Brain sodium MRI-derived priors support the estimation of epileptogenic zones using personalized model-based methods in epilepsy.

Patients presenting with drug-resistant epilepsy are eligible for surgery aiming to remove the regions involved in the production of seizure activities, the so-called epileptogenic zone network (EZN). Thus the accurate estimation of the EZN is crucial. Data-driven, personalized virtual brain models derived from patient-specific anatomical and functional data are used in Virtual Epileptic Patient (VEP) to estimate the EZN via optimization methods from Bayesian inference. The Bayesian inference approach used in previous VEP integrates priors, based on the features of stereotactic-electroencephalography (SEEG) seizures' recordings. Here, we propose new priors, based on quantitative 23Na-MRI. The 23Na-MRI data were acquired at 7T and provided several features characterizing the sodium signal decay. The hypothesis is that the sodium features are biomarkers of neuronal excitability related to the EZN and will add additional information to VEP estimation. In this paper, we first proposed the mapping from 23Na-MRI features to predict the EZN via a machine learning approach. Then, we exploited these predictions as priors in the VEP pipeline. The statistical results demonstrated that compared with the results from current VEP, the result from VEP based on 23Na-MRI prior has better balanced accuracy, and the similar weighted harmonic mean of the precision and recall.

For the first time quantitative ²³Na-MRI were used as prior information to improve estimation of epileptogenic network (EZN) using VEP pipeline, a personalized whole-brain network modeling from patient's specific data. The prior information of EZN can be derived from ²³Na-MRI features using logistic regression predictions. The ²³Na-MRI priors inferred EZNs has a better balanced accuracy than the previously used priors or the no-prior condition.

Predictive modeling for the adsorptive and photocatalytic removal of phenolic contaminants from water using artificial neural networks.

Numerous harmful phenolic contaminants are discharged into water that pose a serious threat to environment where two of the most important purification methodologies for the mitigation of phenolic contaminants are adsorption and photocatalysis. Besides cost, each process has drawbacks in terms of productivity, environmental impact, sludge creation, and the development of harmful by-products. To overcome these limitations, the modeling and optimization of water treatment methods is required. Artificial Intelligence (AI) is employed for the interpretation of treatment-based processes due to powerful learning, simplicity, high estimation accuracy, effectiveness, and improvement of process efficiency where artificial neural networks (ANNs) are most frequently employed for predicting and analyzing the efficiency of processes applied for the mitigation of these phenolic contaminants from water. ANNs are superior to conventional linear regression models because the latter are incapable of dealing with non-linear systems. ANNs can also reduce the operational cost of treating phenol-contaminated water. A correlation coefficient of >0.99 can be achieved using ANN with enhanced phenol mitigation percentage accuracy generally ranging from 80 % to 99.99 %. Using ANN optimization, the maximum phenol mitigation efficiencies achieved were 99.99 % for phenol, 99.93 % for bisphenol A, 99.6 % for nonylphenol, 97.1 % for 2-nitrophenol, 96.6 % for 4-chlorophenol and 90 % for 2,6-dichlorophenol. In numerous ANN models, Levenberg-Marquardt backpropagation algorithm for training was employed using MATLAB software. This study overviews their employment and application for optimization and modeling of removal processes and explicitly discusses the important input and output parameters necessary for better performance of the system. The comparison of ANNs with other AI techniques revealed that ANNs have better predictability for mitigation of most of the phenolic contaminants. Furthermore, several challenges and future prospects have also been discussed.

Large-Scale Mechanistic Models of Brain Circuits with Biophysically and Morphologically Detailed Neurons.

Understanding the brain requires studying its multiscale interactions from molecules to networks. The increasing availability of large-scale datasets detailing brain circuit composition, connectivity, and activity is transforming neuroscience. However, integrating and interpreting this data remains challenging. Concurrently, advances in supercomputing and sophisticated modeling tools now enable the development of highly detailed, large-scale biophysical circuit models. These mechanistic multiscale models offer a method to systematically integrate experimental data, facilitating investigations into brain structure, function, and disease. This review, based on a Society for Neuroscience 2024 MiniSymposium, aims to disseminate recent advances in large-scale mechanistic modeling to the broader community. It highlights (1) examples of current models for various brain regions developed through experimental data integration; (2) their predictive capabilities regarding cellular and circuit mechanisms underlying experimental recordings (e.g., membrane voltage, spikes, local-field potential, electroencephalography/magnetoencephalography) and brain function; and (3) their use in simulating biomarkers for brain diseases like epilepsy, depression, schizophrenia, and Parkinson's, aiding in understanding their biophysical underpinnings and developing novel treatments. The review showcases state-of-the-art models covering hippocampus, somatosensory, visual, motor, auditory cortical, and thalamic circuits across species. These models predict neural activity at multiple scales and provide insights into the biophysical mechanisms underlying sensation, motor behavior, brain signals, neural coding, disease, pharmacological interventions, and neural stimulation. Collaboration with experimental neuroscientists and clinicians is essential for the development and validation of these models, particularly as datasets grow. Hence, this review aims to foster interest in detailed brain circuit models, leading to cross-disciplinary collaborations that accelerate brain research.

Predicting the Progression of Mild Cognitive Impairment to Alzheimer's Dementia Using Recurrent Neural Networks With a Series of Neuropsychological Tests.

BACKGROUND AND PURPOSE: The prevalence of Alzheimer's dementia (AD) is increasing as populations age, causing immense suffering for patients, families, and communities. Unfortunately, no treatments for this neurodegenerative disease have been established. Predicting AD is therefore becoming more important, because early diagnosis is the best way to prevent its onset and delay its progression. METHODS: Mild cognitive impairment (MCI) is the stage between normal cognition and AD, with large variations in its progression. The disease can be effectively managed by accurately predicting the probability of MCI progressing to AD over several years. In this study we used the Alzheimer's Disease Neuroimaging Initiative dataset to predict the progression of MCI to AD over a 3-year period from baseline. We developed and compared various recurrent neural network (RNN) models to determine the predictive effectiveness of four neuropsychological (NP) tests and magnetic resonance imaging (MRI) data at baseline. RESULTS: The experimental results confirmed that the Preclinical Alzheimer's Cognitive Composite score was the most effective of the four NP tests, and that the prediction performance of the NP tests improved over time. Moreover, the gated recurrent unit model exhibited the best performance among the prediction models, with an average area under the receiver operating characteristic curve of 0.916. CONCLUSIONS: Timely prediction of progression from MCI to AD can be achieved using a series of NP test results and an RNN, both with and without using the baseline MRI data.

Modeling interdependence between climatic factors, commodities, and financial markets.

This paper introduces a comprehensive approach to studying the impact of climate-related factors on commodity and financial markets using network analysis. We utilize a Bayesian network Vector Autoregressive model to investigate whether climate risk significantly influ-ences commodity prices and financial market returns. Our findings provide evidence of a climate effect on major commodities and global financial markets. Specifically, we identify Crude oil, Cotton, and Sugar as the commodities most affected by climate risk, with Gold demonstrating the least susceptibility. Additionally, we observe that climate-related risk on commodities is likely propagated by patterns such as PNA, NN1, and AO. In terms of financial markets, we find that stock markets in Hong Kong, India, and Spain are the most susceptible to climate risk, while Switzerland's market appears to be the least affected. Furthermore, we document evidence that climate-related risk capable of altering financial markets is likely propagated by factors like ENP, NN1, and WH. Overall, our study underscores the intricate relationship between climate factors and market dynamics, highlighting the importance of considering climate risk in assessing market behavior and performance.

Causal inference over stochastic networks.

Claiming causal inferences in network settings necessitates careful consideration of the often complex dependency between outcomes for actors. Of particular importance are treatment spillover or outcome interference effects. We consider causal inference when the actors are connected via an underlying network structure. Our key contribution is a model for causality when the underlying network is endogenous; where the ties between actors and the actor covariates are statistically dependent. We develop a joint model for the relational and covariate generating process that avoids restrictive separability and fixed network assumptions, as these rarely hold in realistic social settings. While our framework can be used with general models, we develop the highly expressive class of Exponential-family Random Network models (ERNM) of which Markov random fields and Exponential-family Random Graph models are special cases. We present potential outcome-based inference within a Bayesian framework and propose a modification to the exchange algorithm to allow for sampling from ERNM posteriors. We present results of a simulation study demonstrating the validity of the approach. Finally, we demonstrate the value of the framework in a case study of smoking in the context of adolescent friendship networks.

Reused Protein Segments Linked to Functional Dynamics.

Protein space is characterized by extensive recurrence, or "reuse," of parts, suggesting that new proteins and domains can evolve by mixing-and-matching of existing segments. From an evolutionary perspective, for a given combination to persist, the protein segments should presumably not only match geometrically but also dynamically communicate with each other to allow concerted motions that are key to function. Evidence from protein space supports the premise that domains indeed combine in this manner; we explore whether a similar phenomenon can be observed at the sub-domain level. To this end, we use Gaussian Network Models (GNMs) to calculate the so-called soft modes, or low-frequency modes of motion for a dataset of 150 protein domains. Modes of motion can be used to decompose a domain into segments of consecutive amino acids that we call "dynamic elements", each of which belongs to one of two parts that move in opposite senses. We find that, in many cases, the dynamic elements, detected based on GNM analysis, correspond to established "themes": Sub-domain-level segments that have been shown to recur in protein space, and which were detected in previous research using sequence similarity alone (i.e. completely independently of the GNM analysis). This statistically significant correlation hints at the importance of dynamics in evolution. Overall, the results are consistent with an evolutionary scenario where proteins have emerged from themes that need to match each other both geometrically and dynamically, e.g. to facilitate allosteric regulation.

MV-GNN: Generation of continuous geometric representations of mitral valve motion from 3D+t echocardiography.

We present a geometric deep-learning method for reconstructing a temporally continuous mitral valve surface mesh from 3D transesophageal echocardiography sequences. Our approach features a supervised end-to-end deep learning architecture that combines a convolutional neural network-based voxel encoder and decoder with a graph neural network-based multi-resolution mesh decoder, all trained on sparse landmark annotations. Key elements of our methodology include a tube-shaped prototype mesh with labeled vertices, a specialized loss function to preserve the known inlet and outlet, and a rigid alignment system for anatomical landmarks. A custom term in the loss function prevents self-intersecting geometries within the valve mesh, promoting point correspondence and facilitating a continuous representation of valve anatomy over time. An ablation study evaluates the impact of different loss term configurations on model performance, highlighting the effectiveness of each individual loss term. Our Mitral Valve Graph Neural Network (MV-GNN) outperforms existing deep-learning methods on most distance metrics for the annulus and leaflets. The continuous valve motion representations generated by our approach (3D+t) exhibit distance measures comparable to our 3D solution, demonstrating its potential for analyzing mitral valve dynamics and enhancing personalized simulations for hemodynamic assessment and therapy planning.

Application of Deep Neural Networks in the Manufacturing Process of Mesenchymal Stem Cells Therapeutics.

Current image-based analysis methods for monitoring cell confluency and status depend on individual interpretations, which can lead to wide variations in the quality of cell therapeutics. To overcome these limitations, images of mesenchymal stem cells cultured adherently in various types of culture vessels were captured and analyzed using a deep neural network. Among the various deep learning methods, a classification and detection algorithm was selected to verify cell confluency and status. We confirmed that the image classification algorithm demonstrates significant accuracy for both single- and multistack images. Abnormal cells could be detected exclusively in single-stack images, as multistack culture was performed only when abnormal cells were absent in the single-stack culture. This study is the first to analyze cell images based on a deep learning method that directly impacts yield and quality, which are important product parameters in stem cell therapeutics.

Enhancing schizophrenia phenotype prediction from genotype data through knowledge-driven deep neural network models.

This article explores deep learning model design, drawing inspiration from the omnigenic model and genetic heterogeneity concepts, to improve schizophrenia prediction using genotype data. It introduces an innovative three-step approach leveraging neural networks' capabilities to efficiently handle genetic interactions. A locally connected network initially routes input data from variants to their corresponding genes. The second step employs an Encoder-Decoder to capture relationships among identified genes. The final model integrates knowledge from the first two and incorporates a parallel component to consider the effects of additional genes. This expansion enhances prediction scores by considering a larger number of genes. Trained models achieved an average AUC of 0.83, surpassing other genotype-trained models and matching gene expression dataset-based approaches. Additionally, tests on held-out sets reported an average sensitivity of 0.72 and an accuracy of 0.76, aligning with schizophrenia heritability predictions. Moreover, the study addresses genetic heterogeneity challenges by considering diverse population subsets.

Advancing critical care recovery: The pivotal role of machine learning in early detection of intensive care unit-acquired weakness.

This editorial explores the significant challenge of intensive care unit-acquired weakness (ICU-AW), a prevalent condition affecting critically ill patients, characterized by profound muscle weakness and complicating patient recovery. Highlighting the paradox of modern medical advances, it emphasizes the urgent need for early identification and intervention to mitigate ICU-AW's impact. Innovatively, the study by Wang et al is showcased for employing a multilayer perceptron neural network model, achieving high accuracy in predicting ICU-AW risk. This advancement underscores the potential of neural network models in enhancing patient care but also calls for continued research to address limitations and improve model applicability. The editorial advocates for the development and validation of sophisticated predictive tools, aiming for personalized care strategies to reduce ICU-AW incidence and severity, ultimately improving patient outcomes in critical care settings.

Relationship evolution shapes inter-brain synchrony in affective sharing: The role of self-expansion.

The development of social relationships influences a person's self-concept, which in turn affects their perceptions and neural correlates in social interactions. This study employed an EEG-based hyperscanning technique and a longitudinal design to investigate how the evolution of interpersonal relationships impacts inter-brain synchrony during nonverbal social-emotional interactions. The framework for this study is based on the self-expansion model. We found that dyads exhibited enhanced affective sharing abilities and increased brain-to-brain synchrony, particularly in the gamma rhythm across the frontal, parietal, and left temporoparietal regions, after seven months together compared to when they first met. Additionally, the results indicate that inter-brain coupling evolves as relationships develop, with synchrony in nonverbal social-emotional interactions increasing as self-expansion progresses. Crucially, in the deep learning model, interpersonal closeness can be successfully classified by inter-brain synchrony during emotional-social interactions. The longitudinal EEG-hyperscanning design of our study allows for capturing dynamic changes over time, offering new insights into the neurobiological foundations of social interaction and the potential of neural synchrony as a biomarker for relationship dynamics.

Construction of neural network model for exercise load monitoring based on yoga training data and rehabilitation therapy.

The Internet of Things is based on the traditional Internet and its purpose is to achieve information exchange between users and devices, as well as between devices. The rapid development of sensor technology, communication network technology, and computer technology has enriched the coverage of the Internet of Things, including a wide range of intelligent applications such as healthcare, smart cities, and smart homes. The development of high-performance computing and machine learning technologies has promoted the wide application of intelligent auxiliary systems in sports medicine. With the rapid development of yoga in the field of sports, athletes can play the various functions of yoga, improve their physical strength and quality, and improve their strength, flexibility, etc., cultivate positive, optimistic, and healthy emotions, and these are conducive to rehabilitation treatment after sports injuries. Therefore, it is feasible and feasible to introduce yoga training into the monitoring of the exercise load of athletes. In this paper, neural network technology was used to break the traditional training method based on experience. Based on yoga training data, through experimental exercise research, it could explore a new effective way to monitor exercise load and rehabilitation treatment, and build an exercise load monitoring model of the Ant Colony Optimization (ACO) neural network. By sorting out the data, statistics and analysis of the data, this article confirmed the effect of yoga training on reducing fatigue after exercise. The experimental results showed that the prediction value obtained by the ACO neural network model was 9.106, and the error was only -0.003 compared to the actual detection value of 9.109. This result showed that the ACO neural network model can perfectly fit the functional relationship between yoga training level and exercise load and has high prediction accuracy. This also marked that the development of high-performance computing systems has entered a new journey in the field of sports and health.

Allosteric modulation of serotonin and dopamine transporters: New insights from computations and experiments.

Human monoamine transporters (MATs) are critical to regulating monoaminergic neurotransmission by translocating their substrates from the synaptic space back into the presynaptic neurons. As such, their primary substrate binding site S1 has been targeted by a wide range of compounds for treating neuropsychiatric and neurodegenerative disorders including depression, ADHD, neuropathic pain, and anxiety disorders. We present here a comparative study of the structural dynamics and ligand-binding properties of two MATs, dopamine transporter (DAT) and serotonin transporter (SERT), with focus on the allosteric modulation of their transport function by drugs or substrates that consistently bind a secondary site S2, proposed to serve as an allosteric site. Our systematic analysis of the conformational space and dynamics of a dataset of 50 structures resolved for DAT and SERT in the presence of one or more ligands/drugs reveals the specific residues playing a consistent role in coordinating the small molecules bound to subsites S2-I and S2-II within S2, such as R476 and Y481 in dDAT and E494, P561, and F556 in hSERT. Further analysis reveals how DAT and SERT differ in their two principal modes of structural changes, PC1 and PC2. Notably, PC1 underlies the transition between outward- and inward-facing states of the transporters as well as their gating; whereas PC2 supports the rearrangements of TM helices near the S2 site. Finally, the examination of cross-correlations between structural elements lining the respective sites S1 and S2 point to the crucial role of coupled motions between TM6a and TM10. In particular, we note the involvement of hSERT residues F335 and G338, and E493-E494-T497 belonging to these two respective helices, in establishing the allosteric communication between S1 and S2. These results help understand the molecular basis of the action of drugs that bind to the S2 site of DAT or SERT. They also provide a basis for designing allosteric modulators that may provide better control of specific interactions and cellular pathways, rather than indiscriminately inhibiting the transporter by targeting its orthosteric site.

Watershed features shape spatial patterns of fish tissue mercury in a boreal river network.

Freshwater mercury (Hg) contamination is a widespread environmental concern but how proximate sources and downstream transport shape Hg spatial patterns in riverine food webs is poorly understood. We measured total Hg (THg) in slimy sculpin (Cottus cognatus) across the Kuskokwim River, a large boreal river in western Alaska and home to subsistence fishing communities which rely on fish for primary nutrition. We used spatial stream network models (SSNMs) to quantify watershed and instream conditions influencing sculpin THg. Spatial covariates for local watershed geology and slope accounted for 55 % of observed variation in sculpin THg and evidence for downstream transport of Hg in sculpins was weak. Empirical semivariograms indicated these spatial covariates accounted for most spatial autocorrelation in observed THg. Watershed geology and slope explained up to 70 % of sculpin THg variation when SSNMs accounted for instream spatial dependence. Our results provide network-wide predictions for fish tissue THg based largely on publicly available geospatial data and open-source software for SSNMs, and demonstrate how these emerging models can be used to understand contaminant behavior in spatially complex aquatic ecosystems.

Learning the Meanings of Function Words From Grounded Language Using a Visual Question Answering Model.

Interpreting a seemingly simple function word like "or," "behind," or "more" can require logical, numerical, and relational reasoning. How are such words learned by children? Prior acquisition theories have often relied on positing a foundation of innate knowledge. Yet recent neural-network-based visual question answering models apparently can learn to use function words as part of answering questions about complex visual scenes. In this paper, we study what these models learn about function words, in the hope of better understanding how the meanings of these words can be learned by both models and children. We show that recurrent models trained on visually grounded language learn gradient semantics for function words requiring spatial and numerical reasoning. Furthermore, we find that these models can learn the meanings of logical connectives and and or without any prior knowledge of logical reasoning as well as early evidence that they are sensitive to alternative expressions when interpreting language. Finally, we show that word learning difficulty is dependent on the frequency of models' input. Our findings offer proof-of-concept evidence that it is possible to learn the nuanced interpretations of function words in a visually grounded context by using non-symbolic general statistical learning algorithms, without any prior knowledge of linguistic meaning.

Prediction of surgery-first approach orthognathic surgery using deep learning models.

The surgery-first approach (SFA) orthognathic surgery can be beneficial due to reduced overall treatment time and earlier profile improvement. The objective of this study was to utilize deep learning to predict the treatment modality of SFA or the orthodontics-first approach (OFA) in orthognathic surgery patients and assess its clinical accuracy. A supervised deep learning model using three convolutional neural networks (CNNs) was trained based on lateral cephalograms and occlusal views of 3D dental model scans from 228 skeletal Class III malocclusion patients (114 treated by SFA and 114 by OFA). An ablation study of five groups (lateral cephalogram only, mandible image only, maxilla image only, maxilla and mandible images, and all data combined) was conducted to assess the influence of each input type. The results showed the average validation accuracy, precision, recall, F1 score, and AUROC for the five folds were 0.978, 0.980, 0.980, 0.980, and 0.998 ; the average testing results for the five folds were 0.906, 0.986, 0.828, 0.892, and 0.952. The lateral cephalogram only group had the least accuracy, while the maxilla image only group had the best accuracy. Deep learning provides a novel method for an accelerated workflow, automated assisted decision-making, and personalized treatment planning.

Computational biology approaches for drug repurposing.

The drug discovery and development (DDD) process greatly relies on the data available in various forms to generate hypotheses for novel drug design. The complex and heterogeneous nature of biological data makes it difficult to utilize or gather meaningful information as such. Computational biology techniques have provided us with opportunities to better understand biological systems through refining and organizing large amounts of data into actionable and systematic purviews. The drug repurposing approach has been utilized to overcome the expansive time periods and costs associated with traditional drug development. It deals with discovering new uses of already approved drugs that have an established safety and efficacy profile, thereby, requiring them to go through fewer development phases. Thus, drug repurposing through computational biology provides a systematic approach to drug development and overcomes the constraints of traditional processes. The current chapter covers the basics, approaches and tools of computational biology that can be employed to effectively develop repurposing profile of already approved drug molecules.

Paradoxical replay can protect contextual task representations from destructive interference when experience is unbalanced.

Experience replay is a powerful mechanism to learn efficiently from limited experience. Despite several decades of compelling experimental results, the factors that determine which experiences are selected for replay remain unclear. A particular challenge for current theories is that on tasks that feature unbalanced experience, rats paradoxically replay the less-experienced trajectory. To understand why, we simulated a feedforward neural network with two regimes: rich learning (structured representations tailored to task demands) and lazy learning (unstructured, task-agnostic representations). Rich, but not lazy, representations degraded following unbalanced experience, an effect that could be reversed with paradoxical replay. To test if this computational principle can account for the experimental data, we examined the relationship between paradoxical replay and learned task representations in the rat hippocampus. Strikingly, we found a strong association between the richness of learned task representations and the paradoxicality of replay. Taken together, these results suggest that paradoxical replay specifically serves to protect rich representations from the destructive effects of unbalanced experience, and more generally demonstrate a novel interaction between the nature of task representations and the function of replay in artificial and biological systems.

The future of psychological treatments: The Marburg Declaration.

Although psychological treatments are broadly recognized as evidence-based interventions for various mental disorders, challenges remain. For example, a substantial proportion of patients receiving such treatments do not fully recover, and many obstacles hinder the dissemination, implementation, and training of psychological treatments. These problems require those in our field to rethink some of our basic models of mental disorders and their treatments, and question how research and practice in clinical psychology should progress. To answer these questions, a group of experts of clinical psychology convened at a Think-Tank in Marburg, Germany, in August 2022 to review the evidence and analyze barriers for current and future developments. After this event, an overview of the current state-of-the-art was drafted and suggestions for improvements and specific recommendations for research and practice were integrated. Recommendations arising from our meeting cover further improving psychological interventions through translational approaches, improving clinical research methodology, bridging the gap between more nomothetic (group-oriented) studies and idiographic (person-centered) decisions, using network approaches in addition to selecting single mechanisms to embrace the complexity of clinical reality, making use of scalable digital options for assessments and interventions, improving the training and education of future psychotherapists, and accepting the societal responsibilities that clinical psychology has in improving national and global health care. The objective of the Marburg Declaration is to stimulate a significant change regarding our understanding of mental disorders and their treatments, with the aim to trigger a new era of evidence-based psychological interventions.