Pavement damage is characterized by concealment and rapid deterioration, posing a serious threat to road traffic quality and safety. This paper proposes an improved pavement defect detection model named SHE-YOLO, which is deployed on the Kendryte K230 embedded platform, enabling a low-cost real-time pavement defect detection system. SHE-YOLO introduces a semantic-aware hypergraph convolution module based on YOLOv11n, enhancing feature association for small targets by integrating a domain-knowledge-driven semantic similarity matrix with an area-adaptive hyperedge attention mechanism. Combined with multi-scale edge enhancement techniques, this model achieves improved detection accuracy for fine cracks and small-scale pavement damages. Experimental validation demonstrates that the proposed model achieves a detection accuracy of 90.96±0.06
We study hyperchaos and extreme events observed in the minimal ensemble of three spiking Rulkov neurons with mutual chemical synaptic couplings earlier proposed in [1]. We discuss the relation between types of chaotic dynamics and extreme events in the system under study. Bifurcation scenarios of the emergence and destruction of complex dynamics containing extreme events were studied in detail. We hope that our study can help to gain more insights into the role of different types of couplings on the appearance and destruction of extreme events in neural networks.
Multi-parameter mapping (MPM) magnetic resonance imaging (MRI) provides parameters sensitive to myelin, iron and water. Conventional analyses treat these parameters individually or via pairwise correlations. We introduce O-information ( Ω ) as a proof-of-principle higher-order interaction framework to quantify how the three myelin-sensitive parameters (magnetisation transfer, longitudinal relaxation rate R_1 and proton density) are interrelated beyond pairwise dependencies. We compute Ω from their joint distribution across cortical grey matter, subcortical grey matter and white-matter bundles and ask how Ω is modulated by the iron-sensitive transverse relaxation rate R_2^* in grey matter and by fibre architecture (neurite density and orientation dispersion) in white matter. In 22 healthy adults, Ω separates tissue classes (cortex near-balanced, subcortex mildly synergistic, white matter strongly synergistic), consistent in 21 of 22 participants. At matched voxel count per region, grey-matter Ω declines with R_2^* and changes sign from redundancy to synergy within the physiological R_2^* range, an effect carried by cortex. A minimal two-compartment forward model [1] ( R_2^*∝[Fe] , following Langkammer et al. 257:455–462, 2010 ) generates the same sign change using constants taken from the literature, but places it at systematically lower R_2^* . Fitting a single effective iron- R_1 coupling closes the gap, at a value below the literature estimate. In white matter, Ω is only weakly coupled to R_2^* ( r=+0.007 , p=0.82 ), well below the grey-matter coupling. Neurite density, not orientation dispersion, is the dominant inter-bundle predictor. Because NODDI and Ω are both derived in part from shared tissue properties, the association does not establish an independent causal link. O-information thus quantifies how myelin-sensitive parameters combine or diverge across tissue compartments, providing a higher-order axis not captured by pairwise analyses.
The HADES experiment, located at the GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt, investigates the equation of state (EoS) of Quantum Chromodynamics (QCD) matter created in central collisions of heavy- and medium-mass nuclei at beam energies of a few GeV, employing both electromagnetic and hadronic probes. In addition, the experiment maintains dedicated programs aimed at the study of hadron properties in p+p and pion-induced reactions. A particular focus of this report is the reconstruction of higher-order cumulant ratios of proton and light-nuclei multiplicity distributions in Ag+Ag collisions, which constitute sensitive probes of critical phenomena due to their direct connection, within the framework of statistical mechanics, to the EoS. Event-by-event (E-by-E) fluctuations are reconstructed using a novel probabilistic approach based on Fuzzy Logic, which effectively circumvents the limitations imposed by incomplete E-by-E particle identification in experimental data. Several strategies for efficiency correction are discussed, and a data-driven event-mixing technique is employed to account for volume (centrality) fluctuations. The fully corrected normalized factorial cumulants of the proton multiplicity distribution, including their acceptance dependence in rapidity, are presented and compared with canonical ensemble baseline calculations that incorporate correlations arising from local attractive interactions. The HADES results extend the trends observed by the STAR experiment toward lower collision energies for the measured cumulant ratios.
A delay line is a fundamental building block for time-based effects in classical signal processing. Preliminary quantum signal-processing circuits found in the literature implement signal delays through quantum time-shifting operations. In this work, we present an optimization that reduces the qubit and gate counts required by these existing quantum delay circuits. The paper begins by reviewing the classical concept of a delay line, comparing it with its quantum counterpart, and discussing previous work on quantum delay operations. The methodology section then introduces the quantum circuits used to implement both the original and optimized versions of the delay operation. The final sections provide a complexity analysis of the different circuits, present simulation examples in which the operation is applied to signals, and discuss possible extensions of the proposed approach. These include scaling the circuits to larger numbers of audio samples, incorporating signal mixing after the time-shifting modules to obtain a complete delay effect in the classical sense, and exploring their creative potential for audio and music applications.
This special issue presents a collection of original research articles on the preparation, characterization, and application of photo-thermo-electric functional materials, with a particular emphasis on the role of spatial charge modulation. Electrons and holes, as the primary charge carriers, govern numerous physicochemical processes fundamental to optoelectronic, thermoelectric, and photo-thermo-electric functional materials. The 13 contributions in this issue address recent advances across multiple scales—from macroscopic photovoltaic manipulation and triboelectric hybrid systems to microscopic control of electron–hole pairs for enhanced light absorption, emission, and photocatalytic efficiency. Topics covered include surface charge modulation for micro-/nano-manipulation, plasmonic sensing platforms, perovskite quantum dots, photocatalytic water-splitting, and emerging device architectures. Collectively, these articles provide a comprehensive snapshot of the current state of the field and offer valuable insights into the design of next-generation functional materials.
This paper presents a study of fast neutron imaging using an AmBe radioisotope neutron source and a plastic scintillator. In the new laboratory of neutron applications, the device for imaging was constructed, with the main components being a plastic scintillator and a high-resolution astro camera. Then, the preliminary testing and first experiments were conducted with the AmBe neutron source. The results of these experiments are presented in this paper with a focus on the first images taken, material contrast, and the image processing of the captured image. Furthermore, this paper describes the methodology used to construct and operate the presented device. The objective of this study was to determine the capabilities of the neutron imaging device when utilized under laboratory conditions. This will lead to the main goal, which is to utilize a D-D neutron generator within the device for fast neutron defectoscopy.
We study a population of self-propelled phase oscillators (swarmalators) in which every firing event is followed by a brief silent interval, as in excitable neurons. The equations of motion are strictly pairwise, but the population spontaneously generates a time-dependent hypergraph of higher-order interactions: a hyperedge of order k is any group of k mutually neighbouring agents that are simultaneously inside their emission (refractory) window. The hypergraph is, therefore, not postulated, as in most of the higher-order synchronization literature, but read off from the pairwise dynamics. We report three qualitative effects. First, the mean number of hyperedges of every order obeys a clean two-parameter power law in the spatial degree and the temporal duty fraction; the temporal exponents are consistent with statistically independent refractory clocks, while the spatial exponents systematically exceed the combinatorial null by an order-independent amount, which we identify as the structural fingerprint of positive pair correlations inherited from the cohesive part of the swarmalator force. Second, local pairwise coherence and local triadic coherence (defined on the emergent triangles) are extremised at two well-separated values of the phase coupling. The interval between them is the triangulation window: a range of coupling in which triadic three-fold organisation is favoured while local pairwise alignment is simultaneously suppressed. A run-by-run rank correlation between the two coherences is significantly negative at both operating points, showing that the two channels of order are competing modes rather than projections of a single coherence. Third, across the geometric regimes examined—spanning roughly a factor of two in load—the triadic-coherence optimum is essentially load-independent while the pairwise optimum scales mildly, and their multiplicative separation is approximately conserved (mean ratio of order ≈ 2 in the higher-order-dominated regimes). In this restricted, but consistent sense, the width of the triangulation window is an approximately geometry-independent feature of the refractory-swarmalator dynamics.
Tremor involves complex interactions within the cortex (CTX)-basal ganglia (BG) network and is especially relevant to Parkinson’s disease (PD). However, the mechanism by which CTX, as the input node, shapes tremor-related activity remains unclear. Based on a computational CTX-BG model, this study examines how cortical dynamics modulate tremor-related activity in the internal segment of the globus pallidus (GPi), the output nucleus of BG. Using the oscillation index, cross-spectral coherence, and controlled frequency-response gain, we show marked parameter selectivity in cortical intrinsic dynamics. Specifically, subthreshold voltage-recovery acts as a frequency-selective resonance switch, whereas spike-triggered recovery feedback enhances recovery-driven rhythmic activity. Cortical modulation of GPi tremor-related activity is mainly mediated by resonance and transmitted primarily through the direct pathway. These findings reveal the dynamical mechanisms by which CTX regulates GPi tremor-related activity and provide theoretical and methodological insight for future interventions targeting resting tremor in PD.
Obstructive sleep apnoea (OSA) and major depressive disorder (MDD) frequently co-occur and both have been associated with autonomic dysregulation. Whether the co-occurrence alters the short-scale dynamical organisation of cardiac control during the cyclic alternating pattern (CAP) of NREM sleep is unknown. We retrospectively analysed overnight polysomnography from 44 adults with OSA, 18 with comorbid MDD diagnosed by psychiatrist-administered Mini-International Neuropsychiatric Interview (MINI) version 5 (OSAD+) and 26 without MDD (OSAD−) and from 10 healthy controls drawn from an institutional CAP repository. All MDD patients were antidepressant-naive at the time of PSG. After delta-band-based CAP A-phase detection on the C4–A1 EEG, we extracted 22 HRV indices spanning time, frequency (Lomb–Scargle) and nonlinear domains from RR intervals occurring inside CAP windows, and compared groups with two-tailed Mann–Whitney U tests, Benjamini–Hochberg FDR-corrected, supplemented by stepwise logistic regression and receiver-operating characteristic (ROC) analysis. The OSAD+ and OSAD− groups did not differ in age, sex, BMI, AHI or CAP rate. Sample entropy (SampEn) during CAP was reduced in OSAD− compared with healthy controls (0.87 ± 0.05 vs 1.05 ± 0.02, p = 0.004) and was not reduced in OSAD+ (1.00 ± 0.05; OSAD+ vs control p = 0.29). The OSAD+ vs OSAD− difference reached nominal significance (raw p = 0.043; AUC = 0.69) but did not survive correction for multiple comparisons across the 22 features (Benjamini–Hochberg-adjusted p = 0.83). A multivariable logistic-regression model retaining SampEn and high-frequency power (HF) achieved an AUC of approximately 0.77 and is reported as hypothesis-generating; no other linear or nonlinear HRV index distinguished the groups. In OSA, the short-scale irregularity of CAP-aligned RR intervals is reduced relative to healthy controls in the absence of MDD and is restored towards control values in the presence of comorbid MDD. The finding is exploratory and requires replication in larger cohorts with prospective insomnia screening (Insomnia Severity Index) and stratification by depression subtype, but it identifies CAP-aligned SampEn as a candidate physiological marker of the depressive phenotype within OSA and motivates a re-framing of “complexity loss” interpretations as context- and microstructure-dependent.
Electronic health records (EHRs) contain vast volumes of clinical information that encode complex relationships between diseases. Traditional approaches to the analysis of interrelated or co-occurring diseases have focused on pairwise associations between diagnoses, missing the higher-order structures that characterise multimorbid patients. The present paper offers a narrative review of existing statistical, machine-learning, and artificial intelligence methods for extracting and analysing complex interrelated or co-occurring diseases from EHRs. The application of diverse approaches based on the analysis of diagnostic codes, of textual data, and of combinations of sources of different modalities within hybrid methods makes it possible to uncover emergent interactions between diseases. The work shows the critical role of natural-language processing of narrative text in extracting clinically relevant information absent from structured codes, which in turn motivates the integrated application of deep-learning methods with graph-based analysis for the identification of disease clusters and patient stratification. The analysis shows that going beyond pairwise comorbidity models enables the discovery of collective disease-progression trajectories, offering new opportunities for personalised medicine.
This study presents a hybrid framework for modeling the impact of climate variability on infectious disease dynamics by integrating a compartmental SEIRC (Susceptible–Exposed–Infectious–Recovered–Climate) model with deep neural networks (DNNs). The proposed approach explicitly incorporates temperature, precipitation, and humidity effects into disease transmission through AI-driven parameter estimation, thereby enhancing predictive accuracy. Rigorous mathematical analysis establishes the epidemiological soundness of the model, including the derivation of the basic reproduction number ℛ_0 and stability conditions for disease-free and endemic equilibria. Sensitivity analysis identifies key parameters influencing ℛ_0 , while Physics-Informed Neural Networks (PINNs) are employed to obtain dynamic solutions of the governing system. Numerical simulations demonstrate the significant role of climate factors and recovery rates in shaping infection trajectories. By coupling mechanistic epidemiological modeling with data-driven adaptability, this framework provides a robust tool for real-time outbreak prediction and public health intervention assessment under climate variability. The results highlight the importance of climate sensitivity coefficients and the effectiveness of artificial intelligence in refining estimates of transmission rates.
This paper presents a study of the behavior of a group of gas bubbles in a liquid flow and their mutual influence on mass transfer. This study is a development of a previous analysis of the influence of the shape of a single bubble on mass transfer parameters in multiphase media: the new study examines the influence of gas content (volume fraction of gas in liquid) and the mutual arrangement of bubbles on gas transfer from bubbles to liquid. The results obtained show that at high flow intensity and a large number of bubbles, an increase in the total interfacial contact area does not necessarily lead to a proportional increase in the mass transfer coefficients k_L and the volumetric coefficient k_L a . Moreover, the efficiency of gas dissolution per bubble may decrease. Illustrative simulations of the distribution of dissolved gas concentration around a group of bubbles demonstrate that the concentration “trail” from a single bubble in an intense flow can cast a diffusion shadow, creating a diffusion shielding effect that affects neighboring bubbles and the group dynamics as a whole. A steady-state mass transfer regime is not established: the relationship between bubbles and dynamic interactions leads to constant fluctuations in concentration fields. The data obtained are important for understanding large-scale gas–liquid systems: it has been shown that the mutual location of bubbles and their collective dynamics significantly affect the efficiency of gas dissolution in liquid, and these effects must be taken into account, including when evaluating mass transfer in real systems.
In this paper, we describe methods for mapping the information contained in a spline model of one cycle of an audio signal to the quantum information setting. We use this mapping to describe methods of sonification of quantum algorithms which use spline models to describe and generate timbre, melody, and rhythm. We show how a musical state vector given by n B-spline coefficients, which represent one cycle of a waveform, can produce these three musical elements simultaneously. We present two modes of correspondence between quantum and musical states. The first mode is derived from proposed signal processing of quantum audio, such as QPAM (quantum probability amplitude modulation), and the second mode is more abstract, modeling the evolution of musical states through unitary evolution of intermediate states in quantum algorithms. Musical states are initialized using approximate timbres derived from instrument samples or synthesized timbres. Melodic contours are derived from spline models of these timbres, using continuous pitch and duration spaces. Musical examples are discussed, and larger compositional works are currently under development.
Aortic valve stenosis (AVS) is a progressive disease characterized by inflammation, lipid infiltration, and calcification, ultimately leading to impaired valvular function. Heart rate variability (HRV) has been associated with autonomic impairment in AVS, yet prior ambulatory studies have reported inconsistent associations between this imbalance and AVS severity. Moreover, the relationship between nonlinear HRV indices, particularly those derived from recurrence quantification analysis (RQA) and echocardiographic parameters under controlled short-term conditions, remains insufficiently explored. This cross-sectional study investigated the correlation between linear and nonlinear RQA-derived HRV indices with the AVS severity. 26 patients (57.7
In this work, we investigate the performance of a hybrid quantum–classical convolutional neural network (HQCNN) designed for reliable brain magnetic resonance imaging (MRI) binary classification in case of distorted inputs. The proposed architecture combines a lightweight convolutional feature extractor with a shallow two-qubit variational quantum circuit (VQC) integrated into the final stage of the classification pipeline through end-to-end differentiable optimization. The influence of VQC topology on learning dynamics and classification performance is analyzed varying the depth of the quantum ansatz and the number of repetitions of the elementary quantum block. Experimental results demonstrate that the proposed HQCNN architecture achieves testing accuracy up to 98.6
This paper introduces quantum circuit methodologies conceptualized as “processing through encoding”, for pointwise multiplication and convolution of complex functions through amplitude encoding. Using a scheme with two distinct auxiliary qubits for two functions, f and h, their pointwise product f(x)h(x) emerges as the amplitudes of a specific subspace of the total quantum state. Using this multiplication capability and the convolution theorem, we construct the convolution f*h by encoding the Fourier coefficients ℱ[f] and ℱ[h] , multiplying them pointwise via our encoding scheme, and applying an inverse quantum Fourier transform to the index qubit register. We apply these methods to derive efficient quantum circuits that effectively compute the discrete derivative of encoded functions. Numerical verification is conducted through quantum circuit simulation where signal recovery is examined under shot noise. Real-time audio signal processing capabilities are also evaluated through a strong scaling test involving highly parallelized statevector simulations within a High Performance Computing facility. This work widens the path toward practical quantum signal processing, with potential applications in signal manipulation and synthesis.
This research explores cholera transmission dynamics through a comprehensive mathematical framework that integrates asymptomatic detection measures and advanced fractional-order operators. We study a generalized SEIRB cholera epidemic model incorporating early detection mechanisms for asymptomatic individuals. The model is formulated using a piecewise modified Atangana-Baleanu-Caputo (mABC) fractional derivative operator with Mittag–Leffler kernel, which effectively captures the crossover behavior and memory effects inherent in disease transmission processes. We studied the fundamental properties of the model including existence, uniqueness, positivity, boundedness, and stability of solutions through rigorous mathematical analysis. Our theoretical framework offers valuable insights for developing effective control strategies and early intervention measures for cholera outbreaks. The piecewise methodology allows for modeling different phases of the epidemic with appropriate mathematical operators, enhancing the model’s flexibility and biological relevance. Secondly, we also performed some numerical simulations, developed and trained a Deep Neural Network (DNN) to serve as an alternative function approximator for the system’s solutions.
