Updating and recalibrating causal probabilistic models on a new target population.

OBJECTIVE: Very often the performance of a Bayesian Network (BN) is affected when applied to a new target population. This is mainly because of differences in population characteristics. External validation of the model performance on different populations is a standard approach to test model's generalisability. However, a good predictive performance is not enough to show that the model represents the unique population characteristics and can be adopted in the new environment. METHODS: In this paper, we present a methodology for updating and recalibrating developed BN models - both their structure and parameters - to better account for the characteristics of the target population. Attention has been given on incorporating expert knowledge and recalibrating latent variables, which are usually omitted from data-driven models. RESULTS: The method is successfully applied to a clinical case study about the prediction of trauma-induced coagulopathy, where a BN has already been developed for civilian trauma patients and now it is recalibrated on combat casualties. CONCLUSION: The methodology proposed in this study is important for developing credible models that can demonstrate a good predictive performance when applied to a target population. Another advantage of the proposed methodology is that it is not limited to data-driven techniques and shows how expert knowledge can also be used when updating and recalibrating the model.

Unsupervised Mixture Models on the Edge for Smart Energy Consumption Segmentation with Feature Saliency.

Smart meter datasets have recently transitioned from monthly intervals to one-second granularity, yielding invaluable insights for diverse metering functions. Clustering analysis, a fundamental data mining technique, is extensively applied to discern unique energy consumption patterns. However, the advent of high-resolution smart meter data brings forth formidable challenges, including non-Gaussian data distributions, unknown cluster counts, and varying feature importance within high-dimensional spaces. This article introduces an innovative learning framework integrating the expectation-maximization algorithm with the minimum message length criterion. This unified approach enables concurrent feature and model selection, finely tuned for the proposed bounded asymmetric generalized Gaussian mixture model with feature saliency. Our experiments aim to replicate an efficient smart meter data analysis scenario by incorporating three distinct feature extraction methods. We rigorously validate the clustering efficacy of our proposed algorithm against several state-of-the-art approaches, employing diverse performance metrics across synthetic and real smart meter datasets. The clusters that we identify effectively highlight variations in residential energy consumption, furnishing utility companies with actionable insights for targeted demand reduction efforts. Moreover, we demonstrate our method's robustness and real-world applicability by harnessing Concordia's High-Performance Computing infrastructure. This facilitates efficient energy pattern characterization, particularly within smart meter environments involving edge cloud computing. Finally, we emphasize that our proposed mixture model outperforms three other models in this paper's comparative study. We achieve superior performance compared to the non-bounded variant of the proposed mixture model by an average percentage improvement of 7.828%.

Naturalistic embodied interactions elicit intuitive physical behaviour in accordance with Newtonian physics.

The success of visuomotor interactions in everyday activities such as grasping or sliding a cup is inescapably governed by the laws of physics. Research on intuitive physics has predominantly investigated reasoning about objects' behaviour involving binary forced choice responses. We investigated how the type of visuomotor response influences participants' beliefs about physical quantities and their lawful relationship implicit in their active behaviour. Participants propelled pucks towards targets positioned at different distances. Analysis with a probabilistic model of interactions showed that subjects adopted the non-linear control prescribed by Newtonian physics when sliding real pucks in a virtual environment even in the absence of visual feedback. However, they used a linear heuristic when viewing the scene on a monitor and interactions were implemented through key presses. These results support the notion of probabilistic internal physics models but additionally suggest that humans can take advantage of embodied, sensorimotor, multimodal representations in physical scenarios.

Expectations about future learning influence moment-to-moment feelings of suspense.

Suspense is a cognitive and affective state that is often experienced in the anticipation of information and contributes to the enjoyment and consumption of entertainment such as movies or sports. Ely et al. proposed a formal definition of suspense which relies upon predictions about future belief updates. In order to empirically evaluate this theory, we designed a task based on the casino card game Blackjack where a variety of suspense dynamics can be experimentally induced. Our behavioural data confirmed the explanatory power of this theory. We further compared this formulation with other heuristic models inspired by studies in other domains such as narratives and found that most heuristic models cannot well account for the specific temporal dynamics of suspense across wide range of game variants. We additionally propose a way to test whether experiencing greater levels of suspense motivates more game-playing. In summary, this work is an initial attempt to link formal models of information and uncertainty with affective cognitive states and motivation.

Stochastic reliability-based risk evaluation and mapping for watershed systems and sustainability (STREAMS).

Mitigating water contamination, improving water security, and increasing sustainability involve environmental awareness and conscientious decision-making by denizens and stakeholders. Achieving such awareness requires visually compelling geospatial decision-making tools that take into account the probabilistic and spatially distributed nature of water contamination. Inspired by the success of weather maps, this paper presents a novel STochastic Reliability-based Risk Evaluation And Mapping for watershed Systems and Sustainability (STREAMS) tool that produces and effectively communicates the risk of water contamination as maps. STREAMS is integrated with ArcGIS geoprocessing tools and uses physics-based reliability theory to compute the spatial distribution of risk, which is defined as the probability of exceeding a safety threshold of water contamination within a watershed. A quantitative analysis of the efficacy of mitigation strategies is conducted by estimating risk reduction from best management practices throughout the entire watershed. Two case studies at different spatial scales are presented, demonstrating STREAMS application to watersheds with varied properties.

Enhancer prediction in the human genome by probabilistic modelling of the chromatin feature patterns.

BACKGROUND: The binding sites of transcription factors (TFs) and the localisation of histone modifications in the human genome can be quantified by the chromatin immunoprecipitation assay coupled with next-generation sequencing (ChIP-seq). The resulting chromatin feature data has been successfully adopted for genome-wide enhancer identification by several unsupervised and supervised machine learning methods. However, the current methods predict different numbers and different sets of enhancers for the same cell type and do not utilise the pattern of the ChIP-seq coverage profiles efficiently. RESULTS: In this work, we propose a PRobabilistic Enhancer PRedictIoN Tool (PREPRINT) that assumes characteristic coverage patterns of chromatin features at enhancers and employs a statistical model to account for their variability. PREPRINT defines probabilistic distance measures to quantify the similarity of the genomic query regions and the characteristic coverage patterns. The probabilistic scores of the enhancer and non-enhancer samples are utilised to train a kernel-based classifier. The performance of the method is demonstrated on ENCODE data for two cell lines. The predicted enhancers are computationally validated based on the transcriptional regulatory protein binding sites and compared to the predictions obtained by state-of-the-art methods. CONCLUSION: PREPRINT performs favorably to the state-of-the-art methods, especially when requiring the methods to predict a larger set of enhancers. PREPRINT generalises successfully to data from cell type not utilised for training, and often the PREPRINT performs better than the previous methods. The PREPRINT enhancers are less sensitive to the choice of prediction threshold. PREPRINT identifies biologically validated enhancers not predicted by the competing methods. The enhancers predicted by PREPRINT can aid the genome interpretation in functional genomics and clinical studies.

Probabilistic Modelling for Unsupervised Analysis of Human Behaviour in Smart Cities.

The growth of urban areas in recent years has motivated a large amount of new sensor applications in smart cities. At the centre of many new applications stands the goal of gaining insights into human activity. Scalable monitoring of urban environments can facilitate better informed city planning, efficient security, regular transport and commerce. A large part of monitoring capabilities have already been deployed; however, most rely on expensive motion imagery and privacy invading video cameras. It is possible to use a low-cost sensor alternative, which enables deep understanding of population behaviour such as the Global Positioning System (GPS) data. However, the automated analysis of such low dimensional sensor data, requires new flexible and structured techniques that can describe the generative distribution and time dynamics of the observation data, while accounting for external contextual influences such as time of day or the difference between weekend/weekday trends. In this paper, we propose a novel time series analysis technique that allows for multiple different transition matrices depending on the data's contextual realisations all following shared adaptive observational models that govern the global distribution of the data given a latent sequence. The proposed approach, which we name Adaptive Input Hidden Markov model (AI-HMM) is tested on two datasets from different sensor types: GPS trajectories of taxis and derived vehicle counts in populated areas. We demonstrate that our model can group different categories of behavioural trends and identify time specific anomalies.

Comparing probabilistic and descriptive analyses of time-dose-toxicity relationship for determining no-observed-adverse-effect level in drug development.

The no-observed-adverse-effect level (NOAEL) of a drug defined from animal studies is important for inferring a maximal safe dose in human. However, several issues are associated with its concept, determination and application. It is confined to the actual doses used in the study; becomes lower with increasing sample size or dose levels; and reflects the risk level seen in the experiment rather than what may be relevant for human. We explored a pharmacometric approach in an attempt to address these issues. We first used simulation to examine the behaviour of the NOAEL values as determined by current common practice; and then fitted the probability of toxicity as a function of treatment duration and dose to data collected from all applicable toxicology studies of a test compound. Our investigation was in the context of an irreversible toxicity that is detected at the end of the study. Simulations illustrated NOAEL's dependency on experimental factors such as dose and sample size, as well as the underlying uncertainty. Modelling the probability as a continuous function of treatment duration and dose simultaneously to data from multiple studies allowed the estimation of the dose, along with its confidence interval, for a maximal risk level that might be deemed as acceptable for human. The model-based data integration also reconciled between-study inconsistency and explicitly provided maximised estimation confidence. Such alternative NOAEL determination method should be explored for its more efficient data use, more quantifiable insight to toxic doses, and the potential for more relevant animal-to-human translation.

A new Bayesian approach for managing bathing water quality at river bathing locations vulnerable to short-term pollution.

Short-term fecal pollution events are a major challenge for managing microbial safety at recreational waters. Long turn-over times of current laboratory methods for analyzing fecal indicator bacteria (FIB) delay water quality assessments. Data-driven models have been shown to be valuable approaches to enable fast water quality assessments. However, a major barrier towards the wider use of such models is the prevalent data scarcity at existing bathing waters, which questions the representativeness and thus usefulness of such datasets for model training. The present study explores the ability of five data-driven modelling approaches to predict short-term fecal pollution episodes at recreational bathing locations under data scarce situations and imbalanced datasets. The study explicitly focuses on the potential benefits of adopting an innovative modeling and risk-based assessment approach, based on state/cluster-based Bayesian updating of FIB distributions in relation to different hydrological states. The models are benchmarked against commonly applied supervised learning approaches, particularly linear regression, and random forests, as well as to a zero-model which closely resembles the current way of classifying bathing water quality in the European Union. For model-based clustering we apply a non-parametric Bayesian approach based on a Dirichlet Process Mixture Model. The study tests and demonstrates the proposed approaches at three river bathing locations in Germany, known to be influenced by short-term pollution events. At each river two modelling experiments ("longest dry period", "sequential model training") are performed to explore how the different modelling approaches react and adapt to scarce and uninformative training data, i.e., datasets that do not include event pollution information in terms of elevated FIB concentrations. We demonstrate that it is especially the proposed Bayesian approaches that are able to raise correct warnings in such situations (> 90 % true positive rate). The zero-model and random forest are shown to be unable to predict contamination episodes if pollution episodes are not present in the training data. Our research shows that the investigated Bayesian approaches reduce the risk of missed pollution events, thereby improving bathing water safety management. Additionally, the approaches provide a transparent solution for setting minimum data quality requirements under various conditions. The proposed approaches open the way for developing data-driven models for bathing water quality prediction against the reality that data scarcity is common problem at existing and prospective bathing waters.

Novel risk index integrating practical operation limits enhances probabilistic contingency ranking for large-scale photovoltaic plant planning.

This research addresses the pressing need for heightened grid security amid increasing uncertainties in photovoltaic PV generation. The research problem lies in the limitations of conventional contingency analysis metrics, failing to adequately consider both contingency occurrences and uncertainties inherent in PV generation. In response, a comprehensive algorithm is proposed that introduces a novel severity function framework, enhancing traditional contingency ranking metrics. This approach incorporates grid remedial actions and refines line and bus voltage classification by considering available correction time, aiming to offer a more robust security assessment. Motivated by the imperative to address uncertainty in PV generation, the proposed work builds on established analysis tools. A probabilistic load flow algorithm manages PV generation uncertainties, utilizing historical data for contingency incidence uncertainty. Additionally, a probabilistic model for PV plants integrates historical solar data, deriving hourly probability density functions to meet grid code requirements, including reactive power considerations. The justification for this work lies in the algorithm's demonstrated efficacy, validated on the IEEE 14-bus network. Results highlight its ability to identify risks associated with line overloading and bus voltage breaches. Comparative evaluations underscore proper coupling buses for security, favoring distributed capacity to mitigate line overloading risks. The study's key results emphasize voltage risk amplification with reactive power omission, stressing the significance of compensation strategies. This research addresses a critical problem, presenting a comprehensive algorithmic solution to enhance grid security amidst uncertainties in PV integration. Findings offer valuable insights for strategically interaction between large scale PV plants and electrical grid, contributing to an improved grid security paradigm in a dynamic and uncertain energy model.

The impact of the proposed revised Australia's microbiological monitoring programme for beef and sheep meat exported to the EU.

The European Commission asks scientific and technical assistance from EFSA to determine the impact of the revision of the Australian monitoring programme on its ability to detect microbiological contamination. Considering that, in 2010, the European Commission determined the current Australian monitoring programme to be equivalent to the EU requirements for microbiological monitoring further to an EFSA scientific assessment, the current and proposed programmes were described and the total number of alerts was compared using a probabilistic modelling approach. In the current programme, only beef and sheep carcasses are monitored using three-class moving window sampling plans, while in the proposed programme, carcass, bulk meat, primal and offal are monitored using four two-class sampling plans and Salmonella testing is excluded. The models revealed that the current programme provides a higher number of alerts for APC, while the proposed monitoring programme provides a higher number of alerts for E. coli. For APC and E. coli combined, the mean, 5th and 95th centiles of the uncertainty distribution of the total number of alerts in the current and the proposed monitoring programme are 201 [179, 227] and 172 [149, 194] for beef, and 199 [175, 222] and 2897 [2795, 3008] for sheep, respectively. For Salmonella, there are no alerts for the proposed programme since sampling is excluded while for the current programme, the estimated mean, 5th and 95th centiles of the uncertainty distribution of the number of alerts for a 5-year period were 143 [126, 144] for heifer/steer, 1.6 [0, 4] for cow/bull and 0 [0, 0] for lamb/sheep. Overall, for APC and E. coli, the estimated total number of alerts was similar (beef) or higher (sheep) for the proposed compared to the current programme. In contrast, Salmonella sampling is excluded from the proposed programme and thus cannot detect the number of current alerts.

AI-based derivation of atrial fibrillation phenotypes in the general and critical care populations.

BACKGROUND: Atrial fibrillation (AF) is the most common heart arrhythmia worldwide and is linked to a higher risk of mortality and morbidity. To predict AF and AF-related complications, clinical risk scores are commonly employed, but their predictive accuracy is generally limited, given the inherent complexity and heterogeneity of patients with AF. By classifying different presentations of AF into coherent and manageable clinical phenotypes, the development of tailored prevention and treatment strategies can be facilitated. In this study, we propose an artificial intelligence (AI)-based methodology to derive meaningful clinical phenotypes of AF in the general and critical care populations. METHODS: Our approach employs generative topographic mapping, a probabilistic machine learning method, to identify micro-clusters of patients with similar characteristics. It then identifies macro-cluster regions (clinical phenotypes) in the latent space using Ward's minimum variance method. We applied it to two large cohort databases (UK-Biobank and MIMIC-IV) representing general and critical care populations. FINDINGS: The proposed methodology showed its ability to derive meaningful clinical phenotypes of AF. Because of its probabilistic foundations, it can enhance the robustness of patient stratification. It also produced interpretable visualisation of complex high-dimensional data, enhancing understanding of the derived phenotypes and their key characteristics. Using our methodology, we identified and characterised clinical phenotypes of AF across diverse patient populations. INTERPRETATION: Our methodology is robust to noise, can uncover hidden patterns and subgroups, and can elucidate more specific patient profiles, contributing to more robust patient stratification, which could facilitate the tailoring of prevention and treatment programs specific to each phenotype. It can also be applied to other datasets to derive clinically meaningful phenotypes of other conditions. FUNDING: This study was funded by the DECIPHER project (LJMU QR-PSF) and the EU project TARGET (10113624).

Methods for multi-omic data integration in cancer research.

Multi-omics data integration is a term that refers to the process of combining and analyzing data from different omic experimental sources, such as genomics, transcriptomics, methylation assays, and microRNA sequencing, among others. Such data integration approaches have the potential to provide a more comprehensive functional understanding of biological systems and has numerous applications in areas such as disease diagnosis, prognosis and therapy. However, quantitative integration of multi-omic data is a complex task that requires the use of highly specialized methods and approaches. Here, we discuss a number of data integration methods that have been developed with multi-omics data in view, including statistical methods, machine learning approaches, and network-based approaches. We also discuss the challenges and limitations of such methods and provide examples of their applications in the literature. Overall, this review aims to provide an overview of the current state of the field and highlight potential directions for future research.

Cumulative dietary risk assessment for French consumers exposed to succinate dehydrogenase inhibitor pesticides.

Consumers are exposed to succinate dehydrogenase inhibitor (SDHI) pesticides through their diet. A cumulative dietary risk assessment for the French population has been performed with French monitoring data (2017-2021) and consumption data from INCA3. The calculation followed a two-tiered approach, using deterministic then probabilistic methods. It was carried out, using European health based guidance values (HBGV) derived for each active substance to characterise their toxicity. In Tier I, the calculated hazard index of 0.12 was below the threshold of 1 and in Tier II, the total margin of exposure at percentile 99.9 remains above the trigger value of 100 (1798 [1631-2311]). In Tier II, the three main risk drivers identified at the upper tail of the distribution were strawberries-fluopyram (19.1%), peaches-fluopyram (14.1%) and table grapes-boscalid (10.5%). Finally, the impact of the major sources of uncertainties was qualitatively evaluated. All together, they were considered of low impact on the outcomes. This work demonstrates the absence of unacceptable chronic risk related to the cumulative exposure of SDHI for French consumers during the 2017-2021 period.

Stochastic modelling of pesticide transport to drinking water sources via runoff and resulting human health risk assessment.

A modelling framework was developed to facilitate a probabilistic assessment of health risks posed by pesticide exposure via drinking water due to runoff, with the inclusion of influential site conditions and in-stream processes. A Monte-Carlo based approach was utilised to account for the inherent variability in pesticide and population properties, as well as site and climatic conditions. The framework presented in this study was developed with an ability to integrate different data sources and adapt the model for various scenarios and locations to meet the users' needs. The results from this model can be used by farm advisors and catchment managers to identify lower risk pesticides for use for given soil and site conditions and implement risk mitigation measures to protect water resources. Pesticide concentrations in surface water, and their risk of regulatory threshold exceedances, were simulated for fifteen pesticides in an Irish case study. The predicted concentrations in surface water were then used to quantify the level of health risk posed to Irish adults and children. The analysis indicated that herbicides triclopyr and MCPA occur in the greatest concentrations in surface water, while mecoprop was associated with the highest potential for health risks. The study found that the modelled pesticides posed little risk to human health under current application patterns and climatic conditions in Ireland using international acceptable intake values. A sensitivity study conducted examined the impact seasonal conditions, timing of application, and instream processes, have on the transport of pesticides to drinking water.

Effects of Obesity on Medial Tibiofemoral Cartilage Mechanics in Females-An Exploration Using Musculoskeletal Simulation and Probabilistic Cartilage Failure Modelling.

This study examined the effects of obesity on cartilage mechanics and longitudinal failure probability at the medial tibiofemoral compartment, using combined musculoskeletal simulation and probabilistic failure modelling approaches. The current investigation examined twenty obese females (BMI > 30.0 kg/m2) and 20 healthy weight (BMI < 25.0 kg/m2) females. Walking kinematics were obtained via an 8-camera optoelectric system, and a force plate was used to collect ground reaction forces. Musculoskeletal simulation and probabilistic failure modelling were utilized to explore medial tibiofemoral forces and cartilage probability. Comparisons between groups were undertaken using linear mixed-effects models. Net peak cartilage forces, stress and strain were significantly larger in the obese group (force = 2013.92 N, stress = 3.03 MPa & strain = 0.25), compared to health weight (force = 1493.21 N, stress 2.26 MPa & strain = 0.19). In addition, medial tibiofemoral cartilage failure probability was also significantly larger in the obese group (42.98%) compared to healthy weight (11.63%). The findings from the current investigation show that obesity has a profoundly negative influence on longitudinal medial knee cartilage health and strongly advocates for the implementation of effective weight management programs into long-term musculoskeletal management strategies.

Probabilistic Modelling of Fracture Toughness of Composites with Discontinuous Reinforcement.

The results presented in the paper are related to the prediction of the effective fracture toughness of particulate composites (KICeff). KICeff was determined using a probabilistic model supported by a cumulative probability function qualitatively following the Weibull distribution. Using this approach, it was possible to model two-phase composites with an arbitrarily defined volume fraction of each phase. The predicted value of the effective fracture toughness of the composite was determined based on the mechanical parameter of the reinforcement (fracture toughness), matrix (fracture toughness, Young's modulus, yield stress), and composite (Young's modulus, yield stress). The proposed method was validated: the determined fracture toughness of the selected composites was in accordance with the experimental data (the authors' tests and literature data). In addition, the obtained results were compared with data captured by means of the rule of mixtures (ROM). It was found that the prediction of KICeff using the ROM was subject to a significant error. Moreover, a study of the effect of averaging the elastic-plastic parameters of the composite, on KICeff, was performed. The results showed that if the yield stress of the composite increased, a decrease in its fracture toughness was noticed, which is in line with the literature reports. Furthermore, it was noted that an increase in the Young's modulus of the composite affected KICeff in the same way as a change in its yield stress.

Dietary intake of acrylamide in the Norwegian EuroMix biomonitoring study: Comparing probabilistic dietary estimates with haemoglobin adduct measurements.

Acrylamide is a probable human carcinogen with widespread exposure via food. The present study compared acrylamide intake measurements obtained from haemoglobin adduct levels and self-registered dietary consumption data in a group of 144 Norwegian healthy adults. Acrylamide adducts to N-terminal valine in haemoglobin were measured and used to estimate the intake via the internal dose approach which showed a median (interquartile range) of 0.24 (0.19-0.30) µg/kg bw/day. Data from weighed food records and food frequency questionnaires from the same individuals were used for probabilistic modelling of the intake of acrylamide. The median acrylamide intake was calculated to be 0.26 (0.16-0.39) and 0.30 (0.23-0.39) µg/kg bw/day, respectively from the two sources of self-registered dietary consumption data. Overall, a relatively good agreement was observed between the methods in pairwise comparison in Bland-Altman plots, with the methods disagreeing with 7% or less of the values. The intake estimates obtained with the two dietary consumption methods and one biomarker method are in line with earlier dietary estimates in the Norwegian population. The Margin of Exposure indicate a possible health risk concern from dietary acrylamide. This is the first study with a comparison in the same individuals of acrylamide intake estimates obtained with these methods.

Risk ranking of macrolide antibiotics - Release levels, resistance formation potential and ecological risk.

Antibiotic resistance (AR) development in natural water bodies is a significant source of concern. Macrolide antibiotics in particular have been identified as pollutants of concern for AR development throughout the literature, as well as by state and international authorities. This study utilises a probabilistic model to examine the risk of AR development arising from human-use macrolide residues, utilising administration rates from Ireland as a case study. Stages modelled included level of administration, excretion, degradation in wastewater, removal in wastewater treatment, assuming conventional activated sludge (CAS) treatment, and dilution. Release estimates per day, as well as risk quotient values for antibiotic resistance development and ecological impact, are generated for erythromycin, clarithromycin, and azithromycin. In the modelled scenario in which conventional activated sludge treatment is utilised in wastewater treatment, this model ranks risk of resistance development for each antibiotic in the order clarithromycin > azithromycin > erythromycin, with mean risk quotient values of 0.50, 0.34 and 0.12, respectively. A membrane bioreactor scenario was also modelled, which reduced risk quotient values for all three macrolides by at least 50 %. Risk of ecological impact for each antibiotic was also examined, by comparing environmental concentrations predicted to safety limits based on toxicity data for cyanobacteria and other organisms from the literature, with azithromycin being identified as the macrolide of highest risk. This study compares and quantifies the risk of resistance development and ecological impact for a high-risk antibiotic group in the Irish context, and demonstrates the potential for risk reduction achieved by adoption of alternative (e.g. membrane bioreactor) technology.

Low- and no-calorie sweetener intakes from beverages - an up-to-date assessment in four regions: Brazil, Canada, Mexico and the United States.

The current assessment estimated exposure to four low- and no-calorie sweeteners (LNCS) (aspartame, acesulfame potassium (AceK), steviol glycosides and sucralose) from beverages in Brazil, Canada, Mexico and the United States, using up-to-date nationally representative consumption data and industry reported-use level information. Two modelling scenarios were applied - the probabilistic model was guided by reported use level data, with estimated intake for an individual leveraging market-weighted average use level of a particular LNCS in any given LNCS-sweetened beverage type, while the distributional (brand-loyal) model assumed consumer behaviour-led patterns, namely that an individual will be brand loyal to a pre-determined beverage type. Consumer-only and general population intake estimates were derived for the overall population and individual age categories, and compared to the respective acceptable daily intake (ADI) as established by the Joint FAO/WHO Expert Committee on Food Additives (JECFA) for each LNCS. The mean, 90th percentile and 95th percentile intake estimates were substantially lower than the ADI in both modelling scenarios, regardless of the population group or market. In the probabilistic model, the highest consumer-only intake was observed for AceK in Brazilian adolescents (95th percentile, 12.4% of the ADI), while the highest 95th percentile intakes in the distributional model were observed for sucralose in Canadian adults at 20.9% of the ADI. This study provides the latest insights into current intakes of LNCS from water-based non-alcoholic LNCS-sweetened beverages in these regions, aligning well with those published elsewhere.