13 results
Now showing1 - 10 of 13
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , A review of the application of machine learning and data mining approaches in continuum materials mechanicsMachine learning tools represent key enablers for empowering material scientists and engineers to accelerate the development of novel materials, processes and techniques. One of the aims of using such approaches in the field of materials science is to achieve high-throughput identification and quantification of essential features along the process-structure-property-performance chain. In this contribution, machine learning and statistical learning approaches are reviewed in terms of their successful application to specific problems in the field of continuum materials mechanics. They are categorized with respect to their type of task designated to be either descriptive, predictive or prescriptive; thus to ultimately achieve identification, prediction or even optimization of essential characteristics. The respective choice of the most appropriate machine learning approach highly depends on the specific use-case, type of material, kind of data involved, spatial and temporal scales, formats, and desired knowledge gain as well as affordable computational costs. Different examples are reviewed involving case-by-case dependent application of different types of artificial neural networks and other data-driven approaches such as support vector machines, decision trees and random forests as well as Bayesian learning, and model order reduction procedures such as principal component analysis, among others. These techniques are applied to accelerate the identification of material parameters or salient features for materials characterization, to support rapid design and optimization of novel materials or manufacturing methods, to improve and correct complex measurement devices, or to better understand and predict fatigue behavior, among other examples. Besides experimentally obtained datasets, numerous studies draw required information from simulation-based data mining. Altogether, it is shown that experiment- and simulation-based data mining in combination with machine leaning tools provide exceptional opportunities to enable highly reliant identification of fundamental interrelations within materials for characterization and optimization in a scale-bridging manner. Potentials of further utilizing applied machine learning in materials science and empowering significant acceleration of knowledge output are pointed out.Publicationtype: Journal ArticleScopus© Citations 240 431 705 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , General multi-fidelity framework for training artificial neural networks with computational models(2019-04-17); ; Training of artificial neural networks (ANNs) relies on the availability of training data. If ANNs have to be trained to predict or control the behavior of complex physical systems, often not enough real-word training data are available, for example, because experiments or measurements are too expensive, time-consuming or dangerous. In this case, generating training data by way of realistic computational simulations is a viable and often the only promising alternative. Doing so can, however, be associated with a significant and often even prohibitive computational cost, which forms a serious bottleneck for the application of machine learning to complex physical systems. To overcome this problem, we propose in this paper a both systematic and general approach. It uses cheap low-fidelity computational models to start the training of the ANN and gradually switches to higher-fidelity training data as the training of the ANN progresses. We demonstrate the benefits of this strategy using examples from structural and materials mechanics. We demonstrate that in these examples the multi-fidelity strategy introduced herein can reduce the total computational cost–compared to simple brute-force training of ANNs–by a half up to one order of magnitude. This multi-fidelity strategy can thus be hoped to become a powerful and versatile tool for the future combination of computational simulations and artificial intelligence, in particular in areas such as structural and materials mechanics.Publicationtype: Journal ArticleScopus© Citations 25 471 421 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Anisotropic stiffness and tensional homeostasis induce a natural anisotropy of volumetric growth and remodeling in soft biological tissues(2019-04-15); ; Growth in soft biological tissues in general results in anisotropic changes of the tissue geometry. It remains a key challenge in biomechanics to understand, quantify, and predict this anisotropy. In this paper, we demonstrate that anisotropic tissue stiffness and the well-known mechanism of tensional homeostasis induce a natural anisotropy of the geometric changes resulting from volumetric growth in soft biological tissues. As a rule of thumb, this natural anisotropy makes differential tissue volume elements dilate mainly in the direction(s) of lowest stiffness. This simple principle is shown to explain the experimentally observed growth behavior in a host of different soft biological tissues without relying on any additional heuristic assumptions or quantities (such as ad hoc defined growth tensors).Publicationtype: Journal Article159 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Hematopoietic stem cell transplantation as treatment for patients with DOCK8 deficiencyBackground: Biallelic variations in the dedicator of cytokinesis 8 (DOCK8) gene cause a combined immunodeficiency with eczema, recurrent bacterial and viral infections, and malignancy. Natural disease outcome is dismal, but allogeneic hematopoietic stem cell transplantation (HSCT) can cure the disease. Objective: To determine outcome of HSCT for DOCK8 deficiency and define possible outcome variables. Methods: We performed a retrospective study of the results of HSCT in a large international cohort of DOCK8-deficient patients. Results: We identified 81 patients from 22 centers transplanted at a median age of 9.7 years (range, 0.7-27.2 years) between 1995 and 2015. After median follow-up of 26 months (range, 3-135 months), 68 (84%) patients are alive. Severe acute (III-IV) or chronic graft versus host disease occurred in 11% and 10%, respectively. Causes of death were infections (n = 5), graft versus host disease (5), multiorgan failure (2), and preexistent lymphoma (1). Survival after matched related (n = 40) or unrelated (35) HSCT was 89% and 81%, respectively. Reduced-toxicity conditioning based on either treosulfan or reduced-dose busulfan resulted in superior survival compared with fully myeloablative busulfan-based regimens (97% vs 78%; P =.049). Ninety-six percent of patients younger than 8 years at HSCT survived, compared with 78% of those 8 years and older (P =.06). Of the 73 patients with chimerism data available, 65 (89%) had more than 90% donor T-cell chimerism at last follow-up. Not all disease manifestations responded equally well to HSCT: eczema, infections, and mollusca resolved quicker than food allergies or failure to thrive. Conclusions: HSCT is curative in most DOCK8-deficient patients, confirming this approach as the treatment of choice. HSCT using a reduced-toxicity regimen may offer the best chance for survival.Publicationtype: Journal Article16 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Mechanics of the stomach: a review of an emerging field of biomechanics(Wiley-VCH, 2019-02-27); ; ; Mathematical and computational modeling of the stomach is an emerging field of biomechanics where several complex phenomena, such as gastric electrophysiology, fluid mechanics of the digesta, and solid mechanics of the gastric wall, need to be addressed. Developing a comprehensive multiphysics model of the stomach that allows studying the interactions between these phenomena remains one of the greatest challenges in biomechanics. A coupled multiphysics model of the human stomach would enable detailed in-silico studies of the digestion of food in the stomach in health and disease. Moreover, it has the potential to open up unprecedented opportunities in numerous fields such as computer-aided medicine and food design. This review article summarizes our current understanding of the mechanics of the human stomach and delineates the challenges in mathematical and computational modeling which remain to be addressed in this emerging area.Publicationtype: Journal ArticleScopus© Citations 57 257 419 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Computational model of gastric motility with active-strain electromechanics(Wiley-VCH, 2018-12-01); ; ; ; We present an electro-mechanical constitutive framework for the modeling of gastric motility, including pacemaker electrophysiology and smooth muscle contractility. In this framework, we adopt a phenomenological description of the gastric tissue. Tissue electrophysiology is represented by a set of two minimal two-variable models and tissue electromechanics by an active-strain finite elasticity approach. We numerically investigate the implication of the spatial distribution of pacemaker cells on the entrainment and synchronization of the slow waves characterizing gastric motility in health and disease. On simple schematic model geometries, we demonstrate that the proposed computational framework is amenable to large scale in-silico analyses of the complex gastric motility including the underlying electro-mechanical coupling.Publicationtype: Journal ArticleScopus© Citations 20 347 430 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Experimental characterization of the biaxial mechanical properties of porcine gastric tissue(Elsevier, 2017-10); ; ; ; Health problems related to the stomach are among the most important sources of morbidity in industrialized countries. There is evidence that mechanics may play an important role in various such pathologies. However, so far experimental data characterizing the mechanical properties of gastric tissue remain scarce, which significantly limits our understanding of the mechanics of the stomach. To help close this gap, we performed biaxial mechanical tests of porcine gastric tissue patches. Our experiments reveal a considerable anisotropy and different mechanical properties in the three major regions of the stomach (fundus, corpus, antrum). Moreover, they demonstrate that the mechanical properties of the gastric wall and the physiological function of the different regions of the stomach are closely related. This finding suggests that further examination of the mechanics of the gastric wall may indeed be a promising avenue of research towards a better understanding of the organic causes of frequent health problems related to the stomach.Publicationtype: Journal Article15 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Higher patient satisfaction with antidepressants correlates with earlier drug release dates across online user-generated medical databases(2017-10); ; ; Studies establishing the use of new antidepressants often rely simply on proving efficacy of a new compound, comparing against placebo and single compound. The advent of large online databases in which patients themselves rate drugs allows for a new Big Data–driven approach to compare the efficacy and patient satisfaction with sample sizes exceeding previous studies. Exemplifying this approach with antidepressants, we show that patient satisfaction with a drug anticorrelates with its release date with high significance, across different online user-driven databases. This finding suggests that a systematic reevaluation of current, often patent-protected drugs compared to their older predecessors may be helpful, especially given that the efficacy of newer agents relative to older classes of antidepressants such as monoamine oxidase inhibitors (MAOIs) and tricyclic antidepressants (TCAs) is as yet quantitatively unexplored.Publicationtype: Journal Article29 27 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Mechanobiological free energy : a variational approach to tensional homeostasis in tissue equivalents(2017-09); Classical engineering materials typically seek to maintain or attain a relaxed, stress-free state, if admitted by the boundary conditions. By contrast, living tissues seek to establish and maintain a certain non-zero target stress, a behavior that is often referred to as tensional homeostasis. Although a growing body of experimental evidence is underpinning the importance of tensional homeostasis for the biomechanics of living tissues, its theoretical implications remain poorly understood. Motivated by tissue culture studies, this paper demonstrates that tensional homeostasis has significant impact on the way how the notion of free energy should be conceptualized in living tissues. It demonstrates that standard variational approaches from continuum mechanics and engineering mechanics such as the principle of minimal total mechanical potential are in practice often insufficient to capture even the qualitative mechanical behavior of living tissues. To overcome this problem, this paper introduces, under assumptions that reflect the situation in tissue culture experiments, the novel concept of mechanobiological free energy. Mechanobiological free energy is shown to share the essential favorable properties of other free energy concepts such as Helmholtz free energy or strain energy but to generalize them in a way suitable to understand mechanobiological remodeling in collagenous tissues inhabited by living cells. In particular, mechanobiological free energy gives rise to a simple variational approach to tensional homeostasis in tissue equivalents.Publicationtype: Journal Article8 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Homogenized constrained mixture models for anisotropic volumetric growth and remodeling(2017-06-01); ; ; Constrained mixture models for soft tissue growth and remodeling have attracted increasing attention over the last decade. They can capture the effects of the simultaneous presence of multiple constituents that are continuously deposited and degraded at in general different rates, which is important to understand essential features of living soft tissues that cannot be captured by simple kinematic growth models. Recently the novel concept of homogenized constrained mixture models was introduced. It was shown that these models produce results which are very similar (and in certain limit cases even identical) to the ones of constrained mixture models based on multi-network theory. At the same time, the computational cost and complexity of homogenized constrained mixture models are much lower. This paper discusses the theory and implementation of homogenized constrained mixture models for anisotropic volumetric growth and remodeling in three dimensions. Previous constrained mixture models of volumetric growth in three dimensions were limited to the special case of isotropic growth. By numerical examples, comparison with experimental data and a theoretical discussion, we demonstrate that there is some evidence raising doubts whether isotropic growth models are appropriate to represent growth and remodeling of soft tissue in the vasculature. Anisotropic constrained mixture models, as introduced in this paper for the first time, may be required to avoid unphysiological results in simulations of vascular growth and remodeling.Publicationtype: Journal Article15