Explore the vast range of titles available.

This work is aimed at exploiting second-order circular suprasegmental hidden Markov models (CSPHMM2s) as classifiers to enhance talking condition recognition in stressful and emotional talking environments (completely two separate environments). The stressful talking environment that has been used in this work uses speech under simulated and actual stress database, while the emotional talking environment uses emotional prosody speech and transcripts database. The achieved results of this work using mel-frequency cepstral coefficients demonstrate that CSPHMM2s outperform each of hidden Markov models, second-order circular hidden Markov models, and suprasegmental hidden Markov models in enhancing talking condition recognition in the stressful and emotional talking environments. The results also show that the performance of talking condition recognition in stressful talking environments leads that in emotional talking environments by 3.67 % based on CSPHMM2s. Our results obtained in subjective evaluation by human judges fall within 2.14 and 3.08 % of those obtained, respectively, in stressful and emotional talking environments based on CSPHMM2s.

The use of visual features in audio-visual speech recognition (AVSR) is justified by both the speech generation mechanism, which is essentially bimodal in audio and visual representation, and by the need for features that are invariant to acoustic noise perturbation. As a result, current AVSR systems demonstrate significant accuracy improvements in environments affected by acoustic noise. In this paper, we describe the use of two statistical models for audio-visual integration, the coupled HMM (CHMM) and the factorial HMM (FHMM), and compare the performance of these models with the existing models used in speaker dependent audio-visual isolated word recognition. The statistical properties of both the CHMM and FHMM allow to model the state asynchrony of the audio and visual observation sequences while preserving their natural correlation over time. In our experiments, the CHMM performs best overall, outperforming all the existing models and the FHMM.

Depression is closely associated with abnormalities in brain function. Traditional static functional connectivity analyses offer limited insight into the temporal variability of brain activity. Recent advances in dynamic analyses enable a deeper understanding of how depression relates to temporal fluctuations in brain activity. This study utilized a large resting-state functional magnetic resonance imaging dataset (  = 696) to examine the association between brain dynamics and depression. Two complementary approaches were employed. Hidden Markov modeling (HMM) was used to identify discrete brain states and quantify their temporal switching patterns; temporal variability was computed within and between large-scale functional networks to capture time-varying fluctuations in functional connectivity. Depression scores were positively associated with switching rate and negatively associated with maximum fractional occupancy. Furthermore, depression scores were significantly associated with greater temporal variability both within and between networks, with particularly strong effects observed in the default mode network, ventral attention network, and frontoparietal network. Together, these findings suggest that individuals with higher depression scores exhibit more unstable brain dynamics. Our findings reveal that individuals with higher depression levels exhibit greater instability in brain state transitions and increased temporal variability in functional connectivity across large-scale networks. This instability in brain dynamics may contribute to difficulties in emotion regulation and cognitive control. By capturing whole-brain temporal patterns, this study offers a novel perspective on the neural basis of depression.

Driven by recent advances in technology, tracking devices allow to collect high-frequency data on the position of players in (association) football matches and in many other sports. Although such data sets are available to every professional team, most teams still rely on time-consuming video analysis when analysing future opponents, for example with regard to how goals were scored or a team’s general style of play. In this contribution, we provide a data-driven approach for automated classification of tactics in football. For that purpose, we consider hidden Markov models (HMMs) to analyse high-frequency tracking data, where the underlying states serve for a team’s tactic. In particular, as space control in football has been considered a major driver of success, we focus on the effective playing space, which is the convex hull created by the players excluding the goalkeeper. This quantity relates to both playing style and team behavior. Using copula-based HMMs, we model jointly the effective playing space of both teams to account for the competitive nature of the game. Our model thus provides an estimate of a team’s playing style at each time point, which can be beneficial for team managers but also of huge interest to football fans.

Fetal heart rate variability (FHRV) is a critical indicator of fetal well-being and autonomic nervous system development during labor. Traditional monitoring methods often provide limited insights, potentially leading to delayed interventions and suboptimal outcomes. This study proposes an advanced predictive analytics approach by integrating approximate entropy analysis with a hidden Markov model (HMM) within a digital twin framework to enhance real-time fetal monitoring. We utilized a dataset of 469 fetal electrocardiogram (ECG) recordings, each exceeding one hour in duration, to ensure sufficient temporal information for reliable modeling. The FHRV data were preprocessed and partitioned into parasympathetic and sympathetic components based on downward and non-downward beat detection. Approximate entropy was calculated to quantify the complexity of FHRV patterns, revealing significant correlations with umbilical cord blood gas parameters, particularly pH levels. The HMM was developed with four hidden states representing discrete pH levels and eight observed states derived from FHRV data. By employing the Baum–Welch and Viterbi algorithms for training and decoding, respectively, the model effectively captured temporal dependencies and provided early predictions of the fetal acid–base status. Experimental results demonstrated that the model achieved 85% training and 79% testing accuracy on the balanced dataset distribution, improving from 78% and 71% on the imbalanced dataset. The integration of this predictive model into a digital twin framework offers significant benefits for timely clinical interventions, potentially improving prenatal outcomes.

There is an urgent need to advance understanding of the neural underpinnings of depression, especially early in the life span. Examination of neural dynamics using resting‐state functional magnetic resonance imaging (fMRI) data can provide indices of neural flexibility, which may provide important new insights for the neurobiology of pediatric depression. Here we applied Hidden Markov Modeling (HMM) to resting‐state fMRI data to investigate neural flexibility in relation to depression and suicidal thinking in children. We utilized data from the Adolescent Brain Cognitive Development℠ Study (ABCD Study), and included data from 10,763 children (9–10 years) who completed two 5‐min resting state fMRI scans at the baseline visit. After applying the NeuroMark framework to the data, HMM was applied with a varying number of states; a six‐state model was selected from candidate models based on between‐scan reliability. We applied linear mixed‐effect modeling to test the relationship between two clinical predictors: current major depressive disorder (MDD) diagnosis and presence of suicidal ideation (SI) with our primary outcome for neural flexibility: the frequency of transitions between HMM‐derived states (“state‐switching”), while including sex, age, and other socio‐demographic variables as covariates. Analyses were conducted both with and without correction for head motion. We also explored relationships with total time and dwell time in each state of the six states. Lower state‐switching during rest was associated with both MDD and SI, although these findings were no longer significant after correcting for head motion. Notably, state‐switching was inversely related to head motion and was higher in females than males. Exploratory analysis showed that MDD was associated with shorter dwell time in one state and longer dwell time in another, suggesting altered temporal persistence of specific neural configurations. Tentative evidence supported our hypothesis that lower state‐switching in children with MDD and SI may reflect a reduction in brain flexibility, potentially contributing to a tendency to become “stuck” in negative patterns of thinking and feeling. However, the relatively low frequency of these problems in late childhood reduced statistical power after correcting for motion. Future research is needed to assess these relationships at later adolescent time points, when higher prevalence of depression and SI and lower prevalence of head motion will allow more powerful tests of these associations.

Background Pseudogenes are non-functional copies of protein coding genes that typically follow a different molecular evolutionary path as compared to functional genes. The inclusion of pseudogene sequences in DNA barcoding and metabarcoding analysis can lead to misleading results. None of the most widely used bioinformatic pipelines used to process marker gene (metabarcode) high throughput sequencing data specifically accounts for the presence of pseudogenes in protein-coding marker genes. The purpose of this study is to develop a method to screen for nuclear mitochondrial DNA segments (nuMTs) in large COI datasets. We do this by: (1) describing gene and nuMT characteristics from an artificial COI barcode dataset, (2) show the impact of two different pseudogene removal methods on perturbed community datasets with simulated nuMTs, and (3) incorporate a pseudogene filtering step in a bioinformatic pipeline that can be used to process Illumina paired-end COI metabarcode sequences. Open reading frame length and sequence bit scores from hidden Markov model (HMM) profile analysis were used to detect pseudogenes. Results Our simulations showed that it was more difficult to identify nuMTs from shorter amplicon sequences such as those typically used in metabarcoding compared with full length DNA barcodes that are used in the construction of barcode libraries. It was also more difficult to identify nuMTs in datasets where there is a high percentage of nuMTs. Existing bioinformatic pipelines used to process metabarcode sequences already remove some nuMTs, especially in the rare sequence removal step, but the addition of a pseudogene filtering step can remove up to 5% of sequences even when other filtering steps are in place. Conclusions Open reading frame length filtering alone or combined with hidden Markov model profile analysis can be used to effectively screen out apparent pseudogenes from large datasets. There is more to learn from COI nuMTs such as their frequency in DNA barcoding and metabarcoding studies, their taxonomic distribution, and evolution. Thus, we encourage the submission of verified COI nuMTs to public databases to facilitate future studies.

Epidemic waves in large metropolitan areas unfold heterogeneously across territories shaped by persistent socioeconomic inequalities. Explaining how transmission intensifies, stabilises, and shifts across the urban landscape remains a central challenge in epidemiology. This study develops a covariate-dependent, non-homogeneous Hidden Markov Model (nHMM) to infer latent transmission regimes from municipality-level COVID-19 incidence in Santiago, Chile. The framework links daily case dynamics to mobility flows and structural socioeconomic indicators within a hierarchical specification that captures inter-municipal heterogeneity. Model selection identifies three statistically distinct and epidemiologically interpretable regimes corresponding to moderate, severe, and critical transmission phases. Transition dynamics reveal marked spatial asymmetries: while increases in mobility consistently elevate escalation risk, structural conditions—such as overcrowding and deficits in urban infrastructure—substantially influence both the probability of entering and the persistence within high-severity regimes. To ensure epidemiological interpretability, regime-conditioned incidence trajectories are mapped to the time-varying reproduction number ( ) through a renewal formulation, enabling coherent propagation of uncertainty from latent-state inference to transmission estimates. By integrating latent regime inference with hierarchical transition modelling and renewal-based transmission analysis, this study distinguishes structural phase shifts from stochastic variability in urban epidemic dynamics. The findings clarify how mobility and entrenched inequality jointly structure transmission risk, providing a scalable and transferable framework for monitoring and anticipating epidemic regime transitions in complex metropolitan systems.

Brain-related disorders are characterized by observable behavioral symptoms, for example, social withdrawal. Smartphones can passively collect behavioral data reflecting digital activities such as communication app usage and calls. These data are collected objectively in real time, avoiding recall bias, and may, therefore, be a useful tool for measuring behaviors related to social functioning. Despite promising clinical utility, analyzing smartphone data is challenging as datasets often include a range of temporal features prone to missingness. Hidden Markov models (HMMs) provide interpretable, lower-dimensional temporal representations of data, allowing for missingness. This study aimed to investigate the HMM as a method for modeling smartphone time series data. We applied an HMM to an aggregate dataset of smartphone measures designed to assess phone-related social functioning in healthy controls (HCs) and participants with schizophrenia, Alzheimer disease (AD), and memory complaints. We trained the HMM on a subset of HCs (91/348, 26.1%) and selected a model with socially active and inactive states. Then, we generated hidden state sequences per participant and calculated their \"total dwell time,\" that is, the percentage of time spent in the socially active state. Linear regression models were used to compare the total dwell time to social and clinical measures in a subset of participants with available measures, and logistic regression was used to compare total dwell times between diagnostic groups and HCs. We primarily reported results from a 2-state HMM but also verified results in HMMs with more hidden states and trained on the whole participant dataset. We identified lower total dwell times in participants with AD (26/257, 10.1%) versus withheld HCs (156/257, 60.7%; odds ratio 0.95, 95% CI 0.92-0.97; false discovery rate [FDR]-corrected P<.001), as well as in participants with memory complaints (57/257, 22.2%; odds ratio 0.97, 95% CI 0.96-0.99; FDR-corrected P=.004). The result in the AD group was very robust across HMM variations, whereas the result in the memory complaints group was less robust. We also observed an interaction between the AD group and total dwell time when predicting social functioning (FDR-corrected P=.02). No significant relationships regarding total dwell time were identified for participants with schizophrenia (18/257, 7%; P>.99). We found the HMM to be a practical, interpretable method for digital phenotyping analysis, providing an objective phenotype that is a possible indicator of social functioning.

Background Approximating the recent phylogeny of N phased haplotypes at a set of variants along the genome is a core problem in modern population genomics and central to performing genome-wide screens for association, selection, introgression, and other signals. The Li & Stephens (LS) model provides a simple yet powerful hidden Markov model for inferring the recent ancestry at a given variant, represented as an N × N distance matrix based on posterior decodings. Results We provide a high-performance engine to make these posterior decodings readily accessible with minimal pre-processing via an easy to use package kalis, in the statistical programming language R . kalis enables investigators to rapidly resolve the ancestry at loci of interest and developers to build a range of variant-specific ancestral inference pipelines on top. kalis exploits both multi-core parallelism and modern CPU vector instruction sets to enable scaling to hundreds of thousands of genomes. Conclusions The resulting distance matrices accessible via kalis enable local ancestry, selection, and association studies in modern large scale genomic datasets.