Analyzing asymmetry in brain hierarchies with a linear state-space model of resting-state fMRI data.

This study challenges the traditional focus on zero-lag statistics in resting-state functional magnetic resonance imaging (rsfMRI) research. Instead, it advocates for considering time-lag interactions to unveil the directionality and asymmetries of the brain hierarchy. Effective connectivity (EC), the state matrix in dynamical causal modeling (DCM), is a commonly used metric for studying dynamical properties and causal interactions within a linear state-space system description. Here, we focused on how time-lag statistics are incorporated within the framework of DCM resulting in an asymmetric EC matrix. Our approach involves decomposing the EC matrix, revealing a steady-state differential cross-covariance matrix that is responsible for modeling information flow and introducing time-irreversibility. Specifically, the system's dynamics, influenced by the off-diagonal part of the differential covariance, exhibit a curl steady-state flow component that breaks detailed balance and diverges the dynamics from equilibrium. Our empirical findings indicate that the EC matrix's outgoing strengths correlate with the flow described by the differential cross covariance, while incoming strengths are primarily driven by zero-lag covariance, emphasizing conditional independence over directionality.

Modeling large-scale brain dynamics offers insight into the main principles of brain self-organization. In particular, the identification of traces of nonequilibrium steady-state dynamics also at the mascroscale level has been recently linked to the presence of intrinsic brain networks. Quantifying these aspects is generally limited by numerical difficulties. However, for resting-state BOLD data, a linear stochastic state-space model has demonstrated efficacy, simplifying analysis. Specifically, the asymmetric structure of effective connectivity, that is, the state interaction matrix, directly reflects nonequilibrium steady-state dynamics and time-irreversibility. By disentangling this asymmetry, we quantified departure from equilibrium and discerned primary directions of information propagation, identifying brain regions as sources or sinks.

An information-theoretic framework for conditional causality analysis of brain networks.

Identifying directed network models for multivariate time series is a ubiquitous problem in data science. Granger causality measure (GCM) and conditional GCM (cGCM) are widely used methods for identifying directed connections between time series. Both GCM and cGCM have frequency-domain formulations to characterize the dependence of time series in the spectral domain. However, the original methods were developed using a heuristic approach without rigorous theoretical explanations. To overcome the limitation, the minimum-entropy (ME) estimation approach was introduced in our previous work (Ning & Rathi, 2018) to generalize GCM and cGCM with more rigorous frequency-domain formulations. In this work, this information-theoretic framework is further generalized with three formulations for conditional causality analysis using techniques in control theory, such as state-space representations and spectral factorizations. The three conditional causal measures are developed based on different ME estimation procedures that are motivated by equivalent formulations of the classical minimum mean squared error estimation method. The relationship between the three formulations of conditional causality measures is analyzed theoretically. Their performance is evaluated using simulations and real neuroimaging data to analyze brain networks. The results show that the proposed methods provide more accurate network structures than the original approach.

This paper introduces a theoretical framework for causal inference in brain networks using time series measurements based on the principle of minimum-entropy regression. Three types of conditional causality measures are derived based on varying formulations of minimum-entropy regressions. The standard time-domain conditional Granger causality measure is formulated as a special case but with a different expression of the frequency-domain measure. The methods were evaluated using simulations and real resting-state functional MRI data of human brains and compared with standard Granger causality measures and directed transfer functions. Two new formulations of minimum-entropy-based causality measures showed better performance than other methods. The algorithms developed from this work may provide new insights to understand information flow in brain networks.

Estimating the impact of physician risky-prescribing on the network structure underlying physician shared-patient relationships.

Social network analysis and shared-patient physician networks have become effective ways of studying physician collaborations. Assortative mixing or "homophily" is the network phenomenon whereby the propensity for similar individuals to form ties is greater than for dissimilar individuals. Motivated by the public health concern of risky-prescribing among older patients in the United States, we develop network models and tests involving novel network measures to study whether there is evidence of homophily in prescribing and deprescribing in the specific shared-patient network of physicians linked to the US state of Ohio in 2014. Evidence of homophily in risky-prescribing would imply that prescribing behaviors help shape physician networks and would suggest strategies for interventions seeking to reduce risky-prescribing (e.g., strategies to directly reduce risky prescribing might be most effective if applied as group interventions to risky prescribing physicians connected through the network and the connections between these physicians could be targeted by tie dissolution interventions as an indirect way of reducing risky prescribing). Furthermore, if such effects varied depending on the structural features of a physician's position in the network (e.g., by whether or not they are involved in cliques-groups of actors that are fully connected to each other-such as closed triangles in the case of three actors), this would further strengthen the case for targeting groups of physicians involved in risky prescribing and the network connections between them for interventions. Using accompanying Medicare Part D data, we converted patient longitudinal prescription receipts into novel measures of the intensity of each physician's risky-prescribing. Exponential random graph models were used to simultaneously estimate the importance of homophily in prescribing and deprescribing in the network beyond the characteristics of physician specialty (or other metadata) and network-derived features. In addition, novel network measures were introduced to allow homophily to be characterized in relation to specific triadic (three-actor) structural configurations in the network with associated non-parametric randomization tests to evaluate their statistical significance in the network against the null hypothesis of no such phenomena. We found physician homophily in prescribing and deprescribing. We also found that physicians exhibited within-triad homophily in risky-prescribing, with the prevalence of homophilic triads significantly higher than expected by chance absent homophily. These results may explain why communities of prescribers emerge and evolve, helping to justify group-level prescriber interventions. The methodology may be applied, adapted or generalized to study homophily and its generalizations on other network and attribute combinations involving analogous shared-patient networks and more generally using other kinds of network data underlying other kinds of social phenomena.

SEMIPARAMETRIC BIVARIATE HIERARCHICAL STATE SPACE MODEL WITH APPLICATION TO HORMONE CIRCADIAN RELATIONSHIP.

The adrenocorticotropic hormone and cortisol play critical roles in stress regulation and the sleep-wake cycle. Most research has been focused on how the two hormones regulate each other in terms of short-term pulses. Few studies have been conducted on the circadian relationship between the two hormones and how it differs between normal and abnormal groups. The circadian patterns are difficult to model as parametric functions. Directly extending univariate functional mixed effects models would result in a large dimensional problem and a challenging nonparametric inference. In this article, we propose a semi-parametric bivariate hierarchical state space model, in which each hormone profile is modeled by a hierarchical state space model, with nonparametric population-average and subject-specific components. The bivariate relationship is constructed by concatenating two latent independent subject-specific random functions specified by a design matrix, leading to a parametric inference on the correlation. We propose a computationally efficient state-space EM algorithm for estimation and inference. We apply the proposed method to a study of chronic fatigue syndrome and fibromyalgia and discover an erratic regulation pattern in the patient group in contrast to a circadian regulation pattern conforming to the day-night cycle in the control group.

Learning long sequences in spiking neural networks.

Spiking neural networks (SNNs) take inspiration from the brain to enable energy-efficient computations. Since the advent of Transformers, SNNs have struggled to compete with artificial networks on modern sequential tasks, as they inherit limitations from recurrent neural networks (RNNs), with the added challenge of training with non-differentiable binary spiking activations. However, a recent renewed interest in efficient alternatives to Transformers has given rise to state-of-the-art recurrent architectures named state space models (SSMs). This work systematically investigates, for the first time, the intersection of state-of-the-art SSMs with SNNs for long-range sequence modelling. Results suggest that SSM-based SNNs can outperform the Transformer on all tasks of a well-established long-range sequence modelling benchmark. It is also shown that SSM-based SNNs can outperform current state-of-the-art SNNs with fewer parameters on sequential image classification. Finally, a novel feature mixing layer is introduced, improving SNN accuracy while challenging assumptions about the role of binary activations in SNNs. This work paves the way for deploying powerful SSM-based architectures, such as large language models, to neuromorphic hardware for energy-efficient long-range sequence modelling.

Insights into Blue Whale (Balaenoptera musculus L.) Population Movements in the Galapagos Archipelago and Southeast Pacific.

The Galapagos Marine Reserve is vital for cetaceans, serving as both a stopover and residency site. However, blue whales, occasionally sighted here, exhibit poorly understood migratory behavior within the Galapagos and the broader Eastern Tropical Pacific. This study, the first to satellite tag blue whales in the Galapagos (16 tagged between 2021 and 2023), explored their behavior in relation to environmental variables like chlorophyll-a concentration, sea surface temperature (SST), and productivity. Key findings show a strong correlation between foraging behavior, high chlorophyll-a levels, productivity, and lower SSTs, indicating a preference for food-rich areas. Additionally, there is a notable association with geomorphic features like ridges, which potentially enhance food abundance. Most tagged whales stayed near the Galapagos archipelago, with higher concentrations observed around Isabela Island, which is increasingly frequented by tourist vessels, posing heightened ship strike risks. Some whales ventured into Ecuador's exclusive economic zone, while one migrated southward to Peru. The strong 2023 El Niño-Southern Oscillation event led to SST and primary production changes, likely impacting whale resource availability. Our study provides crucial insights into blue whale habitat utilization, informing adaptive management strategies to mitigate ship strike risks and address altered migration routes due to climate-driven environmental shifts.

The efficiency of detecting seabird behaviour from movement patterns: the effect of sampling frequency on inferring movement metrics in Procellariiformes.

BACKGROUND: Recent technological advances have resulted in low-cost GPS loggers that are small enough to be used on a range of seabirds, producing accurate location estimates (± 5 m) at sampling intervals as low as 1 s. However, tradeoffs between battery life and sampling frequency result in studies using GPS loggers on flying seabirds yielding locational data at a wide range of sampling intervals. Metrics derived from these data are known to be scale-sensitive, but quantification of these errors is rarely available. Very frequent sampling, coupled with limited movement, can result in measurement error, overestimating movement, but a much more pervasive problem results from sampling at long intervals, which grossly underestimates path lengths. METHODS: We use fine-scale (1 Hz) GPS data from a range of albatrosses and petrels to study the effect of sampling interval on metrics derived from the data. The GPS paths were sub-sampled at increasing intervals to show the effect on path length (i.e. ground speed), turning angles, total distance travelled, as well as inferred behavioural states. RESULTS: We show that distances (and per implication ground speeds) are overestimated (4% on average, but up to 20%) at the shortest sampling intervals (1-5 s) and underestimated at longer intervals. The latter bias is greater for more sinuous flights (underestimated by on average 40% when sampling > 1-min intervals) as opposed to straight flight (11%). Although sample sizes were modest, the effect of the bias seemingly varied with species, where species with more sinuous flight modes had larger bias. Sampling intervals also played a large role when inferring behavioural states from path length and turning angles. CONCLUSIONS: Location estimates from low-cost GPS loggers are appropriate to study the large-scale movements of seabirds when using coarse sampling intervals, but actual flight distances are underestimated. When inferring behavioural states from path lengths and turning angles, moderate sampling intervals (10-30 min) may provide more stable models, but the accuracy of the inferred behavioural states will depend on the time period associated with specific behaviours. Sampling rates have to be considered when comparing behaviours derived using varying sampling intervals and the use of bias-informed analyses are encouraged.

A new visual State Space Model for low-dose CT denoising.

BACKGROUND: Low-dose computed tomography (LDCT) can mitigate potential health risks to the public. However, the severe noise and artifacts in LDCT images can impede subsequent clinical diagnosis and analysis. Convolutional neural networks (CNNs) and Transformers stand out as the two most popular backbones in LDCT denoising. Nonetheless, CNNs suffer from a lack of long-range modeling capabilities, while Transformers are hindered by high computational complexity. PURPOSE: In this study, our main goal is to develop a simple and efficient model that can both focus on local spatial context and model long-range dependencies with linear computational complexity for LDCT denoising. METHODS: In this study, we make the first attempt to apply the State Space Model to LDCT denoising and propose a novel LDCT denoising model named Visual Mamba Encoder-Decoder Network (ViMEDnet). To efficiently and effectively capture both the local and global features, we propose the Mixed State Space Module (MSSM), where the depth-wise convolution, max-pooling, and 2D Selective Scan Module (2DSSM) are coupled together through a partial channel splitting mechanism. 2DSSM is capable of capturing global information with linear computational complexity, while convolution and max-pooling can effectively learn local signals to facilitate detail restoration. Furthermore, the network uses a weighted gradient-sensitive hybrid loss function to facilitate the preservation of image details, improving the overall denoising performance. RESULTS: The performance of our proposed ViMEDnet is compared to five state-of-the-art LDCT denoising methods, including an iterative algorithm, two CNN-based methods, and two Transformer-based methods. The comparative experimental results demonstrate that the proposed ViMEDnet can achieve better visual quality and quantitative assessment outcomes. In visual evaluation, ViMEDnet effectively removes noise and artifacts, while exhibiting superior performance in restoring fine structures and low-contrast structural edges, resulting in minimal deviation of denoised images from NDCT. In quantitative assessment, ViMEDnet obtains the lowest RMSE and the highest PSNR, SSIM, and FSIM scores, further substantiating the superiority of ViMEDnet. CONCLUSIONS: The proposed ViMEDnet possesses excellent LDCT denoising performance and provides a new alternative to LDCT denoising models beyond the existing CNN and Transformer options.

Scale-dependent population drivers inform avian management in a declining saline lake ecosystem.

Shrinking saline lakes provide irreplaceable habitat for waterbird species globally. Disentangling the effects of wetland habitat loss from other drivers of waterbird population dynamics is critical for protecting these species in the face of unprecedented changes to saline lake ecosystems, ideally through decision-making frameworks that identify effective management options and their potential outcomes. Here, we develop a framework to assess the effects of hypothesized population drivers and identify potential future outcomes of plausible management scenarios on a saline lake-reliant waterbird species. We use 36 years of monitoring data to quantify the effects of environmental conditions on the population size of a regionally important breeding colony of American white pelicans (Pelecanus erythrorhynchos) at Great Salt Lake, Utah, US, then forecast colony abundance under various management scenarios. We found that low lake levels, which allow terrestrial predators access to the colony, are probable drivers of recent colony declines. Without local management efforts, we predicted colony abundance could likely decline approximately 37.3% by 2040, although recent colony observations suggest population declines may be more extreme than predicted. Results from our population projection scenarios suggested that proactive approaches to preventing predator colony access and reversing saline lake declines are crucial for the persistence of the Great Salt Lake pelican colony. Increasing wetland habitat and preventing predator access to the colony together provided the most effective protection, increasing abundance 145.4% above projections where no management actions are taken, according to our population projection scenarios. Given the importance of water levels to the persistence of island-nesting colonial species, proactive approaches to reversing saline lake declines could likely benefit pelicans as well as other avian species reliant on these unique ecosystems.

Multi-Cluster Approaches to Radio Sensor Array Channel Modeling and Simulation.

In this paper, we explore the physical propagation environment of radio waves by describing it in terms of distant scattering clusters. Each cluster consists of numerous scattering objects that may exhibit certain statistical properties. By utilizing geometry-based methods, we can study the channel second-order statistics (CSOS), where each distant scattering cluster corresponds to a CSOS, contributes a portion to the Doppler spectrum, and is associated with a state-space multiple-input and multiple-output (MIMO) radio channel model. Consequently, the physical propagation environment of radio waves can be modeled by summing multiple state-space MIMO radio channel models. This approach offers three key advantages: simplicity, the ability to construct the entire Doppler power spectrum from multiple uncorrelated distant scattering clusters, and the capability to obtain the channels contributed by these clusters by summing the individual channels. This methodology enables the reconstruction of the radio wave propagation environment in a simulated manner and is crucial for developing massive MIMO channel models.

Neural Memory State Space Models for Medical Image Segmentation.

With the rapid advancement of deep learning, computer-aided diagnosis and treatment have become crucial in medicine. UNet is a widely used architecture for medical image segmentation, and various methods for improving UNet have been extensively explored. One popular approach is incorporating transformers, though their quadratic computational complexity poses challenges. Recently, State-Space Models (SSMs), exemplified by Mamba, have gained significant attention as a promising alternative due to their linear computational complexity. Another approach, neural memory Ordinary Differential Equations (nmODEs), exhibits similar principles and achieves good results. In this paper, we explore the respective strengths and weaknesses of nmODEs and SSMs and propose a novel architecture, the nmSSM decoder, which combines the advantages of both approaches. This architecture possesses powerful nonlinear representation capabilities while retaining the ability to preserve input and process global information. We construct nmSSM-UNet using the nmSSM decoder and conduct comprehensive experiments on the PH2, ISIC2018, and BU-COCO datasets to validate its effectiveness in medical image segmentation. The results demonstrate the promising application value of nmSSM-UNet. Additionally, we conducted ablation experiments to verify the effectiveness of our proposed improvements on SSMs and nmODEs.

The Impacts of COVID-19 Lockdowns on Road Transport Air Pollution in London: A State-Space Modelling Approach.

The emergence of the COVID-19 pandemic in 2020 led to the implementation of legal restrictions on individual activities, significantly impacting traffic and air pollution levels in urban areas. This study employs a state-space intervention method to investigate the effects of three major COVID-19 lockdowns in March 2020, November 2020, and January 2021 on London's air quality. Data were collected from 20 monitoring stations across London (central, ultra-low emission zone, and greater London), with daily measurements of NOx, PM10, and PM2.5 for four years (January 2019-December 2022). Furthermore, the developed model was adjusted for seasonal effects, ambient temperature, and relative humidity. This study found significant reductions in the NOx levels during the first lockdown: 49% in central London, 33% in the ultra-low emission zone (ULEZ), and 37% in greater London. Although reductions in NOx were also observed during the second and third lockdowns, they were less than the first lockdown. In contrast, PM10 and PM2.5 increased by 12% and 1%, respectively, during the first lockdown, possibly due to higher residential energy consumption. However, during the second lockdown, PM10 and PM2.5 levels decreased by 11% and 13%, respectively, and remained unchanged during the third lockdown. These findings highlight the complex dynamics of urban air quality and underscore the need for targeted interventions to address specific pollution sources, particularly those related to road transport. The study provides valuable insights into the effectiveness of lockdown measures and informs future air quality management strategies.

Teachers' Heart Rate Variability and Behavioral Reactions in Aggressive Interactions: Teachers Can Downregulate Their Physiological Arousal, and Progesterone Favors Social Integrative Teacher Responses.

Aggressive student behavior is considered one of the main risk factors for teacher stress. The present study investigated teachers' physiological and behavioral reactions when facing aggressive student behavior and examined which resources favor adaptive teacher reactions. The sample included 42 teachers. We assessed (a) teacher self-reports (i.e., resources, risk factors, and vital exhaustion) (b) classroom observations, (c) ambulatory assessments of teachers' heart rate and heart rate variability, and (d) teachers' progesterone concentrations in the hair. The present study focused on a subsample of ten teachers (9 females, Mage = 34.70, SD = 11.32) managing classes which were potentially very stressful as they had a high density of aggressive behavior. High levels of work satisfaction, hair progesterone, and a low level of work overload fostered social integrative teacher responses. Moreover, in 75% of the cases, teachers succeeded in downregulating their physiological reaction. Our results support the notion that teachers evaluate stressors in light of their resources. When they perceive their resources as insufficient for coping with a challenging situation, stress arises, and subsequently, they react inefficiently to aggressive behavior. Thus, teacher education could benefit from strengthening teacher resources and strategies for coping with aggressive student behavior.

A dynamic generative model can extract interpretable oscillatory components from multichannel neurophysiological recordings.

Modern neurophysiological recordings are performed using multichannel sensor arrays that are able to record activity in an increasingly high number of channels numbering in the 100s to 1000s. Often, underlying lower-dimensional patterns of activity are responsible for the observed dynamics, but these representations are difficult to reliably identify using existing methods that attempt to summarize multivariate relationships in a post hoc manner from univariate analyses or using current blind source separation methods. While such methods can reveal appealing patterns of activity, determining the number of components to include, assessing their statistical significance, and interpreting them requires extensive manual intervention and subjective judgment in practice. These difficulties with component selection and interpretation occur in large part because these methods lack a generative model for the underlying spatio-temporal dynamics. Here, we describe a novel component analysis method anchored by a generative model where each source is described by a bio-physically inspired state-space representation. The parameters governing this representation readily capture the oscillatory temporal dynamics of the components, so we refer to it as oscillation component analysis. These parameters - the oscillatory properties, the component mixing weights at the sensors, and the number of oscillations - all are inferred in a data-driven fashion within a Bayesian framework employing an instance of the expectation maximization algorithm. We analyze high-dimensional electroencephalography and magnetoencephalography recordings from human studies to illustrate the potential utility of this method for neuroscience data.

AI-enhanced analysis of naturalistic social interactions characterizes interaffective impairments in schizophrenia.

Social deficits in schizophrenia have been attributed to an impaired attunement to mutual interaction, or "interaffectivity". While impairments in emotion recognition and facial expressivity in schizophrenia have been consistently reported, findings on mimicry and social synchrony are inconsistent, and previous studies have often lacked ecological validity. To investigate interaffective behavior in dyadic interactions in a real-world-like setting, 20 individuals with schizophrenia and 20 without mental disorder played a cooperative board game with a previously unacquainted healthy control participant. Facial expression analysis was conducted using Affectiva Emotion AI in iMotions 9.3. The contingency and state space distribution of emotional facial expressions was assessed using Mangold INTERACT. Psychotic symptoms, subjective stress, affectivity and game experience were evaluated through questionnaires. Due to a considerable between-group age difference, age-adjusted ANCOVA was performed. Overall, despite an unchanged subjective experience of the social interaction, individuals with schizophrenia exhibited reduced responsiveness to positive affective stimuli. Subjective game experience did not differ between groups. Descriptively, facial expressions in schizophrenia were generally more negative, with increased sadness and decreased joy. Facial mimicry was impaired specifically regarding joyful expressions in schizophrenia, which correlated with blunted affect as measured by the SANS. Dyadic interactions involving persons with schizophrenia were less attracted toward mutual joyful affective states. Only unadjusted for age, in the absence of emotional stimuli from their interaction partner, individuals with schizophrenia showed more angry and sad expressions. These impairments in interaffective processes may contribute to social dysfunction in schizophrenia and provide new avenues for future research.

Unraveling the impact of therapeutic drug monitoring via machine learning for patients with sepsis.

Clinical studies investigating the benefits of beta-lactam therapeutic drug monitoring (TDM) among critically ill patients are hindered by small patient groups, variability between studies, patient heterogeneity, and inadequate use of TDM. Accordingly, definitive conclusions regarding the efficacy of TDM remain elusive. To address these challenges, we propose an innovative approach that leverages data-driven methods to unveil the concealed connections between therapy effectiveness and patient data, collected through a randomized controlled trial (DRKS00011159; 10th October 2016). Our findings reveal that machine learning algorithms can successfully identify informative features that distinguish between healthy and sick states. These hold promise as potential markers for disease classification and severity stratification, as well as offering a continuous and data-driven "multidimensional" Sequential Organ Failure Assessment (SOFA) score. The positive impact of TDM on patient recovery rates is demonstrated by unraveling the intricate connections between therapy effectiveness and clinically relevant data via machine learning.

State Space Representation of Jiles-Atherton Hysteresis Model and Application for Closed-Loop Control.

Hysteresis is a fundamental characteristic of magnetic materials. The Jiles-Atherton (J-A) hysteresis model, which is known for its few parameters and clear physical interpretations, has been widely employed in simulating hysteresis characteristics. To better analyze and compute hysteresis behavior, this study established a state space representation based on the primitive J-A model. First, based on the five fundamental equations of the J-A model, a state space representation was established through variable substitution and simplification. Furthermore, to address the singularity problem at zero crossings, local linearization was obtained through an approximation method based on the actual physical properties. Based on these, the state space model was implemented using the S-function. To validate the effectiveness of the state space model, the hysteresis loops were obtained through COMSOL finite element software and tested on a permalloy toroidal sample. The particle swarm optimization (PSO) method was used for parameter identification of the state space model, and the identification results show excellent agreement with the simulation and test results. Finally, a closed-loop control system was constructed based on the state space model, and trajectory tracking experiments were conducted. The results verify the feasibility of the state space representation of the J-A model, which holds significant practical implications in the development of magnetically shielded rooms, the suppression of magnetic interference in cold atom clocks, and various other applications.

APRICOT-Mamba: Acuity Prediction in Intensive Care Unit (ICU): Development and Validation of a Stability, Transitions, and Life-Sustaining Therapies Prediction Model.

On average, more than 5 million patients are admitted to intensive care units (ICUs) in the US, with mortality rates ranging from 10 to 29%. The acuity state of patients in the ICU can quickly change from stable to unstable, sometimes leading to life-threatening conditions. Early detection of deteriorating conditions can assist in more timely interventions and improved survival rates. While Artificial Intelligence (AI)-based models show potential for assessing acuity in a more granular and automated manner, they typically use mortality as a proxy of acuity in the ICU. Furthermore, these methods do not determine the acuity state of a patient (i.e., stable or unstable), the transition between acuity states, or the need for life-sustaining therapies. In this study, we propose APRICOT-M (Acuity Prediction in Intensive Care Unit-Mamba), a 1M-parameter state space-based neural network to predict acuity state, transitions, and the need for life-sustaining therapies in real-time among ICU patients. The model integrates ICU data in the preceding four hours (including vital signs, laboratory results, assessment scores, and medications) and patient characteristics (age, sex, race, and comorbidities) to predict the acuity outcomes in the next four hours. Our state space-based model can process sparse and irregularly sampled data without manual imputation, thus reducing the noise in input data and increasing inference speed. The model was trained on data from 107,473 patients (142,062 ICU admissions) from 55 hospitals between 2014-2017 and validated externally on data from 74,901 patients (101,356 ICU admissions) from 143 hospitals. Additionally, it was validated temporally on data from 12,927 patients (15,940 ICU admissions) from one hospital in 2018-2019 and prospectively on data from 215 patients (369 ICU admissions) from one hospital in 2021-2023. Three datasets were used for training and evaluation: the University of Florida Health (UFH) dataset, the electronic ICU Collaborative Research Database (eICU), and the Medical Information Mart for Intensive Care (MIMIC)-IV dataset. APRICOT-M significantly outperforms the baseline acuity assessment, Sequential Organ Failure Assessment (SOFA), for mortality prediction in both external (AUROC 0.95 CI: 0.94-0.95 compared to 0.78 CI: 0.78-0.79) and prospective (AUROC 0.99 CI: 0.97-1.00 compared to 0.80 CI: 0.65-0.92) cohorts, as well as for instability prediction (external AUROC 0.75 CI: 0.74-0.75 compared to 0.51 CI: 0.51-0.51, and prospective AUROC 0.69 CI: 0.64-0.74 compared to 0.53 CI: 0.50-0.57). This tool has the potential to help clinicians make timely interventions by predicting the transition between acuity states and decision-making on life-sustaining within the next four hours in the ICU.

Bayesian Inference of Hidden Cognitive Performance and Arousal States in Presence of Music.

Goal: Poor arousal management may lead to reduced cognitive performance. Specifying a model and decoder to infer the cognitive arousal and performance contributes to arousal regulation via non-invasive actuators such as music. Methods: We employ a Bayesian filtering approach within an expectation-maximization framework to track the hidden states during the [Formula: see text]-back task in the presence of calming and exciting music. We decode the arousal and performance states from the skin conductance and behavioral signals, respectively. We derive an arousal-performance model based on the Yerkes-Dodson law. We design a performance-based arousal decoder by considering the corresponding performance and skin conductance as the observation. Results: The quantified arousal and performance are presented. The existence of Yerkes-Dodson law can be interpreted from the arousal-performance relationship. Findings display higher matrices of performance within the exciting music. Conclusions: The performance-based arousal decoder has a better agreement with the Yerkes-Dodson law. Our study can be implemented in designing non-invasive closed-loop systems.

Circadian and environmental signal integration in a natural population of Arabidopsis.

Plants sense and respond to environmental cues during 24 h fluctuations in their environment. This requires the integration of internal cues such as circadian timing with environmental cues such as light and temperature to elicit cellular responses through signal transduction. However, the integration and transduction of circadian and environmental signals by plants growing in natural environments remains poorly understood. To gain insights into 24 h dynamics of environmental signaling in nature, we performed a field study of signal transduction from the nucleus to chloroplasts in a natural population of Arabidopsis halleri. Using several modeling approaches to interpret the data, we identified that the circadian clock and temperature are key regulators of this pathway under natural conditions. We identified potential time-delay steps between pathway components, and diel fluctuations in the response of the pathway to temperature cues that are reminiscent of the process of circadian gating. We found that our modeling framework can be extended to other signaling pathways that undergo diel oscillations and respond to environmental cues. This approach of combining studies of gene expression in the field with modeling allowed us to identify the dynamic integration and transduction of environmental cues, in plant cells, under naturally fluctuating diel cycles.