263134 research outputs found
Sort by
Het gebruik van rectale organoïden voor diagnostiek en gepersonaliseerde geneeskunde bij mucoviscidose
No full textIn recent decades, cystic fibrosis (CF) has evolved from a devastating disease with a dire prognosis to a chronic but manageable disease in most cases. However, challenges still remain in the treatment and diagnosis of CF.
In most cases, a diagnosis of CF is straightforward: the combination of a clinical presentation with either a high sweat chloride concentration (SCC) or identification of 2 CF-causing CFTR variants is sufficient. When the SCC is intermediate and the CFTR genotype undefined, the diagnosis remains inconclusive. Nasal potential difference (NPD) and intestinal current measurement (ICM) are advised for these cases. However, these physiological CFTR assays are not readily available and not feasible in most healthcare settings. Further complicating CF diagnostics are the newborns labelled 'cystic fibrosis screen positive, inconclusive diagnosis' (CFSPID) and the wider range of CFTR-related disorders (CFTR-RD), as well as the large number of CFTR variants of unknown significance.
With the advent of CFTR modulators, efficacy of therapy in CF has increased dramatically. Their mode of action is specific to the molecular defect in the CFTR protein, and thus the genotype. Recording drug response in the patient's own tissue could predict treatment efficacy and thus aid in tailoring treatment.
Recent years have seen an increasing use of rectal organoids in CF research. These are grown from stem cells obtained through a rectal biopsy and consist of an epithelial monolayer surrounding a central lumen. Chloride transport and consequent water movement into the lumen is mediated by CFTR. Organoid swelling is thus a measure of CFTR function, either at baseline or when modulated with compounds. Cultures can be expanded indefinitely for use in experiments or biobanking. In this project, the potential of organoids to fill in the gaps in CF research was explored.
Rectal organoid morphology analysis (ROMA) was developed as a new physiological CFTR assay based on quantification of morphological differences between organoids from subjects with and without CF, showing complete discrimination between these two groups. For those with an inconclusive diagnosis, ROMA provided additional information with diagnostic classification in 75% of cases. ROMA can be performed at all ages, the image analysis has been standardised and is largely automated, and analysis can be centralised.
Rectal organoids can also be used to predict response to modulator therapies and to help in characterisation of CFTR variants, as they have the benefit of stably retaining all of the individual's genetic and epigenetic background. A case series is presented including a person with CF (PwCF) with an unusually mild phenotype for the CFTR genotype. By measuring ROMA indices and organoid swelling in response to treatment, the complex CFTR allele carried by this PwCF was further characterised, in turn informing treatment allowing for a personalised medicine approach.
Organoids have thus proven their potential in diagnostics, personalised medicine, and CFTR variant characterisation. Future work on ROMA will focus on inclusion of more equivocal cases, including CFTR-RD and CFSPID. ROMA could also be used to more accurately predict responses to treatments, especially in PwCF carrying CFTR variants with high residual function.status: Publishe
Assembly and expression in extended reality: transposing human rights across realities
No full textstatus: Published onlin
Common Ownership and Market Entry: Evidence from the Pharmaceutical Industry
No full textCommon ownership - where several firms are (partially) owned by the same investors - and its impact on product market competition has recently drawn much
attention. This paper focuses on its implications for market entry. We consider the entry decisions of generic pharmaceutical firms into drug markets that are opened up by the end of regulatory protection and which were previously dominated by a single firm selling the brand name drug. We find robust evidence that an increase in common ownership leads to a significant reduction in generic entry.sponsorship: Newham acknowledges financial support through project 1103419N from the Flemish Science Foundation (FWO). Seldeslachts acknowledges financial support through projects G.0573.15 and G0D1522N from the FWO. Banal-Estanol acknowledges financial support from the Spanish Ministry of Economy and Competitiveness, through the Severo Ochoa Programme for Centres of Excellence in R&D (SEV-2015-0563) and project ECO2016-76998-P, and from the Fundacion Ramon Areces (CISP15S3712). (Flemish Science Foundation (FWO)|1103419N, Flemish Science Foundation (FWO)|G.0573.15, Flemish Science Foundation (FWO)|G0D1522N, Spanish Ministry of Economy and Competitiveness through the Severo Ochoa Programme for Centres of Excellence|SEV-2015-0563, Fundacion Ramon Areces|CISP15S3712)status: Published onlin
Block-gequantiseerd en data-efficiënt deep learning aan de rand: Optimalisatie van implementaties met universele en domeinspecifieke vectorinstructiesets
No full textThe demand for Artificial Intelligence (AI) continues to grow, leading to it becoming an integral aspect in our everyday lives. This surge can largely be attributed to the remarkable progress in Deep Learning (DL), which has consistently pushed performance boundaries often even outperforming human capabilities. Recent trends have focused on performance improvements through two major scaling axes: training data and model sizes. However, such scaling is often infeasible for practical DL on edge devices targeting custom and personalized tasks. This has driven research into data-efficient algorithms, which aim to deliver high task performance with limited data, addressing scenarios where traditional DL algorithms may falter.
While data-efficient algorithms tackle the challenge of data availability, they often overlook the stringent resource constraints inherent to edge devices. Additionally, model adaptation at the edge is often required to effectively handle data drift, distribution mismatches, etc., further complicating deployments. Block-quantization is used in this thesis to enable efficient computation at the edge, while maintaining algorithmic performance even at reduced-precision. The algorithmic landscape for these workloads is evolving rapidly, motivating the use of instruction-programmable processors that are resilient to obsolescence and also avoid high customization design costs. Moreover, improved performance and efficiency of deployments can be achieved through the use of vector instructions that are suitable for diverse data-parallel workloads predominantly found in DL.
This thesis firstly explores the data-efficient DL modality of few-shot image classification task, which overcomes the challenge of learning a new unseen task with very limited training data, as low as 1 or 5 samples per class. To enable efficient DNN adaptation on resource-constrained edge devices, block-quantization is used during the training process in both the forward and backward passes. It is also shown that conventional meta-learning algorithm initializations used for such tasks fail to transfer well for these quantized adaptations and can be improved by a HW-aware meta-learning algorithm that upfront includes the target HW constraints for DNN adaptations.
In the next part, the focus is still on data-efficiency but targeting a different modality of continuous control tasks such as robotic control. The objective here is to attain effective task control while not requiring large number of environmental interactions. The usage of block-quantized DNNs as dynamics models is introduced here, leading to more efficient inference deployments, and in turn more HW-efficient planning in a conventional model-based RL algorithm. A systematic study of the impacts of quantization error on the degradation of control performance is performed, which typically exacerbates with any model prediction inconsistencies due to the effect of compounding error in its future predictions. Optimizing such a process while still maintaining effective control is achieved through flexible block configurations and utilizing higher precision only for more sensitive computations.
Next to these algorithmic impacts of block-quantization, the exploration of the effect of running them on instruction-programmable processors is also performed. Block-scaled GeMM kernels are designed and optimized across multiple design parameters, with only the use of general-purpose vector instructions. System-level analysis is done to benchmark these kernels against standard datatypes. The impact of increasing block sizes for the same workload leads to decreasing cycle counts due to lower exponent handling overhead. Similarly, 2D blocks are introduced as opposed to conventional 1D blocks, also providing increased performances. To reduce the cost of memory transfers, effective fused requantization has also been designed. These kernels were also evaluated for usage in a DNN training workload. A co-optimized process of adaptive training with vector and square block shapes is proposed to trade off GeMM computational speeds and accuracy.
Lastly, the deployments using domain-specific vector instructions were explored for block-quantized DNN computations with a software-defined SIMD HW unit. Such a HW unit enabled high flexibility for runtime variable-precision computations exhibiting different data-level parallelism, that can be effectively utilized for DNNs. Several mixed-precision inference deployments of CNNs were analyzed, providing several HW-SW trade-off optimization potential.
The contributions of this thesis help push the optimization boundaries of deployments of data-efficient DL algorithms at the edge, using general to domain-specific vector instruction sets, with the use of block-quantization techniques. The progress of research in this interdisciplinary intersection is crucial to enable futuristic applications of bringing intelligence to everyday devices in a performant and privacy-driven manner.status: Publishe
Ontwerp van variatiebestendige digitale circuits met extreem lage voedingsspanning
No full textIn the proposed research we will further investigate ultra low power digital circuit techniques. The focus will be on in-situ timing detection with automatic error correction and/or prevention for further power reduction in digital circuits. It also entails the verification of the results with silicon prototypes in advanced CMOS technologies. The techniques that will be studied and further developed should be compatible with the standard digital design flows.status: Publishe
Malmquist Productivity Index from a regulator's perspective
No full textsponsorship: The authors would like to thank the Department of Economics and Management in Pisa, as this paper started when Maria Molinos-Senante came to Pisa as a visiting researcher. Laura Carosi and Giovanna D'Inverno acknowledge financial support from the University of Pisa [grant number PRA_2022-2023_21] under the 'PRA - Progetti di Ricerca di Ateneo' Institutional Research Grants for 'Fostering the role of public services evaluation for innovation, inclusion, and sustainability'. (Universit di Pisa, Department of Economics and Management in Pisa|PRA_2022-2023_21, University of Pisa)status: Publishe
The Use of Research in Humanistic and Existential Therapy
No full textThis chapter focuses on the link between research and the practice of humanistic–existential psychotherapy, and what the research tells us about evidence-based practice. It considers how humanistic–existential psychotherapy practices can be enhanced and enriched through the application of research and how our clients benefit from the use of research in the humanistic–existential psychotherapy room. The chapter starts with the history of integrating research into humanistic–existential psychotherapy, from Rogers, Maslow, and Yalom, all the way to phenomenology and microphenomenology. It presents overview of the contemporary state of research in humanistic–existential psychotherapy—including general outcome research, comparative outcome research, nonclinical research, and the research question of why humanistic–existential psychotherapy works. The chapter turns to the question of how to use research in our psychotherapy practice, looking at therapy as a form of research, outcome monitoring, and process monitoring.status: Accepte
Miniaturisatie van draagbare elektronica in de context van draadloze body area netwerken
No full textWearable medical sensors are transforming healthcare, but their size and comfort often limit usability. This research explores the miniaturization of wireless sensors designed to capture biosignals (such as muscle activity and brain waves) while ensuring optimal comfort and performance. By reducing the size of these devices without compromising signal quality, this work enhances their practicality for continuous health monitoring. These advancements open new possibilities for more effective biomedical diagnostics, improved patient care, and enhanced human-machine interactions.status: Publishe
Automated Insulin Delivery in Pregnancies Complicated by Type 1 Diabetes
No full textAutomated insulin delivery (AID) systems adapt insulin delivery via a predictive algorithm integrated with continuous glucose monitoring and an insulin pump. Automated insulin delivery has become standard of care for glycemic management of people with type 1 diabetes (T1D) outside pregnancy, leading to improvements in time in range, with lower risk for hypoglycemia and improved treatment satisfaction. The use of AID facilitates optimal preconception care, thus more women of reproductive age are becoming pregnant while using AID. The effectiveness and safety in pregnant populations of using AID systems with algorithms for non-pregnant populations may be impacted by requirements for lower glucose targets and existence of increased insulin resistance during gestation. The CamAPS FX is the only AID system approved for use in pregnancy. A large randomized controlled trial (RCT) with this AID system demonstrated a 10.5% increase in time in pregnancy range (an additional 2.5 hours/day) compared with standard insulin therapy in pregnant women with T1D with a baseline glycated hemoglobin A1c (HbA1c) ≥48 mmol/mol (6.5%). A RCT of AID not approved for use in pregnancy (MiniMed 780G) has also demonstrated some benefits of AID compared with standard insulin therapy with improved time in pregnancy range overnight (24 minutes), less hypoglycemia, and improved treatment satisfaction. There is also increasing evidence that AID can be safely continued during delivery and postpartum, while maintaining glycemic goals with lower risk for hypoglycemia. More AID systems are needed with flexible glucose targets in the pregnancy range and possibly with algorithms that better adapt to changing insulin requirements. More evidence is needed on the impact of AID on maternal and neonatal outcomes. We review the current evidence on the use of AID in pregnancy and postpartum.sponsorship: KB received a fundamental clinical investigatorship from FWO Flanders (1800220N). (FWO Flanders|1800220N)status: Published onlin
Tracing 2500 years of human betaherpesvirus 6A and6B diversity through ancient DNA
No full textHuman betaherpesviruses 6A and 6B (HHV-6A/6B) are DNA viruses, which integrate into the human genome, andare best known to cause “sixth disease.” Despite their recent discovery (1980s), they were speculated to have amuch longer history within the human population than modern data suggest. We present the first 11 ancient ge-nomes of HHV-6A and HHV-6B, dating as far back as the 8th to 6th century BCE. We demonstrate that large frac-tions of current HHV-6 diversity were already established by the 14th century CE. Our data corroborate thatHHV-6A/6B integrations stem from ancient founder events. In addition, we show that all known inherited chro-mosomally integrated HHV-6A clades were already represented in historical populations, confirming that HHV-6Ano longer integrates into the germ line within populations of European ancestry and likely endogenized in earlyhuman history.sponsorship: This work was funded and supported by the following: the Estonian Research Foundation Grant PRG1027 (to K.T., M.G.,T.K., S.S., M. Ma., and L.S.),the Austrian Science Fund FWF 10.55776/ESP162 (https://www.fwf.ac.at/en/research-radar/10.55776/ESP162; to M.G.), the Estonian Ministry of Education and Research TK215 (K.T.,T.K., L.S., S.S., B.B., A.K., M. Ma., and H.K.), the FWO Grant G0A4521N (T.K., O.B., M.H.D.L., C.L.S., and K.T.), the agentschap Onroerend Erfgoed "Besieckt ende Besmet"3M220588 (M.H.D.L.,T.K., and O.B.), Wellcome Trust Award no. 2000368/Z/15/Z (C.L.S., S.I., J.E.R., T.K., and C.C.), St. John's College University of Cambridge (to C.L.S.), the Department of Genetics University of Cambridge (to C.J.H.), UKRI award MR/X034828/1 (to L.v.D),the Estonian Research Foundation Grant PRG1899 (to M. Me.), and the Foundation Osiliana (to M. Mae.). (Estonian Ministry of Education and Research|PRG1027, agentschap Onroerend Erfgoed|TK215, Estonian Research Foundation|TK215, Estonian Research Foundation|3M220588, St John's College University of Cambridge|3M220588, Austrian Science Fund|G0A4521N, Department of Genetics University of Cambridge|10.55776/ESP162, Fonds Wetenschappelijk Onderzoek|10.55776/ESP162, Wellcome Trust|2000368/Z/15/Z, UK Research and Innovation|2000368/Z/15/Z, UK Research and Innovation|MR/X034828/1, Estonian Research Foundation Grant|MR/X034828/1, The Foundation Osiliana, Austrian Science Fund FWF, Estonian Ministry of Education and Research, FWO Grant, Agentschap Onroerend Erfgoed "Besieckt ende Besmet", Wellcome Trust, St. John's College University of Cambridge, Department of Genetics University of Cambridge, UKRI award, Estonian Research Foundation Grant, Foundation Osiliana, PRG1899)status: Publishe