263134 research outputs found
Sort by
Over hiaten in de beveiliging tegen foutaanvallen en gecombineerde aanvallen en het efficiënte ontwerp van tegenmaatregelen
No full textCryptographic algorithms are designed to be secure against cryptanalytic attacks targeting their mathematical structure, yet their implementations in real-world devices, such as smart cards, remain susceptible to physical attacks. Such attacks may exploit unintentional leakages arising from the physical characteristics of devices, such as fluctuations in power consumption or electromagnetic emissions, in what are known as side-channel attacks. Alternatively, they may also deliberately disrupt computations to produce exploitable erroneous outputs, referred to as fault attacks. Combined attacks, which pair the capabilities of side-channel attacks and fault attacks, represent an even more powerful threat. To analyze and understand these attacks, researchers rely on theoretical adversary models that systematically abstract the capabilities of adversaries and guide the design of countermeasures.
Despite increasing research efforts, existing approaches to model physical attack adversaries in theory, particularly for fault and combined attacks, often fail to reflect the full intrinsic nature of practical attack scenarios. Consequently, many countermeasures that are provably secure under these models either fail under realistic attack conditions, due to overly weak assumptions, or incur excessive implementation costs, due to overly strong ones. Similarly, the lack of standardized evaluation methodologies results in unnoticed vulnerabilities in countermeasures. Addressing these fundamental challenges in the field forms the first research direction of this thesis. We introduce a parameterized adversary model that more accurately reflects practical adversarial capabilities. Using this model, we reveal discrepancies between theoretical abstractions and real-world adversaries. We further demonstrate how insufficient evaluations can lead to overlooked weaknesses by breaking two state-of-the-art countermeasures, RS-Mask and CAPA. These findings underscore the immediate need for more realistic adversary abstractions and systematic evaluation frameworks.
The second research direction focuses on the design of efficient protection mechanisms against fault and combined attacks. Traditional approaches often rely on intermediate error detection or correction mechanisms, which increase hardware cost and complexity. For combined attacks, achieving security is even more challenging: existing schemes typically assume that fine-grained fault control is unavoidable, making intermediate error correction mechanisms the common state-of-the-art solution. This thesis challenges this assumption. We first introduce two novel fault countermeasure frameworks that avoid intermediate checks or corrections by enforcing fault propagation to the output. The first, based on the notion of \emph{stability}, ensures that any injected fault reaches the output, enabling detection through a single final error check. The second, MAC-stability, extends this idea by integrating message authentication code (MAC) tags as redundancy, improving security and efficiency trade-off over prior MAC-based approaches. Following the same principles, we further introduce combined stability, a novel security notion for combined attacks. Remarkably, combined stability achieves security against combined attacks using only a single error detection mechanism at the end of the computation, challenging the long-standing belief that costly intermediate checks and corrections are necessary.
Taken together, these contributions advance the state of the art in three ways. First, they highlight fundamental limitations in existing adversary models and evaluation frameworks. Second, they introduce novel countermeasure frameworks for efficiently protecting against fault attacks, relying on a single error detection at the end of the computation. Third, they establish a new security notion for combined attacks that, for the first time, extends this single error detection approach to the more challenging combined attacks context.
In conclusion, this thesis bridges the gap between theoretical abstractions and practical adversarial capabilities. It not only deepens our understanding of fault and combined attacks, but also contributes to the development of protection mechanisms, introducing a new direction focusing on efficiency through a single error check, particularly for combined attacks.status: Accepte
De rol van ATP13A2 in mitochondriale-lysosomale interactie
No full textATP13A2 is a late endolysosomal P5B-type transport ATPase that exports polyamines from the late endolysosome to the cytosol. Loss-of-function mutations in this transporter lead to a range of neurodegenerative disorders, which are all characterized by increased oxidative stress and a flawed mitochondrial-lysosomal axis at the cellular level. In addition, overexpression of ATP13A2 in cell models has been reported to provide protection against rotenone, an environmental Parkinson's disease risk factor and mitochondrial toxin. By exporting polyamines from the lysosome, ATP13A2 prevents lysosomal swelling, rupture and subsequent cathepsin B-dependent cell death, thereby contributing to its neuroprotective effect. However, it remained unclear how ATP13A2 provides mitochondrial protection and whether this involves lysosomal polyamine transport.
In this PhD project, we found that polyamines transported by ATP13A2 complement polyamine synthesis in the mitigation of mitochondrial-generated reactive oxygen species (mitoROS). This is a conserved pathway, as key findings were recapitulated in patient fibroblasts with ATP13A2 mutations and in vivo in a C. elegans model. Cells deficient for ATP13A2 or overexpressing a catalytically dead ATP13A2 mutant were sensitized to rotenone, represented by increased mitochondrial superoxide generation, the induction of a mitoROS-dependent stress response, and cell death. We showed that this mitochondrial protective effect of ATP13A2 is at least in part independent of the lysosomal phenotype, since endocytosis of acidic nanoparticles - which rescues the lysosomal pH and functionality - was unable to decrease mitoROS. We also revealed the cellular route of polyamines taken up via ATP13A2, displaying their redistribution to mitochondria in an ATP13A2-dependent manner, indicating that these polyamines may exert their ROS-scavenging effect locally inside mitochondria.
Moreover, in this PhD project we investigated the impact of disease-associated variants on ATP13A2 activity and on other Parkinson's disease linked proteins with a role in mitochondrial-lysosomal communication. We documented that several ATP13A2 variants affected the ATPase activity, polyamine uptake potential and subcellular location demonstrating their pathogenic effect.
Finally, we generated novel tools to further dissect the role of ATP13A2 in mitochondrial-lysosomal interplay. Organelle immunoprecipitation techniques allowed us to isolate highly pure and intact mitochondria and lysosomes, which uncovered subcellular lipid changes in ATP13A2 deficient cells. We also developed and validated photocrosslinkable polyamine probes to determine other polyamine handling proteins and transporter(s) under control of ATP13A2 in the future.
In conclusion, we uncovered a highly conserved antioxidative pathway mediated by ATP13A2's polyamine transport function that protects against mitochondrial oxidative stress. In addition, we described the functional impact of ATP13A2 disease variants, developed organelle- and substrate-specific tools to identify new players under control of ATP13A2, and we revealed alterations in the subcellular lipidome of ATP13A2 deficient cells.status: Publishe
Onderzoek van de Overeenkomst tussen Convolutionele Neurale Netwerken en de Hersenen bij Sociale Cognitie
No full textArtificial Neural Networks (ANNs) attempt to reproduce 'in silico' processes that take place in the human brain. In the domain of AI, ANNs are now the state-of-the-art method for tackling a variety of problems (computer vision, speech recognition, natural language processing, etc.). Here, the similarity to the human brain is mostly incidental: the brain may serve as a source of inspiration, but it is really the performance of the ANN that matters, e.g., how well it can classify images or understand speech. On the other hand, ANNs are also a popular tool in neuroscientific studies. Here, in stark contrast to the AI domain, similarity to the human brain is key. The ANN is of interest only insofar as it can help to better understand the human brain.
This PhD is situated within the interdisciplinary project 'Computational Modeling of Social Cognition and associated Deficits by means of Artificial Neural Networks' (KU Leuven IDN/21/010) that brings together people from Computer Science, Psychology and Neuropsychiatry to address one of the biggest challenges in both AI and neuroscience: understanding social cognition. The concept of social cognition comprises multifaceted interpersonal processes that underly the interpretation of other people's behaviors and how to interact with them. In particular, this thesis asks the question: do Convolutional Neural Networks (CNNs), a common type of ANN that has achieved remarkable successes in many computer vision tasks, form a good basis to model the human brain for the specific task of social cognition?
Indeed, the successes of CNNs have mainly been booked on low-level tasks, such as, e.g., identifying objects in pictures. Social cognition goes beyond such low-level tasks by necessitating the integration of multiple information streams, and potentially calling upon prior, latent knowledge. Nevertheless, earlier work has established certain correspondences between the visual processing stream in the human brain and CNNs, making them an ideal starting candidate to model this task.
The motivation for this work resides in the fact that how social cognition manifests in the brain is not entirely understood yet. In particular, it is not well understood how neurodivergent brain structures—either through lifelong conditions such Autism Spectrum Disorder or neurodegenerative conditions such as Frontotemporal Dementia—affect social skills. If a computational model were to be found that manages to capture the essence of neurotypical social cognitive processing, this model could then potentially be used to investigate how altered brain mechanisms, reflected in altered network structures, affect social cognition.
This type of research, whereby computer models are used to model the brain, is generally referred to as Computational Psychiatry or Computational Neuroscience. The present dissertation opens with a short, non-technical introduction to this field. Following this, a novel dataset for 'Emotion Recognition in the Wild'—i.e., using non-stages scenes—called FindingEmo is described, including an analysis of why previously existing datasets were insufficient for the task at hand. Using this dataset as a starting point, a method, EmoCAM, is introduced that allows to investigate the importance of specific object classes for the detection of specific emotions by a CNN. Essentially, the question is asked: What information does the network use to make its assessment? Next, making use of both FindingEmo and EmoCAM, the question of to what extent CNNs align with the human brain for the task of valence appraisal is explored by performing a correlation analysis with fMRI data obtained from neurotypical individuals. Some attention is also devoted to a limitation of CNNs, and how despite showing similarities with human visual processing they also show weaknesses not exhibited by humans, in the subsequent chapter focused on how color alterations alter CNN predictions. Finally, the FWO ID-N project 'Social cognition in silico and in vivo' is briefly described. This project forms a direct continuation of, and will build directly on results presented in this dissertation.status: Accepte
IP6K3 laat metabole aanpassing van nierkankercellen toe tijdens metastatische uitgroei
No full textIn the last years, immunotherapy has emerged as a promising treatment for cancer patients. Immune system-based cancer therapies offer a rapid and durable activity, mostly because, once the immune system is activated, it can potentiate a self-propagating and adaptable response. Indeed, immunotherapy leads to durable clinical responses, but only in a fraction of patients and certain tumor types.
Changes in cancer cell metabolism can impair the outcome of immunotherapy treatments. Therefore, a deeper understanding of the metabolic reprogrammingthat occurs in the tumor microenvironment can lead to the identification of new biomarkers and targets for innovative cancer therapies.
With the aim to identify metabolic features associated with immunotherapy resistance, the host lab performed transcriptome and metabolome profiling of three different murine cancer models (MC38 colon carcinoma, CT26 colon carcinoma, and Panc02 pancreatic ductal adenocarcinoma) treated with anti-programmed cell death-1 (PD-1) or anti cytotoxic-T-lymphocyte-associated antigen 4 (CTLA-4) monoclonal antibodies. From the transcriptome analysis, the lab identified a pool of deregulated genes in PD-1 resistance. Among these, a gene coding for a protein that belongs to the inositol phosphokinase (IPK) family, was found to be upregulated under PD-1 treatment, suggesting that it may be involved in immunotherapy resistance. We have preliminary evidence that the knock out (KO) of this specific IPK leads to a significant reduction in tumor volume, even more evident when combining the KO with immunotherapy treatments. These results suggest that the KO of this inositol phosphokinase allows immunotherapy response in vivo.
This project will further unravel the role of this IPK in cancer cells during tumor progression and potentially identify a novel approach to enable optimal immunotherapy in refractory tumors.status: Accepte
Optimaliseren van klimaatslim bosbeheer: het in balans brengen van droogteresiliëntie van bomen en microklimaatbuffering in gematigde bossen
No full textClimate change poses major challenges to temperate forests, with rising prevalence and severity of disturbances threatening tree vitality, microclimate integrity, and biodiversity. A proposed strategy to enhance tree resilience to drought is increasing the thinning intensity, but evidence remains inconsistent, particularly in deciduous forests. Additionally, thinning alters forest structures, an important driver of the forest microclimate. These changes can have cascading effects on the rest of the forest ecosystem, such as altering forest-related species communities. Research integrating thinning effects on multiple forest aspects is, however, lacking. Therefore, this dissertation aimed to investigate the potential of thinning to balance tree growth, drought resilience, forest microclimate stability, and macroarthropod biodiversity in European beech (Fagus sylvatica) and pedunculate oak (Quercus robur) dominated forests in central Belgium.
Chapter 2 investigated the effects of thinning on annual tree growth and stability of growth under drought using tree-ring analyses and historical standing stock reconstructions. Results demonstrated that growth stability remained unaffected by stand density for both European beech and pedunculate oak. Stand density reductions might dampen the interannual growth variability related to drought severity, although this effect was inconsistent. Standing stocks of ~500 m³/ha (≈30 m²/ha basal area) produced the highest individual growth rates, outperforming denser and sparser stands. Thinning, thus, enhances tree growth but provides limited relief from drought stress.
Chapter 3 employed daily dendrometer measurements to analyze intra-annual dynamics of tree water deficit and growth across two contrasting years. European beech showed reduced drought stress at intermediate competition levels (~30 to 35 m²/ha basal area), whereas oak remained insensitive, possibly reflecting its lower drought-susceptibility and already open stand context. Growth, however, improved consistently with reduced competition for beech, while oak growth remained stable. These results confirm the limited potential of thinning to reduce drought stress and highlight its role in supporting individual tree productivity. The inconsistent drought stress effects may be because thinning benefits can be offset by higher individual tree water requirements linked to canopy expansion and evaporative demand increases caused by rising local temperatures after thinning.
Chapter 4 examined forest microclimate and macroarthropod communities in relation to stand structure. Canopy cover emerged as the primary driver of the forest's thermal buffering capacity, with closed stands strongly decoupling forest conditions from macroclimate fluctuations. Reductions in canopy cover not only weakened temperature buffering but could amplify heat extremes (canopy cover < 60%), underscoring the importance of dense canopies for climate regulation. Wet soils, associated with low basal areas, cooled the forest, albeit to a lesser extent than canopy cover. Biodiversity responses were complex and species-specific: woodlice thrived under warm and moist microclimates, while the ground beetles we caught in the forest favored structural openness but were negatively affected by higher temperatures overall. These contrasting responses highlight the role of habitat heterogeneity in supporting diverse species communities.
Chapter 5 integrated the dendrometer and microclimate data in a single structural equation model. Both tree growth and soil moisture benefited from basal area reductions, whereas microclimate stability most strongly relied on canopy cover. Tree growth and soil moisture cooled forest-floor temperatures, although this cooling was limited compared to the warming associated with canopy reductions. Management trade-offs, thus, became explicit, with no configuration that could optimize all ecosystem services simultaneously.
Collectively, these results reveal fundamental trade-offs: thinning enhances growth but does not consistently alleviate drought stress, while closed canopies remain crucial for microclimate buffering. Biodiversity, meanwhile, benefits from habitat heterogeneity. These trade-offs necessitate a mosaic-like approach with diversified management strategies across the landscape at various spatial scales to reconcile timber production, biodiversity, and climate regulation, creating multifunctional and resilient forest landscapes.status: Accepte
Samenwerkend Probleemoplossen in de Context van Non-Formeel Volwassenenonderwijs: De Samenhang Tussen Individuele Inputvariabelen, Teamprocessen en Flow
No full textThis joint PhD investigates the development and assessment of collaborative problem-solving (CPS) in non-formal adult education, guided by the Context-Input-Mediator-Output-Input (CIMOI) framework. It explores how individual input factors—such as personality traits and team roles—relate to team processes and emergent states, including team flow. Across six studies, the thesis contributes design-based insights into CPS training, proposes methods for assessing verbal and non-verbal team interaction using multimodal learning analytics, and examines how personal characteristics shape communication patterns and flow experiences in CPS. By bridging various research fields, this work offers theoretical, methodological, and practical implications for lifelong learning and professional development.status: Publishe
Challenges and Solutions on the Road Towards Smart Lighting with Projected LED Arrays
No full textDynamic illumination systems with an adaptive radiation pattern can be realized by combining high-resolution LED arrays with suitable projection optics. If such adaptive lighting fixtures are furthermore combined with one or more camera sensors, smart lighting systems can be developed. Such systems are already in use for automotive applications, i.e. for adaptive exterior car lighting, but have not yet been commercialized for general lighting. In this paper some important challenges and potential solutions are discussed on the road towards smart lighting with projected LED arrays for general lighting applications.status: Publishe
Perceptie van kleur in virtuele omgevingen
No full textUnderstanding and predicting the perception and appearance of light and color in realistic scenes is one of the ultimate goals of color science. Today's fundamental research on color appearance and color perception is typically based on simplified stimuli and viewing conditions: a flat, uniform stimuli on a neutral flat uniform background illuminated by a single (quasi)neutral light source. However, more complex, naturalistic stimuli that are more difficult to systematically generate and control might provide cues to and activate certain processes in the human visual and cognitive system that might remain quasi-silent using artificial simple stimuli. Advances in virtual reality provide the opportunity to use highly realistic stimuli in an unprecedentedly controlled way, making it a promising innovative, efficient and economical research tool with potentially groundbreaking consequences for the study, understanding and prediction of perception of realistic scenes. However, current knowledge and understanding of how (well) color perception in virtual reality corresponds to that in actual reality is limited. This project will therefore experimentally characterize and compare major, not yet fully understood, aspects of color perception (chromatic adaptation, simultaneous contrast and spatial effects) in actual and virtual reality: to 1) deepen our understanding of these fundamental processes and 2) establish mathematical methods to predict perception in real and virtual environments. Data that will be collected are observer responses (written, verbal, task,etc) to visual stimuli and the optical (spectral & colorimetric) measurements of the latter. Additional data needed for further analysis and modelling of individual variability of color perception, such as observer gender, age, colour deficiency, etc will also be collected.status: Accepte
Verbeterde reconstructie van snelle CT-scans op basis van a priori informatie
No full textAn industrial computed tomography (CT) system is applied for defect detection and dimensional analysis for internal features of manufactured parts or assemblies. In-line CT is a non-destructive three-dimensional (3-D) imaging technology integrated directly into production lines for quality control. The process of quality control using CT begins with acquiring two-dimensional (2-D) X-ray projections from various angles of a rotating object, usually spanning 0 to 360 degrees. These projections are recorded by a detector panel, where each image demonstrates the attenuation of X-ray intensity resulting from the X-rays' interaction with the atoms of the object. Applying dedicated algorithms, a 3-D volume is reconstructed from the 2-D projections and subsequently analysed for quality control, e.g., CT metrology. To accelerate the scanning process and increase throughput in production, reducing the number of projections through sparser angular sampling is an efficient approach. However, this results in under-sampled CT data or few-view CT data, which compromises the quality of reconstruction and accuracy of the CT metrology.
This manuscript addresses the challenge by enhancing the reliability and precision of fast scanning strategies through exploiting a priori knowledge and developing dedicated reconstruction algorithms for in-line CT. Three reconstruction algorithms based on total variation (TV) minimization are proposed. To overcome the limitation inherent to the TV-based method, which may remove some small features during reconstruction, a novel technique for estimating the noise level is introduced. The estimated noise level is incorporated into the TV-type method using two distinct strategies. In the first, the noise level estimation supports the automation of TV-type reconstruction by selecting the optimal regularization parameter according to the noise level, leading to the unconstrained split Bregman algorithm (USB-TV). In the second strategy, the noise level serves as a constraint within the Bregman iterative procedure, enabling the "error forgetting" feature. This yields the constrained split Bregman algorithm (CSB-TV), enhancing image contrast and metrology.
The third developed algorithm utilizes infimal convolution regularization to incorporate a priori knowledge. In this model, the reconstruction is segmented into two main components: structure and pores, penalized by TV-type and ℓ1-norm minimization, respectively. The final volume is obtained by summing these two components. The effectiveness of the proposed methods is evaluated against full-view reconstructions obtained using standard techniques such as the Feldkamp-Davis-Kress (FDK) algorithm. Throughout all experiments, using only between 2%–12% of data required by FDK, the proposed algorithms yielded a higher signal-to-noise ratio with metrology results comparable to those of full-view FDK.status: Accepte
