اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدMechanics Informed Fluoroscopy of Esophageal Transport
112
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Fluoroscopy is a radiographic procedure for evaluating esophageal disorders such as achalasia, dysphasia and gastroesophageal reflux disease (GERD). It performs dynamic imaging of the swallowing process and provides anatomical detail and a qualitative idea of how well swallowed fluid is transported through the esophagus. In this work, we present a method called mechanics informed fluoroscopy (FluoroMech) that derives patient-specific quantitative information about esophageal function. FluoroMech uses a Convolutional Neural Network to perform segmentation of image sequences generated from the fluoroscopy, and the segmented images become input to a one-dimensional model that predicts the flow rate and pressure distribution in fluid transported through the esophagus. We have extended this model by developing a FluoroMech reference model to identify and estimate potential physiomarkers such as esophageal wall stiffness and active relaxation ahead of the peristaltic wave in the esophageal musculature. FluoroMech requires minimal computational time, and hence can potentially be applied clinically in the diagnosis of esophageal disorders.
قيم البحث
اقرأ أيضاً
In this work, we extend our previous esophageal transport model using an immersed boundary (IB) method with discrete fiber-based structures, to one using a continuum mechanics-based model that is approximated based on finite elements (IB-FE). To deal
with the leakage of flow when the Lagrangian mesh becomes coarser than the fluid mesh, we employ adaptive interaction quadrature points for Lagrangian-Eulerian interaction equations based on a previous work. In particular, we introduce a new anisotropic adaptive interaction quadrature rule. The new rule permits us to vary the interaction quadrature points not only at each time-step and element but also at different orientations per element. For the material model, we extend our previous fiber-based model to a continuum-based model. We first study a case in which a three-dimensional short tube is dilated. Results match very well with those from the implicit FE method. We remark that in our IB-FE case, the three-dimensional tube undergoes a very large deformation and the Lagrangian mesh-size becomes about 6 times of Eulerian mesh-size. To validate the method in handling fiber-matrix material models, we perform a second study on dilating a long fiber-reinforced tube. Errors are small when we compare numerical solutions with analytical solutions. The technique is then applied to the problem of esophageal transport. We present three cases that differ in the material model and muscle fiber architecture. The overall transport features are consistent with those from the previous model. We remark that the continuum-based model can handle more realistic and complicated material behavior. This is demonstrated in our third case with spatially varying fiber architecture. We find this unique muscle fiber architecture could generate a so-called pressure transition zone. This suggests an important role of muscle fiber architecture in esophageal transport.
Machine learning-based approaches outperform competing methods in most disciplines relevant to diagnostic radiology. Interventional radiology, however, has not yet benefited substantially from the advent of deep learning, in particular because of two
reasons: 1) Most images acquired during the procedure are never archived and are thus not available for learning, and 2) even if they were available, annotations would be a severe challenge due to the vast amounts of data. When considering fluoroscopy-guided procedures, an interesting alternative to true interventional fluoroscopy is in silico simulation of the procedure from 3D diagnostic CT. In this case, labeling is comparably easy and potentially readily available, yet, the appropriateness of resulting synthetic data is dependent on the forward model. In this work, we propose DeepDRR, a framework for fast and realistic simulation of fluoroscopy and digital radiography from CT scans, tightly integrated with the software platforms native to deep learning. We use machine learning for material decomposition and scatter estimation in 3D and 2D, respectively, combined with analytic forward projection and noise injection to achieve the required performance. On the example of anatomical landmark detection in X-ray images of the pelvis, we demonstrate that machine learning models trained on DeepDRRs generalize to unseen clinically acquired data without the need for re-training or domain adaptation. Our results are promising and promote the establishment of machine learning in fluoroscopy-guided procedures.
Numerical models are increasingly used for non-invasive diagnosis and treatment planning in coronary artery disease, where service-based technologies have proven successful in identifying hemodynamically significant and hence potentially dangerous va
scular anomalies. Despite recent progress towards clinical adoption, many results in the field are still based on a deterministic characterization of blood flow, with no quantitative assessment of the variability of simulation outputs due to uncertainty from multiple sources. In this study, we focus on parameters that are essential to construct accurate patient-specific representations of the coronary circulation, such as aortic pressure waveform, intramyocardial pressure and quantify how their uncertainty affects clinically relevant model outputs. We construct a deformable model of the left coronary artery subject to a prescribed inlet pressure and with open-loop outlet boundary conditions, treating fluid-structure interaction through an Arbitrary-Lagrangian-Eulerian frame of reference. Random input uncertainty is estimated directly from repeated clinical measurements from intra-coronary catheterization and complemented by literature data. We also achieve significant computational cost reductions in uncertainty propagation thanks to multifidelity Monte Carlo estimators of the outputs of interest, leveraging the ability to generate, at practically no cost, one- and zero-dimensional low-fidelity representations of left coronary artery flow, with appropriate boundary conditions. The results demonstrate how the use of multi-fidelity control variate estimators leads to significant reductions in variance and accuracy improvements with respect to traditional Monte-Carlo.
Despite the significant progress over the last 50 years in simulating flow problems using numerical discretization of the Navier-Stokes equations (NSE), we still cannot incorporate seamlessly noisy data into existing algorithms, mesh-generation is co
mplex, and we cannot tackle high-dimensional problems governed by parametrized NSE. Moreover, solving inverse flow problems is often prohibitively expensive and requires complex and expensive formulations and new computer codes. Here, we review flow physics-informed learning, integrating seamlessly data and mathematical models, and implementing them using physics-informed neural networks (PINNs). We demonstrate the effectiveness of PINNs for inverse problems related to three-dimensional wake flows, supersonic flows, and biomedical flows.
Quantization together with quantum dynamics can be simultaneously formulated as the problem of finding an appropriate flat connection on a Hilbert bundle over a contact manifold. Contact geometry treats time, generalized positions and momenta as poin
ts on an underlying phase-spacetime and reduces classical mechanics to contact topology. Contact quantization describes quantum dynamics in terms of parallel transport for a flat connection; the ultimate goal being to also handle quantum systems in terms of contact topology. Our main result is a proof of local, formal gauge equivalence for a broad class of quantum dynamical systems-just as classical dynamics depends on choices of clocks, local quantum dynamics can be reduced to a problem of studying gauge transformations. We further show how to write quantum correlators in terms of parallel transport and in turn matrix elements for Hilbert bundle gauge transformations, and give the path integral formulation of these results. Finally, we show how to relate topology of the underlying contact manifold to boundary conditions for quantum wave functions.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
