarding the application of these methods to species complexes and our efforts to document true biodiversity. This paper has online supplementary materials.
We propose a novel algorithm called Unique Brain Network Identification Number, UBNIN for encoding the brain networks of individual subjects. To realize this objective, we employed structural MRI on 180 Parkinsons disease PD patients and 70 healthy controls HC from the National Institute of Mental Health and Neurosciences, India. We parcellated each subjects brain volume and constructed an individual adjacency matrix using the correlation between the gray matter volumes of every pair of regions. The unique code is derived from values representing connections for every node i, weighted by a factor of 2^1-i. The numerical representation UBNIN was observed to be distinct for each individual brain network, which may also be applied to other neuroimaging modalities. This model may be implemented as a neural signature of a persons unique brain connectivity, thereby making it useful for brainprinting applications. Additionally, we segregated the above datasets into five age cohorts to study the variation in network topology over age. Sparsity was adopted as the threshold estimate to binarize each age-based correlation matrix. For each age cohort, a decreasing trend was observed in the mean clustering coefficient with increasing sparsity. Significantly different clustering coefficients were noted in PD between age cohort B and C, C and E, and in HC between E and B, E and C, E and D, and C and D. Our findings suggest network connectivity patterns change with age, indicating network disruption may be due to the underlying neuropathology. Varying clustering coefficients for different cohorts indicate that information transfer between neighboring nodes changes with age. This provides evidence of age related brain shrinkage and network degeneration. We also discuss limitations and provide an open-access link to software codes and a help file for the entire study.
Can we identify the weights of a neural network by probing its input-output mapping? At first glance, this problem seems to have many solutions because of permutation, overparameterisation and activation function symmetries. Yet, we show that the incoming weight vector of each neuron is identifiable up to sign or scaling, depending on the activation function. Our novel method 'Expand-and-Cluster' can identify layer sizes and weights of a target network for all commonly used activation functions. Expand-and-Cluster consists of two phases: (i) to relax the non-convex optimisation problem, we train multiple overparameterised student networks to best imitate the target function; (ii) to reverse engineer the target network's weights, we employ an ad-hoc clustering procedure that reveals the learnt weight vectors shared between students -- these correspond to the target weight vectors. We demonstrate successful weights and size recovery of trained shallow and deep networks with less than 10\% overhead in the layer size and describe an `ease-of-identifiability' axis by analysing 150 synthetic problems of variable difficulty.
The identification of distribution network topology and parameters is a critical problem that lays the foundation for improving network efficiency, enhancing reliability, and increasing its capacity to host distributed energy resources. Network identification problems often involve estimating a large number of parameters based on highly correlated measurements, resulting in an ill-conditioned and computationally demanding estimation process. We address these challenges by proposing two admittance matrix estimation methods. In the first method, we use the eigendecomposition of the admittance matrix to generalize the notion of stationarity to electrical signals and demonstrate how the stationarity property can be used to facilitate a maximum a posteriori estimation procedure. We relax the stationarity assumption in the second proposed method by employing Linear Minimum Mean Square Error (LMMSE) estimation. Since LMMSE estimation is often ill-conditioned, we introduce an approximate well-conditioned solution based on eigenvalue truncation. Our quantitative results demonstrate the improvement in computational efficiency compared to the state-of-the-art methods while preserving the estimation accuracy.
Functional networks characterize the coordinated neural activity observed by functional neuroimaging. The prevalence of different networks during resting state periods provide useful features for predicting the trajectory of neurodegenerative diseases. Techniques for network estimation rely on statistical correlation or dependence between voxels. Due to the large number of voxels, rather than consider the voxel-to-voxel correlations between all voxels, a small set of seed voxels are chosen. Consequently, the network identification may depend on the selected seeds. As an alternative, we propose to fit first-order linear models with sparse priors on the coefficients to model activity across the entire set of cortical grey matter voxels as a linear combination of a smaller subset of voxels. We propose a two-stage algorithm for voxel subset selection that uses different sparsity-inducing regularization approaches to identify subject-specific causally predictive voxels. To reveal the functional networks among these voxels, we then apply independent component analysis (ICA) to model these voxels' signals as a mixture of latent sources each defining a functional network. Based on the inter-subject similarity of the sources' spatial patterns we identify independent sources that are well-matched across subjects but fail to match the independent sources from a group-based ICA. These are resting state networks, common across subjects that group ICA does not reveal. These complementary networks could help to better identify neurodegeneration, a task left for future work.
Several studies on ship maneuvering models have been conducted using captive model tests or computational fluid dynamics (CFD) and physical models, such as the maneuvering modeling group (MMG) model. A new system identification method for generating a low-speed maneuvering model using recurrent neural networks (RNNs) and free running model tests is proposed in this study. We especially focus on a low-speed maneuver such as the final phase in berthing to achieve automatic berthing control. Accurate dynamic modeling with minimum modeling error is highly desired to establish a model-based control system. We propose a new loss function that reduces the effect of the noise included in the training data. Besides, we revealed the following facts - an RNN that ignores the memory before a certain time improved the prediction accuracy compared with the "standard" RNN, and the random maneuver test was effective in obtaining an accurate berthing maneuver model. In addition, several low-speed free running model tests were performed for the scale model of the M.V. Esso Osaka. As a result, this paper showed that the proposed method using a neural network model could accurately represent low-speed maneuvering motions.
Identification methods for dynamic networks typically require prior knowledge of the network and disturbance topology, and often rely on solving poorly scalable non-convex optimization problems. While methods for estimating network topology are available in the literature, less attention has been paid to estimating the disturbance topology, i.e., the (spatial) noise correlation structure and the noise rank in a filtered white noise representation of the disturbance signal. In this work we present an identification method for dynamic networks, in which an estimation of the disturbance topology precedes the identification of the full dynamic network with known network topology. To this end we extend the multi-step Sequential Linear Regression and Weighted Null Space Fitting methods to deal with reduced rank noise, and use these methods to estimate the disturbance topology and the network dynamics in the full measurement situation. As a result, we provide a multi-step least squares algorithm with parallel computation capabilities and that rely only on explicit analytical solutions, thereby avoiding the usual non-convex optimizations involved. Consequently we consistently estimate dynamic networks of Box Jenkins model structure, while keeping the computational burden low. We provide a consistency proof that includes path-based data informativity conditions for allocation of excitation signals in the experimental design. Numerical simulations performed on a dynamic network with reduced rank noise clearly illustrate the potential of this method.
In classical approaches of dynamic network identification, in order to identify a system (module) embedded in a dynamic network, one has to formulate a Multi-input-Single-output (MISO) identification problem that requires identification of a parametric model for all the modules constituting the MISO setup including (possibly) the noise model, and determine their model order. This requirement leads to model order selection steps for modules that are of no interest to the experimenter which increases the computational complexity for large-sized networks. Also, identification using a parametric noise model (like BJ method) can suffer from local minima, however neglecting the noise model has its impact on the variance of the estimates. In this paper, we provide a two-step identification approach to avoid these problems. The first step involves performing a non-parametric indirect approach for a MISO identification problem to get the non-parametric frequency response function (FRF) estimates and its variance as a function of frequency. In the second step, the estimated non-parametric FRF of the target module is smoothed using a parametric frequency domain estimator with the estimated variance from the previous step as the non-parametric noise model. The developed approach is practical with weak assumptions on noise, uses the available toolbox, requires a parametric model only for the target module of interest, and uses a non-parametric noise model to reduce the variance of the estimates. Numerical simulations illustrate the potentials of the introduced method in comparison with the classical identification methods.
Complex networked systems can be modeled as graphs with nodes representing the agents and links describing the dynamic coupling between them. Previous work on network identification has shown that the network structure of linear time-invariant (LTI) systems can be reconstructed from the joint power spectrum of the data streams. These results assumed that data is perfectly measured. However, real-world data is subject to many corruptions, such as inaccurate time-stamps, noise, and data loss. We show that identifying the structure of linear time-invariant systems using corrupt measurements results in the inference of erroneous links. We provide an exact characterization and prove that such erroneous links are restricted to the neighborhood of the perturbed node. We extend the analysis of LTI systems to the case of Markov random fields with corrupt measurements. We show that data corruption in Markov random fields results in spurious probabilistic relationships in precisely the locations where spurious links arise in LTI systems.
A novel unsupervised deep learning method is developed to identify individual-specific large scale brain functional networks (FNs) from resting-state fMRI (rsfMRI) in an end-to-end learning fashion. Our method leverages deep Encoder-Decoder networks and conventional brain decomposition models to identify individual-specific FNs in an unsupervised learning framework and facilitate fast inference for new individuals with one forward pass of the deep network. Particularly, convolutional neural networks (CNNs) with an Encoder-Decoder architecture are adopted to identify individual-specific FNs from rsfMRI data by optimizing their data fitting and sparsity regularization terms that are commonly used in brain decomposition models. Moreover, a time-invariant representation learning module is designed to learn features invariant to temporal orders of time points of rsfMRI data. The proposed method has been validated based on a large rsfMRI dataset and experimental results have demonstrated that our method could obtain individual-specific FNs which are consistent with well-established FNs and are informative for predicting brain age, indicating that the individual-specific FNs identified truly captured the underlying variability of individualized functional neuroanatomy.
In this paper, the problem of synthesizing a distributed controller from data is considered, with the objective to optimize a model-reference control criterion. We establish an explicit ideal distributed controller that solves the model-reference control problem for a structured reference model. On the basis of input-output data collected from the interconnected system, a virtual experiment setup is constructed which leads to a network identification problem. We formulate a prediction-error identification criterion that has the same global optimum as the model-reference criterion, when the controller class contains the ideal distributed controller. The developed distributed controller synthesis method is illustrated on an academic example network of nine subsystems and the influence of the controller interconnection structure on the achieved closed-loop performance is analyzed.
Gene expression represents a fundamental interface between genes and environment in the development and ongoing plasticity of the human organism. Individual differences in gene expression are likely to underpin much of human diversity, including psychiatric illness. Gene expression shows a distinct regulatory pattern in different tissues. Therefore, brain tissue analysis provides insights into brain disorder mechanisms. Furthermore, mechanistic understanding of gene regulatory pattern can be provided through studying the underlying relationships as a complex network. Identification of brain specific gene relationships provides a complementary framework in which to tackle the complex dysregulations that occur in neuropsychiatric and other neurological disorders. Using a systems approach established in Mendelian randomization and Bayesian Network, we integrated genetic and transcriptomic data from the common-mind consortium and identified transcriptomic causal networks in observational studies. Focusing on Schizophrenia disorder, we identified high impact genes and revealed their underlying pathways in brain tissue. In addition, we generated novel hypotheses including genes as causes of the schizophrenia-associated genes and new genes associated with Schizophrenia. This approach may facilitate a better understanding of the disease mechanism that is complementary to molecular experimental studies especially for complex systems and large-scale data sets.
Conserved moieties are groups of atoms that remain intact in all reactions of a metabolic network. Identification of conserved moieties gives insight into the structure and function of metabolic networks and facilitates metabolic modelling. All moiety conservation relations can be represented as nonnegative integer vectors in the left null space of the stoichiometric matrix corresponding to a biochemical network. Algorithms exist to compute such vectors based only on reaction stoichiometry but their computational complexity has limited their application to relatively small metabolic networks. Moreover, the vectors returned by existing algorithms do not, in general, represent conservation of a specific moiety with a defined atomic structure. Here, we show that identification of conserved moieties requires data on reaction atom mappings in addition to stoichiometry. We present a novel method to identify conserved moieties in metabolic networks by graph theoretical analysis of their underlying atom transition networks. Our method returns the exact group of atoms belonging to each conserved moiety as well as the corresponding vector in the left null space of the stoichiometric matrix. It can be implemented as a pipeline of polynomial time algorithms. Our implementation completes in under five minutes on a metabolic network with more than 4,000 mass balanced reactions. The scalability of the method enables extension of existing applications for moiety conservation relations to genome-scale metabolic networks. We also give examples of new applications made possible by elucidating the atomic structure of conserved moieties.
We consider the problem of identifying parameters of a particular class of Markov chains, called Bernoulli Autoregressive (BAR) processes. The structure of any BAR model is encoded by a directed graph. Incoming edges to a node in the graph indicate that the state of the node at a particular time instant is influenced by the states of the corresponding parental nodes in the previous time instant. The associated edge weights determine the corresponding level of influence from each parental node. In the simplest setup, the Bernoulli parameter of a particular node's state variable is a convex combination of the parental node states in the previous time instant and an additional Bernoulli noise random variable. This paper focuses on the problem of edge weight identification using Maximum Likelihood (ML) estimation and proves that the ML estimator is strongly consistent for two variants of the BAR model. We additionally derive closed-form estimators for the aforementioned two variants and prove their strong consistency.
Learning the differential statistical dependency network between two contexts is essential for many real-life applications, mostly in the high dimensional low sample regime. In this paper, we propose a novel differential network estimator that allows integrating various sources of knowledge beyond data samples. The proposed estimator is scalable to a large number of variables and achieves a sharp asymptotic convergence rate. Empirical experiments on extensive simulated data and four real-world applications (one on neuroimaging and three from functional genomics) show that our approach achieves improved differential network estimation and provides better supports to downstream tasks like classification. Our results highlight significant benefits of integrating group, spatial and anatomic knowledge during differential genetic network identification and brain connectome change discovery.
Real-world face recognition requires an ability to perceive the unique features of an individual face across multiple, variable images. The primate visual system solves the problem of image invariance using cascades of neurons that convert images of faces into categorical representations of facial identity. Deep convolutional neural networks (DCNNs) also create generalizable face representations, but with cascades of simulated neurons. DCNN representations can be examined in a multidimensional "face space", with identities and image parameters quantified via their projections onto the axes that define the space. We examined the organization of viewpoint, illumination, gender, and identity in this space. We show that the network creates a highly organized, hierarchically nested, face similarity structure in which information about face identity and imaging characteristics coexist. Natural image variation is accommodated in this hierarchy, with face identity nested under gender, illumination nested under identity, and viewpoint nested under illumination. To examine identity, we caricatured faces and found that network identification accuracy increased with caricature level, and--mimicking human perception--a caricatured distortion of a face "resembled" its veridical counterpart. Caricatures improved performance by moving the identity away from other identities in the face space and minimizing the effects of illumination and viewpoint. Deep networks produce face representations that solve long-standing computational problems in generalized face recognition. They also provide a unitary theoretical framework for reconciling decades of behavioral and neural results that emphasized either the image or the object/face in representations, without understanding how a neural code could seamlessly accommodate both.