Multi-resolution Model Research Articles

Editorial Issue 34.2

Editorial Issue 34.2

An attention-based multi-resolution deep learning model for automatic A-phase detection of cyclic alternating pattern in sleep using single-channel EEG

Sleep is a crucial part of human well-being. Many people suffer from various sleep disorders and insufficient sleep. The detection of the cyclic alternating pattern (CAP) of electroencephalogram (EEG) activity during sleep is essential for identifying and monitoring these problems. In this paper, we present a multi-resolution deep neural network model with temporal and channel attention for detecting A-phase and its subtypes. A multi-branch one-dimensional convolutional neural network (1D-CNN) is employed where each branch has different kernel sizes to extract features of different frequency resolutions automatically. An attention-based transformer network exploits the dynamic and temporal relationship between CAP event features extracted from the single-channel EEG data. Our model achieves 90.31% accuracy, 95.30% specificity, and 65.73% F1-Score in A-phase detection and 86.72% accuracy, 89.53% specificity, and 59.59% F1-Score in the detection of its subtypes, superior performance as compared to those of the recent approaches.

Fast, Accurate, and System-Specific Variable-Resolution Modeling of Proteins.

In recent years, a few multiple-resolution modeling strategies have been proposed, in which functionally relevant parts of a biomolecule are described with atomistic resolution, with the remainder of the system being concurrently treated using a coarse-grained model. In most cases, the parametrization of the latter requires lengthy reference all-atom simulations and/or the usage of off-shelf coarse-grained force fields, whose interactions have to be refined to fit the specific system under examination. Here, we overcome these limitations through a novel multiresolution modeling scheme for proteins, dubbed coarse-grained anisotropic network model for variable resolution simulations, or CANVAS. This scheme enables a user-defined modulation of the resolution level throughout the system structure; a fast parametrization of the potential without the necessity of reference simulations; and the straightforward usage of the model on the most commonly used molecular dynamics platforms. The method is presented and validated with two case studies, the enzyme adenylate kinase and the therapeutic antibody pembrolizumab, by comparing the results obtained with the CANVAS model against fully atomistic simulations. The modeling software, implemented in Python, is made freely available for the community on a collaborative github repository.

Multiple time-instances features based approach for reference-free speech quality measurement

This paper investigates the problem of measuring speech quality of received speech signal without employing the original speech signal. The problem of deterioration of the speech quality arises due to noise present in the surroundings. To this line, we propose a multiple time-instances (MTI) features-based approach for reference-free speech quality measurement model. A voice activity detector (VAD) is exploited primarily for calculating the number of active speech chunks of a speech signal. For these chunks and their successive combinations called here batches, multi-resolution auditory model (MRAM), mel-frequency cepstral coefficients (MFCC) and line spectral frequencies (LSF) features are extracted and called as MTI features. It is hypothesized that the MTI features are capable in capturing the distortions caused by time-localized effects of short-time transients, impulsive noise, and its differences from the plosive sounds. The MTI metric estimates (MTI-ME) are calculated corresponding to these MTI features employing the Gaussian mixture model (GMM) probabilistic technique. The overall objective speech quality of a speech signal is then determined as a linear combination of optimally weighted MTI-ME corresponding to distinct active speech chunks and their successive combinations, that is, batches of that speech signal. Minimum mean square error criterion or Pearson’s correlation maximization criterion is employed for computing optimal weights. In addition, a deep neural network (DNN)-based speech quality model is also developed for calculating a single objective speech quality while considering all active speech chunks together. Pearson’s correlation coefficient and weighted average correlation are exploited for evaluating the performance. Results demonstrate that the proposed model achieves promising improvement over the standard speech quality model (P.563) and improves correlation values by around 37%.

An AI-assisted tool for efficient prostate cancer diagnosis in low-grade and low-volume cases

Pathologists diagnose prostate cancer by core needle biopsy. In low-grade and low-volume cases, they look for a few malignant glands out of hundreds within a core. They may miss a few malignant glands, resulting in repeat biopsies or missed therapeutic opportunities. This study developed a multi-resolution deep-learning pipeline to assist pathologists in detecting malignant glands in core needle biopsies of low-grade and low-volume cases. Analyzing a gland at multiple resolutions, our model exploited morphology and neighborhood information, which were crucial in prostate gland classification. We developed and tested our pipeline on the slides of a local cohort of 99 patients in Singapore. Besides, we made the images publicly available, becoming the first digital histopathology dataset of patients of Asian ancestry with prostatic carcinoma. Our multi-resolution classification model achieved an area under the receiver operating characteristic curve (AUROC) value of 0.992 (95% confidence interval [CI]: 0.985-0.997) in the external validation study, showing the generalizability of our multi-resolution approach.

Acoustic shape optimization based on isogeometric boundary element method with subdivision surfaces

This study proposes a new acoustic shape optimization approach by combining isogeometric subdivision surfaces with the boundary element method. The geometry of the structural boundary is constructed by the Catmull–Clark subdivision surfaces, which are able to provide multiresolution subdivision hierarchy models to meet different precision requirements. The acoustic propagation in the unbounded domain is simulated using the boundary element method. The control points in subdivision meshes are set as design variables, and the minimization of sound pressure at the observation points is selected as the design objective. The acoustic shape sensitivities are calculated by the sensitivity boundary integral equation based on the direct differentiation method. In each iteration the geometry is updated from a direct optimized shape, then a secondary processing is considered to obtain smooth surfaces. Compared with the conventional optimization methods, we eliminate jagged geometry by multiresolution approach without needing the cumbersome meshing procedure and volume parameterization. Several numerical experiments demonstrate the effectiveness and validity of the developed shape optimization approach.

Robust Method for Reservoir Simulation History Matching Using Bayesian Inversion and Long-Short-Term Memory Network-Based Proxy

Summary History matching is a critical process used for calibrating simulation models and assessing subsurface uncertainties. This common technique aims to align the reservoir models with the observed data. However, achieving this goal is often challenging due to the nonuniqueness of the solution, underlying subsurface uncertainties, and usually the high computational cost of simulations. The traditional approach is often based on trial and error, which is exhaustive and labor-intensive. Some analytical and numerical proxies combined with Monte Carlo simulations are used to reduce the computational time. However, these approaches suffer from low accuracy and may not fully capture subsurface uncertainties. This study proposes a new robust method using Bayesian Markov chain Monte Carlo (MCMC) to perform assisted history matching under uncertainties. We propose a novel three-step workflow that includes (1) multiresolution low-fidelity models to guarantee high-quality matching; (2) long-short-term memory (LSTM) network as a low-fidelity model to reproduce continuous time response based on the simulation model, combined with Bayesian optimization to obtain the optimum low-fidelity model; and (3) Bayesian MCMC runs to obtain the Bayesian inversion of the uncertainty parameters. We perform sensitivity analysis on the LSTM’s architecture, hyperparameters, training set, number of chains, and chain length to obtain the optimum setup for Bayesian-LSTM history matching. We also compare the performance of predicting the recovery factor (RF) using different surrogate methods, including polynomial chaos expansions (PCE), kriging, and support vector machines for regression (SVR). We demonstrate the proposed method using a water flooding problem for the upper Tarbert formation of the 10th SPE comparative model. This study case represents a highly heterogeneous nearshore environment. Results showed that the Bayesian-optimized LSTM has successfully captured the physics in the high-fidelity model. The Bayesian-LSTM MCMC produces an accurate prediction with narrow ranges of uncertainties. The posterior prediction through the high-fidelity model ensures the robustness and accuracy of the workflow. This approach provides an efficient and practical history-matching method for reservoir simulation and subsurface flow modeling with significant uncertainties.

Exploring the Evolution of Typhoon Lekima (2019) Moving Offshore Northeast of Taiwan with a Multi-Resolution Global Model

Typhoon Lekima occurred in early August 2019 and moved northwestward toward Taiwan. During offshore passage, the typhoon underwent rapid intensification, with a northward deflected track, moving closer to northeastern Taiwan. A global model, MPAS, at a multi-resolution of 60-15-3 km, is utilized with ensemble forecasts to investigate the dynamic processes causing the track deflection and intensity change as well as identify the track uncertainty to initial perturbed conditions under the topographic effects of the Central Mountain Range (CMR). For spinning up the typhoon vortex in ensemble forecasts, dynamic vortex initialization has been enforced with a 3 km resolution targeted at the Taiwan area. As one specific member track is similar to the best track, the track deflection is significantly reduced in the absence of the Taiwan terrain, highlighting the role of the topographic effects of the CMR. For these tracks with similar deflection, the northward deflection is caused by the induced strong flow to the east of the typhoon center in response to the re-circulating flow around southern Taiwan, which produces the wavenumber-one gyre in the asymmetric flow difference to drive the vortex northward. The typhoon translation around the Taiwan terrain is dominated by the changing wavenumber-one horizontal potential vorticity (PV) advection during the track deflection in the ensemble forecasts. The formation of an intense PV tongue along the upper eyewall is a facilitation precondition of RI, while RI can be significantly enhanced in the presence of an intense lower-stratospheric PV core near the upper eye, which is produced by the radial inflow of the developed transverse vortex circulation over the upper-level outflow layer.

Construction of rhombic triacontahedron discrete global grid systems

ABSTRACT Discrete Global Grid System (DGGS) is a new multi-resolution geospatial data modeling and processing scheme for the digital earth. The icosahedron is commonly regarded as an ideal polyhedron for constructing DGGSs with small distortions; however, the shape of its face is triangular, making it difficult to incorporate the matrix structure used for geospatial data storage and parallel computing. To overcome this limitation, this study utilizes the rhombic triacontahedron (RT) as the basic polyhedron to construct DGGSs. An equal-area projection between the surface of RT and the sphere is developed and used to design a grid-generation algorithm for the aperture 4 hexagonal DGGS based on RT. Compared with the equal-area DGGS based on the icosahedron, the proposed scheme results in smaller angular projection distortions, with the mean and standard deviation decreasing by 41.6% and 30.9%, respectively. The grid cells of the RT DGGS also achieve more optimized geometric characteristics in shape compactness, length deviation, and angle deviation than those in the icosahedron DGGS. Additionally, the cross-surface computation efficiency provides advantages in code conversion to latitude and longitude and proximity queries. Furthermore, the use of RT offers a new and better framework within the context of DGGS research and application.

Practical Formalism-Based Approaches for Multi-Resolution Modeling and Simulation

Multi-resolution modeling (MRM) has been considered as an ideal form of simulation to acquire low-resolution scalability as well as high-resolution modeled details. Although both practical and theoretical interests exist in MRM, actual implementations were quite different in terms of cases and methods. Specifically, MRM implementations range from parameter-based interoperation to model exchanges with different resolutions, yet it is difficult to observe a method that focuses on both of these aspects. To this end, this paper introduces a formalism or multi-resolution translational Discrete Event System Specification (MRT-DEVS). Focusing on the practical perspective, MRT-DEVS intends to ease the implementation’s difficulty and reduce the simulation’s execution costs. Specifically, MRT-DEVS embeds state and event translation functions into the model’s specifications so that it enables MRM with less complex mechanisms in terms of operations. Using the provided case study and a reduction to other MRM methods, the theoretical soundness of the proposed method is supported. Moreover, we discussed the pros and the cons of the proposed method from various MRM perspectives. We expect that with all the provided information, MRMS users would consider the proposed method as a practical option to implement their models.

Optimization-Based High-Frequency Circuit Miniaturization through Implicit and Explicit Constraint Handling: Recent Advances

Miniaturization trends in high-frequency electronics have led to accommodation challenges in the integration of the corresponding components. Size reduction thereof has become a practical necessity. At the same time, the increasing performance demands imposed on electronic systems remain in conflict with component miniaturization. On the practical side, the challenges related to handling design constraints are aggravated by the high cost of system evaluation, normally requiring full-wave electromagnetic (EM) analysis. Some of these issues can be alleviated by implicit constraint handling using the penalty function approach. Yet, its performance depends on the arrangement of the penalty factors, necessitating a costly trial-and-error procedure to identify their optimum setup. A workaround is offered by the recently proposed algorithms with automatic adaptation of the penalty factors using different adjustment schemes. However, these intricate strategies require a continuous problem-dependent adaptation of the penalty function throughout the entire optimization process. Alternative methodologies have been proposed by taking an explicit approach to handle the inequality constraints, along with correction-based control over equality conditions, the combination of which proves to be demonstrably competitive for some miniaturization tasks. Nevertheless, optimization-based miniaturization, whether using implicit or explicit constraint handling, remains a computationally expensive task. A reliable way of reducing the aforementioned costs is the incorporation of multi-resolution EM fidelity models into the miniaturization procedure. Therein, the principal operation is based on the simultaneous monitoring of factors such as quality of the constraint satisfaction, as well as algorithm convergence status. This paper provides an overview of the abovementioned size-reduction algorithms, in which theoretical considerations are illustrated using a number of antenna and microwave circuit case studies.

High Performance Adaptive Physics Refinement to Enable Large-Scale Tracking of Cancer Cell Trajectory.

The ability to track simulated cancer cells through the circulatory system, important for developing a mechanistic understanding of metastatic spread, pushes the limits of today's supercomputers by requiring the simulation of large fluid volumes at cellular-scale resolution. To overcome this challenge, we introduce a new adaptive physics refinement (APR) method that captures cellular-scale interaction across large domains and leverages a hybrid CPU-GPU approach to maximize performance. Through algorithmic advances that integrate multi-physics and multi-resolution models, we establish a finely resolved window with explicitly modeled cells coupled to a coarsely resolved bulk fluid domain. In this work we present multiple validations of the APR framework by comparing against fully resolved fluid-structure interaction methods and employ techniques, such as latency hiding and maximizing memory bandwidth, to effectively utilize heterogeneous node architectures. Collectively, these computational developments and performance optimizations provide a robust and scalable framework to enable system-level simulations of cancer cell transport.

Multiresolution Data Assimilation for Auroral Energy Flux and Mean Energy Using DMSP SSUSI, THEMIS ASI, and An Empirical Model

AbstractWe apply a multiresolution Gaussian process model (Lattice Kriging) to combine satellite observations, ground‐based observations, and an empirical auroral model, to produce the assimilation of auroral energy flux and mean energy over high‐latitude regions. Compared to a simple padding, the assimilation coherently combines various data inputs leading to continuous transitions between different datasets. The multiresolution modeling capability is achieved by allocating multiple layers of basis functions with different resolutions. Higher‐resolution fitting results capture more mesoscale (10–100 s km) structures such as auroral arcs, than the low‐resolution ones and the empirical model. To better reconcile different datasets, two preprocessing steps, temporal interpolation of satellite data and spatial down‐sampling of low‐fidelity data, are implemented. The inherent smoothing effect of the fitting, which causes an unrealistic spreading of the aurora, is mitigated by a post processing step: the K Nearest Neighbor (KNN) algorithm. KNN identifies the probability of a region with significant aurora and thereby eliminates those regions with low values. Thereby, this methodology can be used to maintain realistic and mesoscale auroral structures without boundary issues. We then run the Thermosphere Ionosphere Electrodynamics General Circulation Model (TIEGCM) driven by the high‐ and low‐resolution auroral assimilations and compare total electron contents (TECs). TIEGCM driven by data assimilation produces enhanced TECs by a factor of ∼2 than the one driven by the empirical aurora, and high‐resolution results show mesoscale structures. Our study shows the value of incorporating realistic auroral inputs via assimilation to drive ionosphere‐thermosphere models for better understanding the consequences of mesoscale phenomena.

Research on co-simulation of multi-resolution models based on HLA

Intelligence, hierarchy, and multi-resolution (MR) modeling are vital to the complex system simulation. Based on the proposed MR-Agent-Discrete EVent System (MR-Agent-DEVS) specification, this paper discusses the co-simulation of MR models under high-level architecture (HLA). First, the structure and formal description of MR-Agent-DEVS federate models were proposed. The communication mechanism and simulation process of MR-Agent-DEVS federate models were analyzed. Finally, an example of distributed interactive simulation for emergency material supply was given. The results of the simulation test show that HLA enhances the reusability of MR-Agent-DEVS models, knowledge updating and model coupling enrich the interoperability of MR-Agent-DEVS models, and the real-time memory of knowledge base in Agent ensures the consistency of MR-Agent-DEVS models in aggregation and disaggregation.

Three-dimensional Face Recognition Method Based on Multiresolution Model and Fuzzy Random Matrix

Three-dimensional face recognition is one of the hotspots of biometric recognition and has a wide range of applications in the fields of information security, video surveillance, video tracking, and so on. The core part of 3D face recognition is to establish the corresponding 3D face model, and the key of building the model is how to obtain the shape model and accurate texture mapping. The problem has not been well solved in the field of face reconstruction. Based on the background, this paper makes an in-depth study and proposes a three-dimensional face recognition method based on multiresolution model and fuzzy random matrix. The face model reconstruction method of image and model can accurately obtain the contour change of face and the representation of specific features through multiresolution model, reduce the inaccurate description of error and noise in samples through fuzzy random matrix, and enhance the effectiveness of image information of classification and recognition. The experimental outcomes exhibit that the 3D face focus approach based totally on multiresolution mannequin and fuzzy random matrix efficiently improves the evaluation effectivity of the model, improves the great of mannequin matching, improves the function extraction in the cognizance process, and improves the attention rate.

3D multi-resolution SPH modeling of the water entry dynamics of free-fall lifeboats

Lifeboats can be launched through free-fall in extreme maritime accidents to fast leave the hazardous area, which deserves particular attention because it is tightly related to the safety of lifeboats and the carried passengers as well. This paper, within the framework of a fully 3D and multi-resolution Smoothed Particle Hydrodynamics (SPH) model, aims at investigating the water entry dynamics of a free-fall lifeboat with different water entry angles. First of all, two rigorous model validations are performed to demonstrate the accuracy, stability and convergence of the present SPH model. Furthermore, the water entry dynamics of the lifeboat with respect to different water entry angles, including the motion characteristics and cavity evolutions, is investigated and discussed in detail based on the SPH results. It is clearly demonstrated that the present SPH model possesses (1) satisfactory accuracy for predicting the water entry dynamics of a lifeboat, and (2) a better ability to capture the splashing jets and scattering droplets than mesh-based solvers thanks to its entirely Lagrangian nature. In addition, it is also demonstrated that the water entry angle exerts significant effects on the water entry dynamics of a lifeboat, which should be taken into account in the design stage to provide safer productions.

Multicriteria, Multiresolution Modeling of Suburban Residential Landscape Alternatives: Water-Efficient Villas in the Arid Middle East

Multicriteria, Multiresolution Modeling of Suburban Residential Landscape Alternatives: Water-Efficient Villas in the Arid Middle East

A multi-scale neurodynamic implementation of incremental grouping

Incremental grouping is a process entailing serial binding of distal image elements into a unified object representation. At the neural level, incremental grouping involves propagation of the enhanced firing rate among feature-tuned neurons in the early visual cortex. Here, we developed a multi-resolution neural model of incremental grouping. In the model, propagation of the enhanced firing rate is achieved by computing the activity difference between two sets of units: attentional or A-units, whose firing rate is modulated by their horizontal collaterals, and non-attentional or N-units that receive only feedforward input. The activity difference is computed on dendrites that act as independent computational subunits. The proposed model employs multiple spatial scales to account for a variable speed of incremental grouping. In addition, the model incorporates the L-junction detection network that enables incremental grouping over L-junctions. Computer simulations show that the timing of attentional modulations in the model is comparable with neurophysiological measurements in monkey primary visual cortex.

Numerical investigations on bionic propulsion problems using the multi-resolution Delta-plus SPH model

The bionic propulsion has garnered increasing attention in both scientific and industrial communities of underwater bionic vehicles on account of its higher efficiency, less noise, and remarkable self-shelter capability. This paper is dedicated to investigating the propulsion mechanism of fish swimming using the Smoothed Particle Hydrodynamics (SPH) method. Two fish-like swimming problems, established with rigid flapping foils and deformable anguilliform swimmers respectively, are simulated using the δ+-SPH model under the laminar flow assumption. First of all, numerical benchmarks are implemented to validate the convergence and accuracy of the SPH model by comparing the SPH results with experimental data quantitatively and qualitatively. Furthermore, the mechanisms of the thrust generation of the bionic swimmers are investigated and discussed in detail from the viewpoint of vortex shedding patterns and several dimensionless parameters describing the motion characteristics. It is demonstrated that the thrust force is tightly related to the shape and scale of the vortex shedding in wakes, and that a higher Reynolds number or larger Strouhal number, as well as a larger motion amplitude, are in favor of improving the thrust. In addition, the numerical results also indicate that the SPH method possesses great potential for the simulation of bionic propulsion problems thanks to its Lagrangian and meshfree nature.

Bayesian multiresolution modeling of georeferenced data: An extension of ‘LatticeKrig’

‘LatticeKrig’ (LK) is a spatial model that is often used for modeling multiresolution spatial data with flexible covariance structures. An extension to LK under a Bayesian framework is proposed that uses integrated nested Laplace approximations (INLA). The extension enables the spatial analysis of non-Gaussian responses in latent Gaussian models, joint spatial modeling with structured and unstructured random effects, and native support for multithreaded parallel likelihood computation. The proposed extended LatticeKrig (ELK) model uses a reparameterization of LK so that the parameters and prior selection are intuitive and interpretable. Priors can be used to make inference robust by penalizing more complex models, and integration over model parameters allows for posterior uncertainty estimates that account for uncertainty in covariance parameters. ELK's ability to reliably resolve multiresolution spatial structure for pointwise and areal predictions is demonstrated in both simulation study and two applications with non-Gaussian observations: a set of 188,717 LiDAR forest canopy height observations in Bonanza Creek Experimental Forest in Alaska, and a set of 1,612 clusters containing counts of secondary education completion from the 2014 Kenya demographic health survey. ELK has improved central predictions as well as uncertainty characterization according to the considered scoring rules when compared against a number of other models, particularly in the forest canopy height application, and performed faster than LK in our tests in part due to its support for and use of parallelization.