151398 research outputs found
Sort by
Quantum-based gated recurrent units for multiclass classification to monitor daily living activities for early disease detection
No full textThe continuous monitoring of activities of daily living (ADLs) can play a vital role in assessing an individuals capability to live independently and enable the possibility for early disease detection. This paper introduces a novel hybrid model, called quantum-based gated recurrent unit - multiclass classifier (QGRU-MC), to enhance ADL classification from wearable sensor data. Using statistical feature extraction from the raw accelerometer sensor signals, the QGRU-MC model demonstrates good performance in activity recognition. Preliminary findings suggest that our model has good potential in healthcare applications, and in particular, can contribute to the advancement of future intelligent systems centered on daily activity monitoring and the promotion of healthy aging
Impacts and mechanisms of biological invasions in global protected areas
No full textProtected areas underpin global biodiversity conservation and sustainability goals, but remain at risk from biological invasion impacts. Anticipated expansions of protected lands and waters towards 2030 under the Kunming-Montreal Global Biodiversity Framework (GBF) will concurrently increase the burden of biological invasions within and from these newly protected areas. To prioritise management efforts and assess the potential risks of current and future designations to address multiple GBF targets, it is essential to quantify impact patterns at large scales across contexts. Here, we provide the first comprehensive global assessment of non-native species impacts in protected areas using the Environmental Impact Classification for Alien Taxa (EICAT)—a semi-quantitative, standardised protocol for evaluating their environmental impacts. We analyse taxonomic, ecological, and spatiotemporal predictors of harm and compare the distribution of impact mechanisms across contexts, using 579 impact cases from a literature review. Harmful impacts from non-native species have been documented across all taxonomic groups and environments, including irreversible extinctions. In particular, non-native plants and fish most often affect native plants and insects. Competition, predation, and structural changes were the most impact mechanisms recorded. Studies have been biassed geographically towards Nearctic, Palearctic and Neotropical protected areas, reinforcing the need for regional strategies to mitigate impacts. Our findings culminate in a series of recommendations to directly and indirectly address GBF targets by managing biological invasion impacts. Overall, expanding the range of studies towards whole ecosystem levels across regions can enrich the understanding of how biological invasions affect protected areas. Based on our results, managers should seek to reduce invasion rates in general across all taxa, but especially prioritise those conferring impacts through consumptive and structural mechanisms.<br/
Some results and questions on r -amenability
No full textThe notion of an r-amenable Banach lattice algebra was introduced by M. Roelands as an order counterpart to the well-known notion of amenability for Banach algebras. In this note, we consider the implications of r-amenability for various classes of Banach lattice algebras, and in particular, the question of whether r-amenability implies amenability. We also pose some questions arising from this research.<br/
Leveraging digital tools to foster resilience in the charity sector: the case of eBay’s charity connect initiative
No full textDigital tools can help mitigate serious threats to social and economic activity created by external shocks and reduced government funding. This paper complements Information Technology for Development (IT4D) studies of resilience by focusing on how the introduction of new digital tools can sustain both economic and social activities, potentially leading to further adaptations. It investigates eBay's Charity Connect initiative, which supported the UK charity sector to continue selling donated items during the COVID-19 lockdown. The findings present three pathways-outcomes used to foster resilience through introducing new digital tools: interim resilience, bounded resilience and enhanced resilience. Using conservation of resources theory, the paper explains the difference between these outcomes and discusses how balancing stress and resource investment can help leverage digital tools and foster greater digital resilience.<br/
Efficient integer-only-inference of gradient boosting decision trees on low-power devices
No full textThere is increasingly interest in developing embedded machine learning hardware as it can offer better performance in terms of privacy, bandwidth efficiency, and scalability. Gradient-boosted decision trees (GBDT) represent a strong candidate as they employ less complex logic, but their efficient implementation in field programmable gate array (FPGA) needsto be explored in detail. In this paper, we propose sophisticated quantisation approaches to balance the dual goals of efficiency and performance. In particular, we introduce quantisation-aware training of GBDT for integer-only and binary arithmetic. Results are presented for implementations on a Zynq UltraScale+ MPSoC FPGA with the best design using only 170 Look-up Tables and 233 flip-flops at a clock speed of 724 MHz. Implementations focused on network intrusion detection and jet substructure classification for large-scale physics experiments are explored. An order of magnitude less FPGA resources are used whilst offering extremely high throughput rate and maintaining accuracy. Code is available at https://github.com/malsharari/QATGBDT
Psychological interventions for depression and anxiety in patients with coronary heart disease, heart failure or atrial fibrillation: Cochrane systematic review and meta-analysis
No full textAimsDepression and anxiety occur frequently in individuals with cardiovascular disease and are associated with poor prognosis. This Cochrane systematic review and meta-analysis assessed the effectiveness of psychological interventions on psychological and clinical outcomes in adults with coronary heart disease (CHD), heart failure (HF) or atrial fibrillation (AF).Methods and resultsCENTRAL, MEDLINE, Embase, PsycINFO, and CINAHL databases were searched from January 2009 to July 2022 for randomized controlled trials of psychological interventions vs. controls in adults with CHD, HF, or AF. Twenty-one studies (n = 2591) were assessed using random-effects models. We found psychological interventions reduced depression [standardized mean difference (SMD) −0.36; 95% confidence interval (CI) −0.65 to −0.06; P = 0.02], anxiety (SMD −0.57; 95% CI −0.96 to −0.18; P = 0.004), and improved mental health-related quality of life (HRQoL) (SMD 0.63, 95% CI 0.01 to 1.26; P = 0.05) (follow-up 6–12 months), but not physical health-related quality of life, all-cause mortality or major adverse cardiovascular events compared with controls. High heterogeneity was present across meta-analyses. Meta-regression analysis showed that psychological interventions designed to target anxiety, were more effective than non-targeted interventions.ConclusionThis review found that psychological interventions improved depression, anxiety and mental HRQoL, with those targeting anxiety to show most benefit. Given the statistical heterogeneity, the precise magnitude of effects remains uncertain. Increasing use of multifactorial psychological interventions shows promise for incorporating patient needs and preferences. Investigation of those at high risk of poor outcomes, comparison of intervention components and those with AF is warranted.<br/
Referral challenges for early-onset colorectal cancer: a qualitative study in UK primary care
No full textIn vivo pharmacokinetic study and PBPK modeling: Comparison between 3D-printed nanocrystals and solid dispersions
No full textThe solubility of drugs remains one of the most challenging aspects of formulation development. Several technologies exist to enhance the properties of poorly soluble drugs, with nanocrystal (NC) and solid dispersion (SD) technologies being among the most important. This work compared NCs and SDs under identical conditions using albendazole as a model drug and 3D printing technology as the delivery method. SDs were initially prepared and characterized, and then compared to the NCs system. Techniques such as TGA, DSC, XRD, FTIR, SEM, and confocal Raman microscopy were employed to assess the solid-state properties and formulation homogeneity. Solubility and dissolution profiles were evaluated under simulated gastric and intestinal conditions. An in vivo pharmacokinetic study was performed in dogs comparing 3D-printed formulations (NC-3D and SD-3D) with a control group treated with the pure drug (ABZ-C). A PBPK model was developed also in dogs to further analyse the results. While no statistically significant differences were observed in the in vitro dissolution profiles in 0.1 N HCl, differences emerged in precipitation time and solubility at intestinal pH (6.8). The pharmacokinetic study revealed improvements in the pharmacokinetic profile of both systems compared to the control, as expected. Between the NCs and the SD, the NC system demonstrated significantly superior pharmacokinetic parameters of interest. The PBPK model helped to explain the differences observed in the in vivo study. The results suggest that nanocrystal technology is more effective at enhancing the in vivo performance of Class II drugs, at least when using albendazole as the model drug.</p
