Are hemodynamics responsible for inflammatory changes in venous vessel walls? : a quantitative study of wall-enhancing intracranial arteriovenous malformation draining veins. - In: Journal of neurosurgery, ISSN 1933-0693, Bd. 0 (2024), 0, S. 1-10
Objective: Signal enhancement of vascular walls on vessel wall MRI might be a biomarker for inflammation. It has been theorized that contrast enhancement on vessel wall imaging (VWI) in draining veins of intracranial arteriovenous malformations (AVMs) may be associated with disease progression and development of venous stenosis. The aim of this study was to investigate the relationship between vessel wall enhancement and hemodynamic stressors along AVM draining veins. Methods: Eight AVM patients with 15 draining veins visualized on VWI were included. Based on MR venography data, patient-specific 3D surface models of the venous anatomy distal to the nidus were segmented. The enhanced vascular wall regions were manually extracted and mapped onto the venous surface models after registration of image data. Using image-based blood flow simulations applying patient-specific boundary conditions based on phase-contrast quantitative MR angiography, hemodynamics were investigated in the enhanced vasculature. For the shear-related parameters, time-averaged wall shear stress (TAWSS), oscillatory shear index (OSI), and relative residence time (RRT) were calculated. Velocity, oscillatory velocity index (OVI), and vorticity were extracted for the intraluminal flow-related hemodynamics. Results: Visual observations demonstrated overlap of enhancement with local lower shear stresses resulting from decreased velocities. Thus, higher RRT values were measured in the enhanced areas. Furthermore, nonenhancing draining veins showed on average slightly higher flow velocities and TAWSS. Significant decreases of 55% (p = 0.03) for TAWSS and of 24% (p = 0.03) for vorticity were identified in enhanced areas compared with near distal and proximal domains. Velocity magnitude in the enhanced region showed a nonsignificant decrease of 14% (p = 0.06). Furthermore, increases were present in the OSI (32%, p = 0.3), RRT (25%, p = 0.15), and OVI (26%, p = 0.3) in enhanced vessel sections, although the differences were not significant. Conclusions: This novel multimodal investigation of hemodynamics in AVM draining veins allows for precise prediction of occurring shear- and flow-related phenomena in enhanced vessel walls. These findings may suggest low shear to be a local predisposing factor for venous stenosis in AVMs.
https://doi.org/10.3171/2024.1.JNS232625
Deontic knowledge representation and reasoning in industrial accident prevention training by means of time travel prevention games. - In: International journal of advanced corporate learning, ISSN 1867-5565, Bd. 17 (2024), 2, S. 4-16
Industrial accident prevention is an issue of societal relevance to avoid loss of human lives, injuries, damage of installations, and financial losses. The authors deploy game-based training in virtual environments where trainees experience challenges of safe operation and disastrous self-induced accidents. Nothing is more affective and, thus, effective than a trainee’s own experience. Time travel prevention games are a game category particularly tailored to the needs of human players who look for opportunities to make good for a damage. Time travel pre-vention games for purposes such as accident prevention in the industries are ad-vantageous due to their conservation of resources including human health and lives. They are affective by allowing for unprecedented learner/player/trainee ex-periences and they are effective due to the fascination of application-oriented game play including opportunities to influence the fate, the latter being less close to reality, but the more attractive and worth telling. For optimal guidance to human trainees, the digital game system needs to learn about the trainees’ strength and weaknesses, about needs and desires. In terms of behavioral sciences, the system observing a human’s behavior hypothesizes theories of mind. In training games, modalities of events/actions are decisive. There are modalities of events/actions such as possibility, unavoidability, and the like as well as obliga-tions and oughts. Training aims at the emergence of cognitive states that are use-ful in practice. The system’s reasoning is deontic.
https://doi.org/10.3991/ijac.v17i2.42975
APP-CEP: adaptive pattern-level privacy protection in complex event processing systems. - In: Proceedings of the 10th International Conference on Information Systems Security and Privacy, Volume 1, (2024), S. 486-497
Although privacy-preserving mechanisms endeavor to safeguard sensitive information at the attribute level, detected event patterns can still disclose privacy-sensitive knowledge in distributed complex event processing systems (DCEP). Events might not be inherently sensitive, but their aggregation into a pattern could still breach privacy. In this paper, we study in the context of APP-CEP the problem of integrating pattern-level privacy in event-based systems by selective assignment of obfuscation techniques to conceal private information. Compared to state-of-the-art techniques, we seek to enforce privacy independent of the actual events in streams. To support this, we acquire queries and privacy requirements using CEP-like patterns. The protection of privacy is accomplished through generating pattern dependency graphs, leading to dynamically appointing those techniques that have no consequences on detecting other sensitive patterns, as well as non-sensitive patterns required to prov ide acceptable Quality of Service. Besides, we model the knowledge that might be possessed by potential adversaries to violate privacy and its impacts on the obfuscation procedure. We assessed the performance of APP-CEP in a real-world scenario involving an online retailer’s transactions. Our evaluation results demonstrate that APP-CEP successfully provides a privacy-utility trade-off. Modeling the background knowledge also effectively prevents adversaries from realizing the modifications in the input streams.
https://doi.org/10.5220/0012358700003648
Increasing resilience of SD-WAN by distributing the control plane [extended version]. - In: IEEE transactions on network and service management, ISSN 1932-4537, (2024), S. 1-13
Modern WAN interconnects utilize SD-WAN to automatically respond to network changes and improve link utilization, latency, and availability. Therefore, they incorporate controllers with a centralized view, which collect network state from managed gateways, calculate suitable forwarding actions, and distribute them accordingly. However, this limits the robustness and availability of the network control plane, especially in the event of node or partial network outages. In this paper, we propose a distributed and highly robust SD-WAN control plane without any central or regional controller. Our solution can handle arbitrary device failures as well as network partitioning. The distributed forwarding decisions are based on user-defined, dynamically evaluated path cost functions, and consider not only path quality but also quality fluctuations. The evaluation shows that our approach can handle several thousand SD-WAN gateways and hundreds of network policies in terms of computation. Further, the communication overhead introduced due to its distributed architecture is discussed and shown to be negligible compared to a central approach. This paper is an extended version of our work published in altheide2023. It describes the information transmitted between sites as well as a strategy for deploying policies, discusses approaches reducing communication bandwidth, introduces grouping of multiple flows without requiring explicit coordination, and provides a detailed analysis of the bandwidth required.
https://doi.org/10.1109/TNSM.2024.3386962
An analysis and simulation framework for systems with classification components. - Setúbal : Scitepress. - 1 Online-Ressource (Seite 50-61)Online-Ausgabe: Proceedings of the 12th International Conference on Model-Based Software and Systems Engineering : February 21-23, 2024, in Rome, Italy / editors: Francisco José Domínguez Mayo, Luís Ferreira Pires and Edwin Seidewitz. - Setúbal : Scitepress, 2024, ISBN 978-989-758-682-8
Machine learning solutions are becoming more widespread as they can solve some classes of problems better than traditional software. Hence, industries look forward to integrating this new technology into their products and workflows. However, this calls for new models and analysis concepts in systems design that can incorporate the properties and effects of machine learning components. In this paper, we propose a framework that allows designing, analyzing, and simulating hardware-software systems that contain deep learning classification components. We focus on the modeling and predicting uncertainty aspects, which are typical for machine-learning applications. They may lead to incorrect results that may negatively affect the entire system’s dependability, reliability, and even safety. This issue is receiving increasing attention as “explain-able” or “certifiable” AI. We propose a Domain-Specific Language with a precise stochastic colored Petri net semantics to model such systems, which then can be simulated and analyzed to compute performance and reliability measures. The language is extensible and allows adding parameters to any of its elements, supporting the definition of additional analysis methods for future modular extensions.
https://doi.org/10.5220/0012357000003645
An exact formula for the hazardous event frequency. - In: IEEE Xplore digital library, ISSN 2473-2001, (2024), insges. 6 S.
The remaining risk of safety-instrumented systems is an important non-functional requirement that is regulated by international standards. Several ways towards computing the safety as a function of its relevant design parameters have been studied in the literature. However, the standard approach only covers two special cases of high or low demand, which simplify the treatment by either ignoring the effects of demand rate or test interval on the safety. More detailed treatments in the literature derive Markov models, which can be numerically analyzed, or approximate solutions using Taylor series expansions etc. This paper introduces closed-form exact formulas for the average probability of failure on demand (PFD) and the resulting hazardous event frequency (HEF, or accident rate), taking into account demand rate and test interval. It integrates all cases of low, high and medium demand in one formula. The derivation is based on an analysis of the cyclostationary semi-Markov stochastic process of the safety-integrated system and its symbolic transient analysis over the test interval.
https://doi.org/10.1109/RAMS51492.2024.10457804
Enhancing person perception for mobile robotics by real-time RGB-D person attribute estimation. - In: IEEE Xplore digital library, ISSN 2473-2001, (2024), S. 914-921
Person attribute estimation is a task of great importance for a variety of real-world robotic applications. While the computer vision community has made impressive progress over the last decade, they often rely on the sole use of RGB images from surveillance datasets and large deep- learning models without considering real-time requirements. By contrast, mobile robotic platforms have to deal with restricted resources but are often equipped with RGB-D cameras, offering complementary modalities. This paper presents an approach to robustly estimate soft-biometric attributes from full-body RGB-D appearances of persons in the surroundings of a mobile robot. The effects of depth, RGB and RGB-D data as input are analyzed, taking into account runtime and resource requirements. On the robotic attribute dataset SRL, it is shown that the presented approach outperforms other state-of-the-art approaches by a large margin. Furthermore, real-time requirements are met when applied on an NVIDIA Jetson AGX Xavier or even on a mobile CPU only. Finally, a previous system for person detection, upper body orientation estimation, and posture classification is integrated to enable an even more comprehensive perception of persons in the surroundings of a mobile robot in one joint approach. The source code for training and application will be made publicly available on GitHub.
https://doi.org/10.1109/SII58957.2024.10417431
Association of genetic risk for age-related macular degeneration with morphological features of the retinal microvascular network. - In: Diagnostics, ISSN 2075-4418, Bd. 14 (2024), 7, 770, S. 1-13
Background: Age-related macular degeneration (AMD) is a multifactorial disease encompassing a complex interaction between aging, environmental risk factors, and genetic susceptibility. The study aimed to determine whether there is a relationship between the polygenic risk score (PRS) in patients with AMD and the characteristics of the retinal vascular network visualized by optical coherence tomography angiography (OCTA). Methods: 235 patients with AMD and 97 healthy controls were included. We used data from a previous AMD PRS study with the same group. The vascular features from different retina layers were compared between the control group and the patients with AMD. The association between features and PRS was then analyzed using univariate and multivariate approaches. Results: Significant differences between the control group and AMD patients were found in the vessel diameter distribution (variance: p = 0.0193, skewness: p = 0.0457) and fractal dimension distribution (mean: p = 0.0024, variance: p = 0.0123). Both univariate and multivariate analyses showed no direct and significant association between the characteristics of the vascular network and AMD PRS. Conclusions: The vascular features of the retina do not constitute a biomarker of the risk of AMD. We have not identified a genotype-phenotype relationship, and the expression of AMD-related genes is perhaps not associated with the characteristics of the retinal vascular network.
https://doi.org/10.3390/diagnostics14070770
An end-to-end machine learning approach with explanation for time series with varying lengths. - In: Neural computing & applications, ISSN 1433-3058, Bd. 36 (2024), 13, S. 7491-7508
An accurate prediction of complex product quality parameters from process time series by an end-to-end learning approach remains a significant challenge in machine learning. A special difficulty is the application of industrial batch process data because many batch processes generate variable length time series. In the industrial application of such methods, explainability is often desired. In this study, a 1D convolutional neural network (CNN) algorithm with a masking layer is proposed to solve the problem for time series of variable length. In addition, a novel combination of 1D CNN and class activation mapping (CAM) technique is part of this study to better understand the model results and highlight some regions of interest in the time series. As a comparative state-of-the-art unsupervised machine learning method, the One-Nearest Neighbours (1NN) algorithm combined with dynamic time warping (DTW) was used. Both methods are investigated as end-to-end learning methods with balanced and unbalanced class distributions and with scaled and unscaled input data, respectively. The FastDTW and DTAIDistance algorithms were investigated for the DTW calculation. The data set is made up of sensor signals that was collected during the production of plastic parts. The objective was to predict a quality parameter of plastic parts during production. For this research, the quality parameter will be a difficult or only destructively measurable parameter and both methods will be investigated for their applicability to this prediction task. The application of the proposed approach to an industrial facility for producing plastic products shows a prediction accuracy of 83.7%. It can improve the reverence method by approximately 1.4%. In addition to the slight increase in accuracy, the CNN training time was significantly reduced compared to the DTW calculation.
https://doi.org/10.1007/s00521-024-09473-9
Modified implicit discretization of the super-twisting controller. - In: IEEE transactions on automatic control, ISSN 1558-2523, Bd. 0 (2024), 0, S. 1-8
In this paper a novel discrete-time realization of the super-twisting controller is proposed. The closed-loop system is proven to converge to an invariant set around the origin in finite time. Furthermore, the steady-state error is shown to be independent of the controller gains. It only depends on the sampling time and the unknown disturbance. The proposed discrete-time controller is evaluated comparative to previously published discrete-time super-twisting controllers by means of the controller structure and in extensive simulation studies. The continuous-time super-twisting controller is capable of rejecting any unknown Lipschitz-continuous perturbation and converges in finite time. Furthermore, the convergence time decreases, if any of the gains is increased. The simulations demonstrate that the closed-loop systems with each of the known controllers lose one of these properties, introduce discretization-chattering, or do not yield the same accuracy level as with the proposed controller. The proposed controller, in contrast, is beneficial in terms of the above described properties.
https://doi.org/10.1109/TAC.2024.3370494