scholarly journals Multimodal Tucker Decomposition for Gated RBM Inference

2021 ◽  
Vol 11 (16) ◽  
pp. 7397
Author(s):  
Mauricio Maldonado-Chan ◽  
Andres Mendez-Vazquez ◽  
Ramon Osvaldo Guardado-Medina
Keyword(s):  
Three Dimensional ◽  
Feature Learning ◽  
Model Parameters ◽  
Affine Transformations ◽  
Tucker Decomposition ◽  
Still Images ◽  
Dimensional Interaction ◽  
Fully Connected ◽  
Image Transformations ◽  
And Training

Gated networks are networks that contain gating connections in which the output of at least two neurons are multiplied. The basic idea of a gated restricted Boltzmann machine (RBM) model is to use the binary hidden units to learn the conditional distribution of one image (the output) given another image (the input). This allows the hidden units of a gated RBM to model the transformations between two successive images. Inference in the model consists in extracting the transformations given a pair of images. However, a fully connected multiplicative network creates cubically many parameters, forming a three-dimensional interaction tensor that requires a lot of memory and computations for inference and training. In this paper, we parameterize the bilinear interactions in the gated RBM through a multimodal tensor-based Tucker decomposition. Tucker decomposition decomposes a tensor into a set of matrices and one (usually smaller) core tensor. The parameterization through Tucker decomposition helps reduce the number of model parameters, reduces the computational costs of the learning process and effectively strengthens the structured feature learning. When trained on affine transformations of still images, we show how a completely unsupervised network learns explicit encodings of image transformations.

scholarly journals A flexible framework for sequential estimation of model parameters in computational hemodynamics

2020 ◽  
Vol 7 (1) ◽  
Author(s):  
Christopher J. Arthurs ◽  
Nan Xiao ◽  
Philippe Moireau ◽  
Tobias Schaeffter ◽  
C. Alberto Figueroa
Keyword(s):  
Unscented Kalman Filter ◽  
Three Dimensional ◽  
Synthetic Data ◽  
Sequential Estimation ◽  
Wall Material ◽  
Patient Specific ◽  
Model Parameters ◽  
Computational Framework ◽  
Lumped Parameter ◽  
Estimation Of Model Parameters

AbstractA major challenge in constructing three dimensional patient specific hemodynamic models is the calibration of model parameters to match patient data on flow, pressure, wall motion, etc. acquired in the clinic. Current workflows are manual and time-consuming. This work presents a flexible computational framework for model parameter estimation in cardiovascular flows that relies on the following fundamental contributions. (i) A Reduced-Order Unscented Kalman Filter (ROUKF) model for data assimilation for wall material and simple lumped parameter network (LPN) boundary condition model parameters. (ii) A constrained least squares augmentation (ROUKF-CLS) for more complex LPNs. (iii) A “Netlist” implementation, supporting easy filtering of parameters in such complex LPNs. The ROUKF algorithm is demonstrated using non-invasive patient-specific data on anatomy, flow and pressure from a healthy volunteer. The ROUKF-CLS algorithm is demonstrated using synthetic data on a coronary LPN. The methods described in this paper have been implemented as part of the CRIMSON hemodynamics software package.

scholarly journals CT based automatic clinical target volume delineation using a dense-fully connected convolution network for cervical Cancer radiation therapy

BMC Cancer ◽  
2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Zhongjian Ju ◽  
Wen Guo ◽  
Shanshan Gu ◽  
Jin Zhou ◽  
Wei Yang ◽  
...  
Keyword(s):  
Cervical Cancer ◽  
Cancer Patients ◽  
Clinical Target Volume ◽  
Target Volume ◽  
Ct Images ◽  
Clinical Samples ◽  
Model Parameters ◽  
Convolutional Network ◽  
Volume Delineation ◽  
Fully Connected

Abstract Background It is very important to accurately delineate the CTV on the patient’s three-dimensional CT image in the radiotherapy process. Limited to the scarcity of clinical samples and the difficulty of automatic delineation, the research of automatic delineation of cervical cancer CTV based on CT images for new patients is slow. This study aimed to assess the value of Dense-Fully Connected Convolution Network (Dense V-Net) in predicting Clinical Target Volume (CTV) pre-delineation in cervical cancer patients for radiotherapy. Methods In this study, we used Dense V-Net, a dense and fully connected convolutional network with suitable feature learning in small samples to automatically pre-delineate the CTV of cervical cancer patients based on computed tomography (CT) images and then we assessed the outcome. The CT data of 133 patients with stage IB and IIA postoperative cervical cancer with a comparable delineation scope was enrolled in this study. One hundred and thirteen patients were randomly designated as the training set to adjust the model parameters. Twenty cases were used as the test set to assess the network performance. The 8 most representative parameters were also used to assess the pre-sketching accuracy from 3 aspects: sketching similarity, sketching offset, and sketching volume difference. Results The results presented that the DSC, DC/mm, HD/cm, MAD/mm, ∆V, SI, IncI and JD of CTV were 0.82 ± 0.03, 4.28 ± 2.35, 1.86 ± 0.48, 2.52 ± 0.40, 0.09 ± 0.05, 0.84 ± 0.04, 0.80 ± 0.05, and 0.30 ± 0.04, respectively, and the results were greater than those with a single network. Conclusions Dense V-Net can correctly predict CTV pre-delineation of cervical cancer patients and can be applied in clinical practice after completing simple modifications.

scholarly journals Using a 4K three-dimensional exoscope system (Vitom 3D) for mastoid surgery during the coronavirus disease 2019 pandemic

2021 ◽  
pp. 1-3
Author(s):  
M Ally ◽  
P Kullar ◽  
G Mochloulis ◽  
A Vijendren
Keyword(s):  
Residual Disease ◽  
Elective Surgery ◽  
Three Dimensional ◽  
Depth Of Field ◽  
High Definition ◽  
Microscopic Surgery ◽  
Lateral Skull Base ◽  
Mastoid Surgery ◽  
And Training

Abstract Objective Microscopic surgery is currently considered the ‘gold standard’ for middle-ear, mastoid and lateral skull base surgery. The coronavirus disease 2019 pandemic has made microscopic surgery more challenging to perform. This work aimed to demonstrate the feasibility of the Vitom 3D system, which integrates a high-definition (4K) view and three-dimensional technology for ear surgery, within the context of the pandemic. Method Combined approach tympanoplasty and ossiculoplasty were performed for cholesteatoma using the Vitom 3D system exclusively. Results Surgery was performed successfully. The patient made a good recovery, with no evidence of residual disease at follow up. The compact system has excellent depth of field, magnification and colour. It enables ergonomic work, improved work flow, and is ideal for teaching and training. Conclusion The Vitom 3D system is considered a revolutionary alternative to microscope-assisted surgery, particularly in light of coronavirus disease 2019. It allows delivery of safe otological surgery, which may aid in continuing elective surgery.

scholarly journals A faster escape does not enhance survival in zebrafish larvae

2017 ◽  
Vol 284 (1852) ◽  
pp. 20170359 ◽  
Author(s):  
Arjun Nair ◽  
Christy Nguyen ◽  
Matthew J. McHenry
Keyword(s):  
High Speed ◽  
Larval Fish ◽  
Body Position ◽  
Three Dimensional ◽  
Limited Range ◽  
The Body ◽  
Model Parameters ◽  
Fish Prey ◽  
Zebrafish Larvae ◽  
Fish Predators

An escape response is a rapid manoeuvre used by prey to evade predators. Performing this manoeuvre at greater speed, in a favourable direction, or from a longer distance have been hypothesized to enhance the survival of prey, but these ideas are difficult to test experimentally. We examined how prey survival depends on escape kinematics through a novel combination of experimentation and mathematical modelling. This approach focused on zebrafish ( Danio rerio ) larvae under predation by adults and juveniles of the same species. High-speed three-dimensional kinematics were used to track the body position of prey and predator and to determine the probability of behavioural actions by both fish. These measurements provided the basis for an agent-based probabilistic model that simulated the trajectories of the animals. Predictions of survivorship by this model were found by Monte Carlo simulations to agree with our observations and we examined how these predictions varied by changing individual model parameters. Contrary to expectation, we found that survival may not be improved by increasing the speed or altering the direction of the escape. Rather, zebrafish larvae operate with sufficiently high locomotor performance due to the relatively slow approach and limited range of suction feeding by fish predators. We did find that survival was enhanced when prey responded from a greater distance. This is an ability that depends on the capacity of the visual and lateral line systems to detect a looming threat. Therefore, performance in sensing, and not locomotion, is decisive for improving the survival of larval fish prey. These results offer a framework for understanding the evolution of predator–prey strategy that may inform prey survival in a broad diversity of animals.

Three-Dimensional Interaction Effects in an Internally Multicracked Pressurized Thick-Walled Cylinder— Part I: Radial Crack Arrays

1996 ◽  
Vol 118 (3) ◽  
pp. 357-363 ◽  
Author(s):  
M. Perl ◽  
C. Levy ◽  
J. Pierola
Keyword(s):  
Radial Crack ◽  
Crack Depth ◽  
Three Dimensional ◽  
Internal Surface ◽  
Interaction Effects ◽  
Life Assessment ◽  
Fatigue Life Assessment ◽  
Dimensional Interaction ◽  
Least Squares Fit ◽  
Wide Range

Under certain conditions, numerous internal surface cracks develop in pressurized thick-walled cylinders, both in the radial and longitudinal directions. For fatigue life assessment of such vessels, the 3-D interaction effects among these cracks on the prevailing stress intensity factors (SIFs) need evaluation. In Part I of this paper, radial crack arrays are considered exclusively. The mode I SIF distribution for a wide range of semi-circular and semi-elliptical cracks are evaluated. The 3-D analysis is performed via the finite element method with the submodeling technique, employing singular elements along the crack front. SIFs are evaluated for arrays of up to n = 180 cracks; for a wide range of crack depth to wall thickness ratios, a/t, from 0.05 to 0.6; and, for various ellipticities of the crack, i.e., the ratio of crack depth to semicrack length, a/c, from 0.2 to 2. Using a least-squares fit, two simple expressions for the most critical (n = 2) SIFs are obtained for sparse and dense crack arrays. The formulas, which are functions of a/t and a/c, are of very good engineering accuracy. The results clearly indicate that the SIFs are considerably affected by the interaction among the cracks in the array as well as the three-dimensionality of the problem. In Part II of this paper, the interaction effects between longitudinal coplanar cracks will be analyzed.

scholarly journals Three-dimensional data-tracking simulations of sprinting using a direct collocation optimal control approach

PeerJ ◽  
2021 ◽  
Vol 9 ◽  
pp. e10975
Author(s):  
Nicos Haralabidis ◽  
Gil Serrancolí ◽  
Steffi Colyer ◽  
Ian Bezodis ◽  
Aki Salo ◽  
...  
Keyword(s):  
Experimental Data ◽  
Reaction Force ◽  
Three Dimensional ◽  
Contact Model ◽  
Model Parameters ◽  
Ground Contact ◽  
Control Approach ◽  
Average Percentage ◽  
Data Tracking ◽  
The Individual

Biomechanical simulation and modelling approaches have the possibility to make a meaningful impact within applied sports settings, such as sprinting. However, for this to be realised, such approaches must first undergo a thorough quantitative evaluation against experimental data. We developed a musculoskeletal modelling and simulation framework for sprinting, with the objective to evaluate its ability to reproduce experimental kinematics and kinetics data for different sprinting phases. This was achieved by performing a series of data-tracking calibration (individual and simultaneous) and validation simulations, that also featured the generation of dynamically consistent simulated outputs and the determination of foot-ground contact model parameters. The simulated values from the calibration simulations were found to be in close agreement with the corresponding experimental data, particularly for the kinematics (average root mean squared differences (RMSDs) less than 1.0° and 0.2 cm for the rotational and translational kinematics, respectively) and ground reaction force (highest average percentage RMSD of 8.1%). Minimal differences in tracking performance were observed when concurrently determining the foot-ground contact model parameters from each of the individual or simultaneous calibration simulations. The validation simulation yielded results that were comparable (RMSDs less than 1.0° and 0.3 cm for the rotational and translational kinematics, respectively) to those obtained from the calibration simulations. This study demonstrated the suitability of the proposed framework for performing future predictive simulations of sprinting, and gives confidence in its use to assess the cause-effect relationships of technique modification in relation to performance. Furthermore, this is the first study to provide dynamically consistent three-dimensional muscle-driven simulations of sprinting across different phases.

Towards an Efficient Validation of Dynamical Whole-brain Models

2021 ◽  
Author(s):  
Kevin J. Wischnewski ◽  
Simon B. Eickhoff ◽  
Viktor K. Jirsa ◽  
Oleksandr V. Popovych
Keyword(s):  
Structural Connectivity ◽  
Three Dimensional ◽  
Bayesian Optimization ◽  
High Dimensional ◽  
Brain Dynamics ◽  
Model Parameters ◽  
Phase Oscillators ◽  
Dimensional Parameter ◽  
Whole Brain ◽  
Brain Models

Abstract Simulating the resting-state brain dynamics via mathematical whole-brain models requires an optimal selection of parameters, which determine the model’s capability to replicate empirical data. Since the parameter optimization via a grid search (GS) becomes unfeasible for high-dimensional models, we evaluate several alternative approaches to maximize the correspondence between simulated and empirical functional connectivity. A dense GS serves as a benchmark to assess the performance of four optimization schemes: Nelder-Mead Algorithm (NMA), Particle Swarm Optimization (PSO), Covariance Matrix Adaptation Evolution Strategy (CMAES) and Bayesian Optimization (BO). To compare them, we employ an ensemble of coupled phase oscillators built upon individual empirical structural connectivity of 105 healthy subjects. We determine optimal model parameters from two- and three-dimensional parameter spaces and show that the overall fitting quality of the tested methods can compete with the GS. There are, however, marked differences in the required computational resources and stability properties, which we also investigate before proposing CMAES and BO as efficient alternatives to a high-dimensional GS. For the three-dimensional case, these methods generated similar results as the GS, but within less than 6% of the computation time. Our results contribute to an efficient validation of models for personalized simulations of brain dynamics.

Inverse modeling of transient three-dimensional core-scale two-phase flows

2021 ◽  
Author(s):  
Andrea Manzoni ◽  
Aronne Dell'Oca ◽  
Martina Siena ◽  
Alberto Guadagnini
Keyword(s):  
Capillary Pressure ◽  
Random Noise ◽  
Three Dimensional ◽  
Population Based ◽  
Model Parameters ◽  
Reference Scenario ◽  
Two Phase ◽  
De Algorithm ◽  
Continuum Scale ◽  
Available Information

<p>We consider transient three-dimensional (3D) two-phase (oil and water) flows, taking place at the core-scale. In this context, we aim at exploiting the full information content associated with available information of (i) the 3D distribution of oil saturation and (ii) the overall pressure difference across the rock sample, to estimate the set of model parameters. We consider a continuum-scale description of the system behavior upon relying on the widely employed Brooks-Corey model for the characterization of relative permeabilities and on the capillary pressure correlation introduced by Skjaeveland et al. (2000). To provide a transparent way of assessing the results of the inversion, we rely on a synthetic reference scenario. The latter is intended to mimic having at our disposal 3D and section-averaged distributions of (time-dependent) oil saturations of the kind that can be acquired during typical laboratory experiments. These are in turn corrupted by way of a random noise, to address the influence of experimental uncertainties. We focus on diverse scenarios encompassing imbibition and drainage conditions. We employ two population-based optimization algorithms, i.e., (i) the particle swarm optimization (PSO); and (ii) the differential evolution (DE), which enable one to effectively tackle the high-dimensionality parameters space (i.e., 12 dimensions in our setting) we consider. Model calibration results are of satisfactory quality for the majority of the tested scenarios, whereas the DE algorithm is associated with highest effectiveness.</p><p><strong>References</strong></p><p>S.M. Skjaeveland; L.M. Siqveland; A. Kjosavik; W.L. Hammervold Thomas; G.A. Virnovsky (2000). Capillary Pressure Correlation for Mixed-Wet Reservoirs SPE Res Eval & Eng 3 (01): 60–67. https://doi.org/10.2118/60900-PA</p>

Sign in / Sign up

Export Citation Format

Share Document