Ground Truth Simulation Research Articles

Disentangling signal and noise in neural responses through generative modeling.

Measurements of neural responses to identically repeated experimental events often exhibit large amounts of variability. This noise is distinct from signal , operationally defined as the average expected response across repeated trials for each given event. Accurately distinguishing signal from noise is important, as each is a target that is worthy of study (many believe noise reflects important aspects of brain function) and it is important not to confuse one for the other. Here, we describe a principled modeling approach in which response measurements are explicitly modeled as the sum of samples from multivariate signal and noise distributions. In our proposed method-termed Generative Modeling of Signal and Noise (GSN)-the signal distribution is estimated by subtracting the estimated noise distribution from the estimated data distribution. Importantly, GSN improves estimates of the signal distribution, but does not provide improved estimates of responses to individual events. We validate GSN using ground-truth simulations and show that it compares favorably with related methods. We also demonstrate the application of GSN to empirical fMRI data to illustrate a simple consequence of GSN: by disentangling signal and noise components in neural responses, GSN denoises principal components analysis and improves estimates of dimensionality. We end by discussing other situations that may benefit from GSN's characterization of signal and noise, such as estimation of noise ceilings for computational models of neural activity. A code toolbox for GSN is provided with both MATLAB and Python implementations.

Deep Learing for Sparse Domain Kalman Filtering With Applications on ECG Denoising and Motility Estimation.

The reconstruction of an input based on a sparse combination of signals, known as sparse coding, has found widespread use in signal processing. In this work, the combination of sparse coding with Kalman filtering is explored and its potential is shown on two use-cases. This work extends the Iterative Shrinkage and Thresholding Algorithm with a Kalman filter in the sparse domain. The resulting method may be implemented as a deep unfolded neural network and may be applied to any signal which has a sparse representation and a known or assumed relation between consecutive measurements. This method is evaluated on the use cases of noise reduction in the electrocardiogram (ECG) and the estimation of object motility. For ECG denoising, the proposed method achieved an improvement in Signal-to-Noise ratio of 18.6 dB, which is comparable to state-of-the-art. In motility estimation, a correlation of 0.84 with ground truth simulations was found. The proposed method was shown to have advantages over sparse coding and Kalman filtering alone. Due to the low complexity and high generalizability of the proposed method, the implementation of context-specific knowledge or an extension to other applications can be readily made. The presented Kalman-ISTA algorithm is a resource-efficient method combining the promise of both sparse coding and Kalman filtering, making it well-suited for various applications.

Ground-truth validation of uni- and multivariate lesion inference approaches.

Lesion analysis aims to reveal the causal contributions of brain regions to brain functions. Various strategies have been used for such lesion inferences. These approaches can be broadly categorized as univariate or multivariate methods. Here we analysed data from 581 patients with acute ischaemic injury, parcellated into 41 Brodmann areas, and systematically investigated the inferences made by two univariate and two multivariate lesion analysis methods via ground-truth simulations, in which we defined a priori contributions of brain areas to assumed brain function. Particularly, we analysed single-region models, with only single areas presumed to contribute functionally, and multiple-region models, with two contributing regions that interacted in a synergistic, redundant or mutually inhibitory mode. The functional contributions could vary in proportion to the lesion damage or in a binary way. The analyses showed a considerably better performance of the tested multivariate than univariate methods in terms of accuracy and mis-inference error. Specifically, the univariate approaches of Lesion Symptom Mapping as well as Lesion Symptom Correlation mis-inferred substantial contributions from several areas even in the single-region models, and also after accounting for lesion size. By contrast, the multivariate approaches of Multi-Area Pattern Prediction, which is based on machine learning, and Multi-perturbation Shapley value Analysis, based on coalitional game theory, delivered consistently higher accuracy and specificity. Our findings suggest that the tested multivariate approaches produce largely reliable lesion inferences, without requiring lesion size consideration, while the application of the univariate methods may yield substantial mis-localizations that limit the reliability of functional attributions.

Inexpensive high fidelity melt pool models in additive manufacturing using generative deep diffusion

Defects in Laser Powder Bed Fusion (L-PBF) parts often result from the meso-scale dynamics of the molten alloy near the laser, known as the melt pool. Experimental in-situ monitoring of the three-dimensional melt pool physical fields is challenging, due to the short length and time scales involved in the process. Multi-physics simulation methods can describe the three-dimensional dynamics of the melt pool, but are computationally expensive at the mesh refinement required for accurate predictions of complex effects. Therefore, we develop a generative deep learning model based on the probabilistic diffusion framework to map low-fidelity simulation information to the high-fidelity counterpart. By doing so, we bypass the computational expense of conducting multiple high-fidelity simulations for analysis by upscaling lightweight coarse mesh simulations. We demonstrate the preservation of key metrics of the melting process between the ground truth simulation data and the diffusion model output, such as the temperature field, the melt pool dimensions and the variability of the keyhole vapor cavity. We predict the melt pool depth within 3 μm based on low-fidelity input data 4× coarser than the high-fidelity simulations, reducing analysis time by two orders of magnitude.

A Novel Poroelastography Method for High-quality Estimation of Lateral Strain, Solid Stress and Fluid Pressure In Vivo.

Assessment of mechanical and transport properties of tissues using ultrasound elasticity imaging requires accurate estimations of the spatiotemporal distribution of volumetric strain. Due to physical constraints such as pitch limitation and the lack of phase information in the lateral direction, the quality of lateral strain estimation is typically significantly lower than the quality of axial strain estimation. In this paper, a novel lateral strain estimation technique based on the physics of compressible porous media is developed, tested and validated. This technique is referred to as "Poroelastography-based Ultrasound Lateral Strain Estimation" (PULSE). PULSE differs from previously proposed lateral strain estimators as it uses the underlying physics of internal fluid flow within a local region of the tissue as theoretical foundation. PULSE establishes a relation between spatiotemporal changes in the axial strains and corresponding spatiotemporal changes in the lateral strains, effectively allowing assessment of lateral strains with comparable quality of axial strain estimators. We demonstrate that PULSE can also be used to accurately track compression-induced solid stresses and fluid pressure in cancers using ultrasound poroelastography (USPE). In this study, we report the theoretical formulation for PULSE and validation using finite element (FE) and ultrasound simulations. PULSE-generated results exhibit less than 5% percentage relative error (PRE) and greater than 90% structural similarity index (SSIM) compared to ground truth simulations. Experimental results are included to qualitatively assess the performance of PULSE in vivo. The proposed method can be used to overcome the inherent limitations of non-axial strain imaging and improve clinical translatability of USPE.

Surface-based anthropomorphic bone structures for use in high-resolution simulated medical imaging

Objective. Virtual imaging trials enable efficient assessment and optimization of medical image devices and techniques via simulation rather than physical studies. These studies require realistic, detailed ground-truth models or phantoms of the relevant anatomy or physiology. Anatomical structures within computational phantoms are typically based on medical imaging data; however, for small and intricate structures (e.g. trabecular bone), it is not reasonable to use existing clinical data as the spatial resolution of the scans is insufficient. In this study, we develop a mathematical method to generate arbitrary-resolution bone structures within virtual patient models (XCAT phantoms) to model the appearance of CT-imaged trabecular bone. Approach. Given surface definitions of a bone, an algorithm was implemented to generate stochastic bicontinuous microstructures to form a network to define the trabecular bone structure with geometric and topological properties indicative of the bone. For an example adult male XCAT phantom (50th percentile in height and weight), the method was used to generate the trabecular structure of 46 chest bones. The produced models were validated in comparison with published properties of bones. The utility of the method was demonstrated with pilot CT and photon-counting CT simulations performed using the accurate DukeSim CT simulator on the XCAT phantom containing the detailed bone models. Main results. The method successfully generated the inner trabecular structure for the different bones of the chest, having quantiative measures similar to published values. The pilot simulations showed the ability of photon-counting CT to better resolve the trabecular detail emphasizing the necessity for high-resolution bone models. Significance. As demonstrated, the developed tools have great potential to provide ground truth simulations to access the ability of existing and emerging CT imaging technology to provide quantitative information about bone structures.

Deep-LASI: deep-learning assisted, single-molecule imaging analysis of multi-color DNA origami structures

Single-molecule experiments have changed the way we explore the physical world, yet data analysis remains time-consuming and prone to human bias. Here, we introduce Deep-LASI (Deep-Learning Assisted Single-molecule Imaging analysis), a software suite powered by deep neural networks to rapidly analyze single-, two- and three-color single-molecule data, especially from single-molecule Förster Resonance Energy Transfer (smFRET) experiments. Deep-LASI automatically sorts recorded traces, determines FRET correction factors and classifies the state transitions of dynamic traces all in ~20–100 ms per trajectory. We benchmarked Deep-LASI using ground truth simulations as well as experimental data analyzed manually by an expert user and compared the results with a conventional Hidden Markov Model analysis. We illustrate the capabilities of the technique using a highly tunable L-shaped DNA origami structure and use Deep-LASI to perform titrations, analyze protein conformational dynamics and demonstrate its versatility for analyzing both total internal reflection fluorescence microscopy and confocal smFRET data.

Machine learning of hidden variables in multiscale fluid simulation

Solving fluid dynamics equations often requires the use of closure relations that account for missing microphysics. For example, when solving equations related to fluid dynamics for systems with a large Reynolds number, sub-grid effects become important and a turbulence closure is required, and in systems with a large Knudsen number, kinetic effects become important and a kinetic closure is required. By adding an equation governing the growth and transport of the quantity requiring the closure relation, it becomes possible to capture microphysics through the introduction of ‘hidden variables’ that are non-local in space and time. The behavior of the ‘hidden variables’ in response to the fluid conditions can be learned from a higher fidelity or ab-initio model that contains all the microphysics. In our study, a partial differential equation simulator that is end-to-end differentiable is used to train judiciously placed neural networks against ground-truth simulations. We show that this method enables an Euler equation based approach to reproduce non-linear, large Knudsen number plasma physics that can otherwise only be modeled using Boltzmann-like equation simulators such as Vlasov or particle-in-cell modeling.

Deep Q-learning to globally optimize a k-D parameter search for medical imaging.

Estimation of the global optima of multiple model parameters is valuable for precisely extracting parameters that characterize a physical environment. This is especially useful for imaging purposes, to form reliable, meaningful physical images with good reproducibility. However, it is challenging to avoid different local minima when the objective function is nonconvex. The problem of global searching of multiple parameters was formulated to be a k-D move in the parameter space and the parameter updating scheme was converted to be a state-action decision-making problem. We proposed a novel Deep Q-learning of Model Parameters (DQMP) method for global optimization which updated the parameter configurations through actions that maximized the Q-value and employed a Deep Reward Network (DRN) designed to learn global reward values from both visible fitting errors and hidden parameter errors. The DRN was constructed with Long Short-Term Memory (LSTM) layers followed by fully connected layers and a rectified linear unit (ReLU) nonlinearity. The depth of the DRN depended on the number of parameters. Through DQMP, the k-D parameter search in each step resembled the decision-making of action selections from 3k configurations in a k-D board game. The DQMP method was evaluated by widely used general functions that can express a variety of experimental data and further validated on imaging applications. The convergence of the proposed DRN was evaluated, which showed that the loss values of six general functions all converged after 12 epochs. The parameters estimated by the DQMP method had relative errors of less than 4% for all cases, whereas the relative errors achieved by Q-learning (QL) and the Least Squares Method (LSM) were 17% and 21%, respectively. Furthermore, the imaging experiments demonstrated that the imaging of the parameters estimated by the proposed DQMP method were the closest to the ground truth simulation images when compared to other methods. The proposed DQMP method was able to achieve global optima, thus yielding accurate model parameter estimates. DQMP is promising for estimating multiple high-dimensional parameters and can be generalized to global optimization for many other complex nonconvex functions and imaging of physical parameters.

Preclinical validation of the advection diffusion flow estimation method using computational patient specific coronary tree phantoms.

Coronary computed tomography angiography (CCTA) does not allow the quantification of reduced blood flow due to coronary artery disease (CAD). In response, numerical methods based on the CCTA image have been developed to compute coronary blood flow and assess the impact of disease. However to compute blood flow in the coronary arteries, numerical methods require specification of boundary conditions that are difficult to estimate accurately in a patient-specific manner. We describe herein a new noninvasive flow estimation method, called Advection Diffusion Flow Estimation (ADFE), to compute coronary artery flow from CCTA to use as boundary conditions for numerical models of coronary blood flow. ADFE uses image contrast variation along the tree-like structure to estimate flow in each vessel. For validating this method we used patient specific software phantoms on which the transport of contrast was simulated. This controlled validation setting enables a direct comparison between estimated flow and actual flow and a detailed investigation of factors affecting accuracy. A total of 10 CCTA image data sets were processed to extract all necessary information for simulating contrast transport. A spectral element method solver was used for computing the ground truth simulations with high accuracy. On this data set, the ADFE method showed a high correlation coefficient of between estimated flow and the ground truth flow together with an average relative error of only . Comparing the ADFE method with the best method currently available (TAFE) for image-based blood flow estimation, which showed a correlation coefficient of and average error of , it can be concluded that the ADFE method has the potential to significantly improve the quantification of coronary artery blood flow derived from contrast gradients in CCTA images.

MitoTNT: Mitochondrial Temporal Network Tracking for 4D live-cell fluorescence microscopy data.

Mitochondria form a network in the cell that rapidly changes through fission, fusion, and motility. Dysregulation of this four-dimensional (4D: x,y,z,time) network is implicated in numerous diseases ranging from cancer to neurodegeneration. While lattice light-sheet microscopy has recently made it possible to image mitochondria in 4D, quantitative analysis methods for the resulting datasets have been lacking. Here we present MitoTNT, the first-in-class software for Mitochondrial Temporal Network Tracking in 4D live-cell fluorescence microscopy data. MitoTNT uses spatial proximity and network topology to compute an optimal tracking assignment. To validate the accuracy of tracking, we created a reaction-diffusion simulation to model mitochondrial network motion and remodeling events. We found that our tracking is >90% accurate for ground-truth simulations and agrees well with published motility results for experimental data. We used MitoTNT to quantify 4D mitochondrial networks from human induced pluripotent stem cells. First, we characterized sub-fragment motility and analyzed network branch motion patterns. We revealed that the skeleton node motion is correlated along branch nodes and is uncorrelated in time. Second, we identified fission and fusion events with high spatiotemporal resolution. We found that mitochondrial skeleton nodes near the fission/fusion sites move nearly twice as fast as random skeleton nodes and that microtubules play a role in mediating selective fission/fusion. Finally, we developed graph-based transport simulations that model how material would distribute on experimentally measured mitochondrial temporal networks. We showed that pharmacological perturbations increase network reachability but decrease network resilience through a combination of altered mitochondrial fission/fusion dynamics and motility. MitoTNT's easy-to-use tracking module, interactive 4D visualization capability, and powerful post-tracking analyses aim at making temporal network tracking accessible to the wider mitochondria research community.

The Gradient-Boosting Method for Tackling High Computing Demand in Underwater Acoustic Propagation Modeling

Agent-based models return spatiotemporal information used to process time series of specific parameters for specific individuals called “agents”. For complex, advanced and detailed models, this typically comes at the expense of high computing times and requires access to important computing resources. This paper provides an example on how machine learning and artificial intelligence can help predict an agent-based model’s output values at regular intervals without having to rely on time-consuming numerical calculations. Gradient-boosting XGBoost under GNU package’s R was used in the social-ecological agent-based model 3MTSim to interpolate, in the time domain, sound pressure levels received at the agents’ positions that were occupied by the endangered St. Lawrence Estuary and Saguenay Fjord belugas and caused by anthropomorphic noise of nearby transiting merchant vessels. A mean error of 3.23 ± 3.76(1σ) dB on received sound pressure levels was predicted when compared to ground truth values that were processed using rigorous, although time-consuming, numerical algorithms. The computing time gain was significant, i.e., it was estimated to be 10-fold higher than the ground truth simulation, whilst maintaining the original temporal resolution.

Learning multiobjective rough terrain traversability

We present a method that uses high-resolution topography data of rough terrain, and ground vehicle simulation, to predict traversability. Traversability is expressed as three independent measures: the ability to traverse the terrain at a target speed, energy consumption, and acceleration. The measures are continuous and reflect different objectives for planning that go beyond binary classification. A deep neural network is trained to predict the traversability measures from the local heightmap and target speed. To produce training data, we use an articulated vehicle with wheeled bogie suspensions and procedurally generated terrains. We evaluate the model on laser-scanned forest terrains, previously unseen by the model. The model predicts traversability with an accuracy of 90%. Predictions rely on features from the high-dimensional terrain data that surpass local roughness and slope relative to the heading. Correlations show that the three traversability measures are complementary to each other. With an inference speed 3000 times faster than the ground truth simulation and trivially parallelizable, the model is well suited for traversability analysis and optimal path planning over large areas.

Combining endocardial mapping and electrocardiographic imaging (ECGI) for improving PVC localization: A feasibility study

IntroductionAccurate reconstruction of cardiac activation wavefronts is crucial for clinical diagnosis, management, and treatment of cardiac arrhythmias. Furthermore, reconstruction of activation profiles within the intramural myocardium has long been impossible because electrical mapping was only performed on the endocardial surface. Recent advancements in electrocardiographic imaging (ECGI) have made endocardial and epicardial activation mapping possible. We propose a novel approach to use both endocardial and epicardial mapping in a combined approach to reconstruct intramural activation times. ObjectiveTo implement and validate a combined epicardial/endocardial intramural activation time reconstruction technique. MethodsWe used 11 simulations of ventricular activation paced from sites throughout myocardial wall and extracted endocardial and epicardial activation maps at approximate clinical resolution. From these maps, we interpolated the activation times through the myocardium using thin-plate-spline radial basis functions. We evaluated activation time reconstruction accuracy using root-mean-squared error (RMSE) of activation times and the percent of nodes within 1 ms of the ground truth. ResultsReconstructed intramural activation times showed an RMSE and percentage of nodes within 1 ms of the ground truth simulations of 3 ms and 70%, respectively. In the worst case, the RMSE and percentage of nodes were 4 ms and 60%, respectively. ConclusionWe showed that a simple, yet effective combination of clinical endocardial and epicardial activation maps can accurately reconstruct intramural wavefronts. Furthermore, we showed that this approach provided robust reconstructions across multiple intramural stimulation sites.

Distant Collision Response in Rigid Body Simulations

AbstractWe use a finite element model to predict the vibration response of objects in a rigid body simulation, such that rigid objects are augmented to provide a plausible elastic collision response between distant objects due to vibration. We start with a generalized eigenvalue decomposition of the elastic model to precompute a response to an impact at any point on an elastic object with fixed boundary conditions. Then, given a collision between objects, we generate an approximate response impulse to distribute to other objects already in contact with the colliding bodies. This can lead to distant impacts causing an object to slip, or a delicate stack of objects to fall. We also use a geodesic distance based spatial attenuation approximation for travelling waves in objects to respond to an impact at one contact with an impulse at other locations. This response ultimately allows a long distance relationship between contacts, both across a single object being struck, but also traversing the contact graph of a larger collection of objects. We qualitatively validate our approach with a ground truth simulation, and demonstrate a number of scenarios where a long distance relationship between contacts is valuable.

Predictive regression modeling with MEG/EEG: from source power to signals and cognitive states

Predicting biomedical outcomes from Magnetoencephalography and Electroencephalography (M/EEG) is central to applications like decoding, brain-computer-interfaces (BCI) or biomarker development and is facilitated by supervised machine learning. Yet, most of the literature is concerned with classification of outcomes defined at the event-level. Here, we focus on predicting continuous outcomes from M/EEG signal defined at the subject-level, and analyze about 600 MEG recordings from Cam-CAN dataset and about 1000 EEG recordings from TUH dataset. Considering different generative mechanisms for M/EEG signals and the biomedical outcome, we propose statistically-consistent predictive models that avoid source-reconstruction based on the covariance as representation. Our mathematical analysis and ground-truth simulations demonstrated that consistent function approximation can be obtained with supervised spatial filtering or by embedding with Riemannian geometry. Additional simulations revealed that Riemannian methods were more robust to model violations, in particular geometric distortions induced by individual anatomy. To estimate the relative contribution of brain dynamics and anatomy to prediction performance, we propose a novel model inspection procedure based on biophysical forward modeling. Applied to prediction of outcomes at the subject-level, the analysis revealed that the Riemannian model better exploited anatomical information while sensitivity to brain dynamics was similar across methods. We then probed the robustness of the models across different data cleaning options. Environmental denoising was globally important but Riemannian models were strikingly robust and continued performing well even without preprocessing. Our results suggest each method has its niche: supervised spatial filtering is practical for event-level prediction while the Riemannian model may enable simple end-to-end learning.

Energy management of a thermally coupled fuel cell system and metal hydride tank

Being produced from renewable energy, hydrogen is one of the most efficient energy carriers of the future. Using metal alloys, hydrogen can be stored and transported at a low cost, in a safe and effective manner. However, most metals react with hydrogen to form a compound called metal hydride (MH). This reaction is an exothermic process, and as a result releases heat. With sufficient heat supply, hydrogen can be released from the as-formed metal hydride. In this work, we propose an integrated power system of a proton exchange membrane fuel cell (PEMFC) together with a hydride tank designed for vehicle use. We investigate different aspects for developing metal hydride tanks and their integration in the PEMFC, using water as the thermal fluid and a FeTi intermetallic compound as the hydrogen storage material. Ground truth simulations show that the annular metal hydride tank meets the hydrogen requirements of the fuel cell, but to the detriment of the operating temperature of the fuel cell (FC).

A learned shape-adaptive subsurface scattering model

Subsurface scattering, in which light refracts into a translucent material to interact with its interior, is the dominant mode of light transport in many types of organic materials. Accounting for this phenomenon is thus crucial for visual realism, but explicit simulation of the complex internal scattering process is often too costly. BSSRDF models based on analytic transport solutions are significantly more efficient but impose severe assumptions that are almost always violated, e.g. planar geometry, isotropy, low absorption, and spatio-directional separability. The resulting discrepancies between model and usage lead to objectionable errors in renderings, particularly near geometric features that violate planarity. This article introduces a new shape-adaptive BSSRDF model that retains the efficiency of prior analytic methods while greatly improving overall accuracy. Our approach is based on a conditional variational autoencoder, which learns to sample from a reference distribution produced by a brute-force volumetric path tracer. In contrast to the path tracer, our autoencoder directly samples outgoing locations on the object surface, bypassing a potentially lengthy internal scattering process. The distribution is conditional on both material properties and a set of features characterizing geometric variation in a neighborhood of the incident location. We use a low-order polynomial to model the local geometry as an implicitly defined surface, capturing curvature, thickness, corners, as well as cylindrical and toroidal regions. We present several examples of objects with challenging medium parameters and complex geometry and compare to ground truth simulations and prior work.

Controlling Motion Blur in Synthetic Long Time Exposures

AbstractIn a photo, motion blur can be used as an artistic style to convey motion and to direct attention. In panning or tracking shots, a moving object of interest is followed by the camera during a relatively long exposure. The goal is to get a blurred background while keeping the object sharp. Unfortunately, it can be difficult to impossible to precisely follow the object. Often, many attempts or specialized physical setups are needed.This paper presents a novel approach to create such images. For capturing, the user is only required to take a casually recorded hand‐held video that roughly follows the object. Our algorithm then produces a single image which simulates a stabilized long time exposure. This is achieved by first warping all frames such that the object of interest is aligned to a reference frame. Then, optical flow based frame interpolation is used to reduce ghosting artifacts from temporal undersampling. Finally, the frames are averaged to create the result.As our method avoids segmentation and requires little to no user interaction, even challenging sequences can be processed successfully. In addition, artistic control is available in a number of ways. The effect can also be applied to create videos with an exaggerated motion blur. Results are compared with previous methods and ground truth simulations. The effectiveness of our method is demonstrated by applying it to hundreds of datasets. The most interesting results are shown in the paper and in the supplemental material.

A volumetric deep Convolutional Neural Network for simulation of mock dark matter halo catalogues

For modern large-scale structure survey techniques it has become standard practice to test data analysis pipelines on large suites of mock simulations, a task which is currently prohibitively expensive for full N-body simulations. Instead of calculating this costly gravitational evolution, we have trained a three-dimensional deep Convolutional Neural Network (CNN) to identify dark matter protohalos directly from the cosmological initial conditions. Training on halo catalogues from the Peak Patch semi-analytic code, we test various CNN architectures and find they generically achieve a Dice coefficient of ~92% in only 24 hours of training. We present a simple and fast geometric halo finding algorithm to extract halos from this powerful pixel-wise binary classifier and find that the predicted catalogues match the mass function and power spectra of the ground truth simulations to within ~10%. We investigate the effect of long-range tidal forces on an object-by-object basis and find that the network's predictions are consistent with the non-linear ellipsoidal collapse equations used explicitly by the Peak Patch algorithm.