Application Of Risk Assessment Research Articles

Health Risk Assessment Using Machine Learning: Systematic Review

According to the World Health Organization, chronic illnesses account for over 70% of deaths globally, underscoring the need for effective health risk assessment (HRA). While machine learning (ML) has shown potential in enhancing HRA, no systematic review has explored its application in general health risk assessments. Existing reviews typically focus on specific conditions. This paper reviews published articles that utilize ML for HRA, and it aims to identify the model development methods. A systematic review following Tranfield et al.’s three-stage approach was conducted, and it adhered to the PRISMA protocol. The literature was sourced from five databases, including PubMed. Of the included articles, 42% (11/26) addressed general health risks. Secondary data sources were most common (14/26, 53.85%), while primary data were used in eleven studies, with nine (81.81%) using data from a specific population. Random forest was the most popular algorithm, which was used in nine studies (34.62%). Notably, twelve studies implemented multiple algorithms, while seven studies incorporated model interpretability techniques. Although these studies have shown promise in addressing digital health inequities, more research is needed to include diverse sample populations, particularly from underserved communities, to enhance the generalizability of existing models. Furthermore, model interpretability should be prioritized to ensure transparent, trustworthy, and broadly applicable healthcare solutions.

Research on Financial Investment Risk Assessment Techniques and Applications

In recent years, the financial markets have undergone rapid development and significant volatility, leading to a pronounced urgency and complexity in investment risk assessment. This study explores the technologies and applications of financial investment risk evaluation, offering an in-depth analysis of the theoretical foundations and practical operations of quantitative, qualitative, and hybrid assessment methodologies. Traditional quantitative techniques, such as the Value at Risk (VaR) model, provide crucial objective data support, while qualitative approaches enrich the breadth of risk identification through expert opinions and seasoned experience. Hybrid methods synergistically combine the strengths of both, adapting to the intricacies of the market environment. The integration of big data and artificial intelligence technologies injects innovative momentum into risk assessment, enabling more precise and timely risk monitoring. Case applications span multi-layered risk management strategies within banks, insurance companies, and hedge funds, offering valuable insights for diverse financial institutions. The study posits that efficient risk assessment relies not only on technological advancements but also necessitates consideration of market dynamics and regulatory environment fluctuations. Continual innovation and practical optimization are pivotal in addressing future challenges.

MiRNA Profiling of Areca Nut-Induced Carcinogenesis in Head and Neck Cancer.

While miRNAs are increasingly recognized for their role in tumorigenesis, their involvement in head and neck cancer (HNC) remains insufficiently explored. Additionally, the carcinogenic mechanisms of areca nut, a major habitual carcinogen in Southeast Asia, are not well understood. This study adopts a systematic approach to identify miRNA profiles associated with areca nut-induced HNC. Using miRNA microarray analysis, we identified 292 miRNAs dysregulated in areca nut-treated HNC cells, with 136 upregulated and 156 downregulated. Bioinformatic analysis of the TCGA-HNSC dataset uncovered a set of 692 miRNAs relevant to HNC development, comprising 449 overexpressed and 243 underexpressed in tumor tissues. Integrating these datasets, we defined a signature of 84 miRNAs, including 39 oncogenic miRNAs (OncomiRs) and 45 tumor-suppressive miRNAs (TsmiRs), highlighting their pivotal role in areca nut-induced carcinogenesis. MultiMiR analysis identified 740 genes cross-regulated by eight hub TsmiRs, significantly impacting key cancer-related pathways (p53, PI3K-AKT, MAPK, and Ras) and critical oncogenic processes. Moreover, we validated miR-499a-5p as a vital regulator, demonstrating its ability to mitigate areca nut-induced cancer progression by reducing cell migration, invasion, and chemoresistance. Thus, this miRNA signature addresses a crucial gap in understanding the molecular underpinnings of areca nut-induced carcinogenesis and offers a promising platform for clinical applications in risk assessment, diagnosis, and prognosis of areca nut-associated malignancies.

Methods for Quantifying Source-Specific Air Pollution Exposure to Serve Epidemiology, Risk Assessment, and Environmental Justice.

Identifying sources of air pollution exposure is crucial for addressing their health impacts and associated inequities. Researchers have developed modeling approaches to resolve source-specific exposure for application in exposure assessments, epidemiology, risk assessments, and environmental justice. We explore six source-specific air pollution exposure assessment approaches: Photochemical Grid Models (PGMs), Data-Driven Statistical Models, Dispersion Models, Reduced Complexity chemical transport Models (RCMs), Receptor Models, and Proximity Exposure Estimation Models. These models have been applied to estimate exposure from sources such as on-road vehicles, power plants, industrial sources, and wildfires. We categorize these models based on their approaches for assessing emissions and atmospheric processes (e.g., statistical or first principles), their exposure units (direct physical measures or indirect measures/scaled indices), and their temporal and spatial scales. While most of the studies we discuss are from the United States, the methodologies and models are applicable to other countries and regions. We recommend identifying the key physical processes that determine exposure from a given source and using a model that sufficiently accounts for these processes. For instance, PGMs use first principles parameterizations of atmospheric processes and provide source impacts exposure variability in concentration units, although approaches within PGMs for source attribution introduce uncertainties relative to the base model and are difficult to evaluate. Evaluation is important but difficult-since source-specific exposure is difficult to observe, the most direct evaluation methods involve comparisons with alternative models.

Application of domino effect quantitative risk assessment to Natech accident triggered by earthquakes in a liquor storage tank area

Application of domino effect quantitative risk assessment to Natech accident triggered by earthquakes in a liquor storage tank area

Continuous Glucose Data Construction and Risk Assessment Application of Diabetic Retinopathy Complications for Patients with Type 2 Diabetes Mellitus

Managing diabetes mellitus (DM) includes achieving acceptable blood glucose levels and minimizing the risk of complications from DM. The appropriate glucose sensing method is continuous glucose monitoring (CGM). Effective evaluation metrics that reflect glucose fluctuations can be realized. However, compared with self-monitoring of blood glucose (SMBG), CGM data are not easy to obtain. Therefore, this article studies a fusion model to achieve this objective, including Gaussian process regression (GPR) and long short-term memory (LSTM). Compared with the three commonly used LSTM, GPR, and support vector machine, the proposed model can construct accurate results. By using the constructed CGM data, the conventional metrics, such as the mean amplitude of glycemic excursion (MAGE), mean blood glucose (MBG), standard deviation (SD), and time in range (TIR), are calculated. These metrics and other variables are input into statistical methods to realize diabetic retinopathy risk assessment. In this way, the relationship between the glycemic variability of the constructed CGM data by the mathematical model and DR could be achieved. The utilized statistical methods include single-factor analysis and binary multivariate logistic regression analysis. Results show that fasting blood glucose, disease course, history of hypertension, MAGE and TIR are independent risk factors for DR.

Research progress on the application of virtual artificial intelligence in risk assessment and diagnosis of periodontal disease

Periodontal disease is a common and frequently-occurring disease in China. Early detection, diagnosis, and treatment of periodontal disease are of great significance for achieving universal oral health and even systemic health. Artificial intelligence endows machines with the ability to mimic human intelligent behavior, and is commonly used in medical field with both physical and virtual forms. Virtual artificial intelligence empowers traditional experience in the application of periodontal disease risk assessment and diagnosis, with the potential to develop a variety of oral health screening tools. It helps to provide new evidence for the prognosis of periodontal disease, improve the accuracy and efficiency of diagnosis, reduce technical sensitivity and further promote the periodontal treatment transformation from "treatment-oriented" to "prevention-oriented". This paper reviews the current applications and progresses of virtual artificial intelligence in periodontal risk assessment and diagnosis, as well as its limitations, providing ideas for future researches on the application of virtual artificial intelligence in this field.

Probabilistic Approach for Assessing the Occupational Risk of Olfactometric Examiners: Methodology Description and Application to Real Exposure Scenario

Human examiners, known as panelists, are exposed to an unknown occupational exposure risk while determining odor concentration (Cod) using dynamic olfactometry. In the literature, a few papers, based on a deterministic approach, have been proposed to establish this occupational risk. As a result, the purpose of this study is to develop and apply a probabilistic approach, based on the randomization of exposure parameters, for assessing and evaluating the occupational exposure risk among olfactometric examiners. In this methodology, the risk is assessed by computing the hazard index (HI) and inhalation risk (IR) to determine the non-carcinogenic and carcinogenic risks. To randomize the exposure parameters, a Monte Carlo simulation was described and then applied in real exposure scenario to establish the exposure risk in terms of probability. Therefore, a one-year survey of the working activity of olfactometric examiners of Laboratorio Olfattometrico of Politecnico di Milano university was conducted. Based on this data collection (exposure parameters and chemical data, divided according to sample categories), a randomized exposure scenario was constructed to estimate the probability and cumulative distribution function of risk parameters. Different distributions were obtained for different industrial samples categories and were compared with respect to acceptability criteria (the value of HI and IR at 95th percentile of distribution). The elaboration provided evidence that negligible non-carcinogenic and carcinogenic risks are associated with the panelists’ activity, according to an entire annual dataset. The application of probabilistic risk assessment provides a more comprehensive and effective characterization of the general exposure scenario for olfactometric examiners, surpassing the limitations of a deterministic approach. This method can be extended to future exposure scenarios and enables the selection of the most effective risk management strategies to protect the health of olfactometric examiners.

Non-Conventional Data for Farming-Related Air Pollution: Contributions to Modelling and Risk Assessment in the Lombardy Region, Italy

Air pollution is one of the most critical global health concerns today. While emissions from industrial activities and combustion processes are the primary threats to air quality, intensive farming activities also contribute significantly, especially through ammonia emissions that promote the formation of secondary pollutants, such as particulate matter. Advancements in air quality research have been achieved by enhancements in emissions characterisation, modelling techniques, and sensor technology, expanding the availability of air pollution data beyond traditional ground sensor observations, which are often lacking in rural agricultural areas. Accordingly, this paper demonstrates the advantages of integrating traditional and non-conventional data to investigate farming-related air pollution through a case study in the Lombardy Region, Northern Italy. The study incorporates an array of data sources, including ground sensors and atmospheric composition model estimates. The concurrent utilisation of these diverse datasets is explored through machine learning modelling, focusing on assessing the influence of agricultural activities on particulate matter distribution patterns. Finally, the reliability of non-conventional air pollution data for health risk assessment applications is also investigated. The paper critically discusses the main findings based on empirical results, highlighting the significance of integrating multiple data sources to complement traditional air quality monitoring while outlining the main limitations in terms of the accuracy and usability of such non-conventional data.

Natural Language Processing Risk Assessment Application Developed for Marble Quarries

In this study, by using the texts describing the hazards and precautions taken during text mining, the necessary processes were carried out to first estimate the probability value and severity value of the risk and then calculate the risk values by Natural Language Processing analysis. In order to be used within the scope of the study, two data sets were generated from the data in the risk assessment report prepared by applying the L-type matrix risk assessment in marble quarries between 2015 and 2021. Stochastic Gradient Descent (SGD) was used for classification and prediction by analyzing text data. One data set was used to analyze the probability value of the risk and the other was used to analyze the severity value of the risk. In light of the results, when a text containing hazard and precaution information was entered, a system was developed that analyzed this text, estimated the probability and severity values, and calculated the risk assessment score. The application of the SGD algorithm to learning models developed on text data yielded an accuracy rate of 91.2% in the risk probability data set and 97.5% in the risk severity data set. The results indicated that the models were capable of conducting automatic risk assessment on text data and of effectively predicting the requisite probability and severity values. Due to the high accuracy rates obtained during the study, this risk assessment software was recommended for use in marble quarries.

Assessing the feasibility of pre-emptive biological control against the emerald ash borer, Agrilus planipennis, an imminent biosecurity threat to Europe

As the globalisation of trade increases, so does the spread of arthropod pests, leading to a growing focus on biosecurity preparedness. One approach to this is pre-emptive biological control, involving the importation of classical biological control (CBC) agents for risk assessments and acquiring approval for their release prior to the expected arrival of their target pests. This aims to mitigate the economic and/or environmental consequences of a delayed biological control response to pest invasions. Guidelines to assess the feasibility of pre-emptive biological control for high priority pests were recently developed. Emerald ash borer (EAB), Agrilus planipennis, is an invasive woodboring pest of ash (Fraxinus spp.) in North America, European Russia and Ukraine, and is spreading westward into Europe, threatening the future of European ash (Fraxinus excelsior). We applied the aforementioned guidelines to assess the feasibility of pre-emptive biological control in Europe using four EAB parasitoids, already released in North America for CBC. Three of the parasitoids; Oobius agrili, Spathius galinae, and Tetrastichus planipennisi, were found suitable for pre-emptive biological control. The fourth parasitoid, Spathius agrili, was found to have limited establishment in new environmental conditions, and was therefore deemed less suitable for pre-emptive biological control of EAB in Europe. This assessment can inform scientists and regulators in Europe on the most promising EAB parasitoids that should be considered for pre-emptive applications for importation and risk assessment to acquire pre-approval for immediate release should the target pest subsequently be discovered. In turn, this study contributes to the development of biosecurity preparedness against EAB’s imminent spread throughout Europe.

Making waves: Breaking the bottleneck of recycling drinking water treatment residue for practical engineering applications in water pollution control

Drinking water treatment residue (DWTR), an inevitable byproduct of water treatment plants, is typically recycled to control water pollution. DWTR poses a low environmental risk and has the potential to function as a functional material for various applications. However, the practical engineering applications of DWTR are limited. These limitations arise from a disconnect between fundamental research and the practical needs of engineering applications, creating a bottleneck for the effective recycling of DWTR. Previous studies have primarily focused on exploring potential DWTR recycling methods that reuse Al, Fe, Mn, Ca, Si, and organic C. However, the varying properties of DWTR obtained from different water treatment plants tend to differ with respect to potential recycling methods, confusing managers and engineers in using relevant knowledge to guide practical engineering applications. To address this challenge, the author advocates for a shift in research toward establishing guidelines that provide direct guidance for practical engineering applications of DWTR. The key components of these guidelines should include risk assessment, capability evaluation, and environmental application procedures with sustainability assessment to break the bottleneck associated with the recycling of DWTR.

A Regional-Scale Assessment-Based SARS-CoV-2 Variants Control Modeling with Implications for Infection Risk Characterization.

The emergence and progression of highly divergent SARS-CoV-2 variants have posed increased risks to global public health, triggering the significant impacts on countermeasures since 2020. However, in addition to vaccination, the effectiveness of non-pharmaceutical interventions, such as social distancing, masking, or hand washing, on different variants of concern (VOC) remains largely unknown. This study provides a mechanistic approach by implementing a control measure model and a risk assessment framework to quantify the impacts of control measure combinations on the transmissions of five VOC (Alpha, Beta, Delta, Gamma, and Omicron), along with a different perspective of risk assessment application. We applied uncontrollable ratios as an indicator by adopting basic reproduction number (R 0) data collected from a regional-scale survey. A risk assessment strategy was established by constructing VOC-specific dose-response profiles to implicate practical uses in risk characterization when exposure data are available. We found that social distancing alone was ineffective without vaccination in almost all countries and VOC when the median R 0 was greater than two. Our results indicated that Omicron could not be contained, even when all control measure combinations were applied, due to its low threshold of infectivity (~3×10-4 plague-forming unit (PFU) mL-1). To facilitate better decision-making in future interventions, we provide a comprehensive evaluation of how combined control measures impact on different countries and various VOC. Our findings indicate the potential application of threshold estimates of infectivity in the context of risk communication and policymaking for controlling future emerging SARS-CoV-2 variant infections.

Estimating Breakfast Characteristics Using Continuous Glucose Monitoring and Machine Learning in Adults With or at Risk of Type 2 Diabetes

Background: Continuous glucose monitoring (CGM) systems allow detailed assessment of postprandial glucose responses (PPGR), offering new insights into food choices’ impact on dysglycemia. However, current approaches to analyze PPGR using a CGM require manual meal logging, limiting the scalability of CGM-driven applications like personalized nutrition and at-home diabetes risk assessment. Objective: We propose a machine learning (ML) framework to automatically identify and characterize breakfast-related PPGRs from CGM profiles in adults at risk of or living with noninsulin-treated type 2 diabetes (T2D). Methods: Our PPGR estimation framework uses a random forest ML algorithm trained on 15 adults without diabetes who wore a CGM for up to four weeks. The algorithm performance was evaluated on a held-out subset of the participants’ CGM data as well as on an external validation data set of 36 individuals at risk for or with noninsulin-treated T2D. Results: Our algorithm’s estimations of breakfast PPGRs displayed no statistically significant differences to annotated PPGRs, in terms of incremental area under the curve and glucose rise ( P > .05 for both data sets), while a small difference in prebreakfast glucose was found in the nondiabetes data set ( P = .005) but not in the validation T2D data set ( P = .18). Conclusions: We designed an ML framework to automatically estimate the timing of meal events from CGM data in individuals without diabetes and in individuals at risk or with T2D. This could provide a more scalable approach for analyzing postprandial glycemia, increasing the feasibility of CGM-based precision nutrition and diabetes risk assessment applications.

Automotive Consumer Loans Risk Assessment Predictive Modeling Using Generative Adversarial Network and Stacked Autoencoder Neural Networks

In this study, we propose a predictive model for automotive consumer loans risk assessment, leveraging Generative Adversarial Networks (GANs) and Stacked Autoencoder Neural Networks (SAEs). We address issues of high dimensionality and sparsity inherent in big data environments and tackle the class imbalance problem using GANs. Feature selection is effectively carried out using SAEs. Experimental results prove the superiority of our approach over traditional neural networks and our model without GANs or SAEs. The proposed model shows significant potential for application in personal credit risk assessment within automotive finance and beyond. Future work is aimed at extending and improving our model and applying it to other domains.

A simplified model for seismic risk assessment of three-dimensional irregular steel moment frame buildings

Seismic risk assessment of irregular buildings requires nonlinear dynamic analysis and three-dimensional (3D) models. However, creating and analyzing such models can be intricate and time-consuming, making the assessment process challenging. Furthermore, this complexity escalates when evaluating building stocks or generating a dataset of buildings with diverse attributes for use in learning algorithms. Therefore, this paper proposes an energy-based model for irregular steel moment-resisting frame buildings to simplify 3D models and improve the time efficiency of structural assessment compared to conventional 3D models. The proposed model is derived from the fish-bone model but has enhanced configuration, elastic, and inelastic properties. These improvements aim to broaden its applicability to encompass 3D irregular mid-rise buildings. Four-, six-, and nine-story irregular buildings in plan and height are utilized to validate the proposed model. The responses investigated in these buildings include eigenvalues, pushover curves, time history responses, incremental dynamic analysis results, fragility curves, and risk analysis outcomes. The findings demonstrate that the simplified model responses are in close agreement with those of 3D models. The comparison of results between the proposed simplified model and existing models indicates that the accuracy of the simplified model is at least 2.2 times that of the existing models. Moreover, it provides an average error reduction of 60 % in contrast to the existing models. These advancements have been achieved while the proposed simplified model maintains a level of time efficiency comparable to the current models. The results from the analysis of the created buildings using the proposed simplified model indicate that the model is suitable for application in the seismic risk assessment of building inventories and generating datasets for future research involving machine learning-based methodologies.

Estimation of Serum Malondialdehyde (a Marker of Oxidative Stress) as a Predictive Biomarker for the Severity of Coronary Artery Disease (CAD) and Cardiovascular Outcomes.

Background Coronary artery disease (CAD) is influenced by oxidative stress, which is a critical yet often overlooked factor in disease progression. While traditional biomarkers such as cholesterol levels and blood pressure are commonly used, they do not fully capture oxidative damage. Malondialdehyde (MDA), a byproduct of lipid peroxidation, offers additional insights into oxidative stress and CAD severity. Unlike conventional markers, such as low-density lipoprotein (LDL) cholesterol, which primarily reflects lipid levels, and high-sensitivity C-reactive protein (hs-CRP), which indicates inflammation, MDA directly measures oxidative damage. This makes MDA a potentially valuable complement to these traditional biomarkers, providing a more nuanced understanding of CAD risk. Despite its potential, the role of MDA in clinical assessments remains underexplored. This study aims to address this gap by evaluating MDA's effectiveness as a complementary biomarker, enhancing the assessment of CAD risk and progression beyond what is provided by existing markers. Objective This study aims to assess serum MDA levels in relation to CAD severity to explore its potential as a non-invasive biomarker for disease progression and cardiovascular outcomes. Methodology This cross-sectional study was conducted at the Department of Cardiology, Mardan Medical Complex Teaching Hospital, Pakistan, from June 2023 to May 2024. Patients were divided into different groups with varying severity of CAD. The one-way ANOVA was used to assess differences among groups, and Pearson's correlation coefficient explored relationships between MDA and all study variables. Simple linear regression analyzed associations between MDA levels, patient groups, and other variables, controlling for covariates. MDA's potential as a predictive biomarker was assessed through ROC curve analysis, with statistical significance set at a p-value < 0.05. Results A total of 133 patients were included in the study, categorized based on CAD severity into mild (n=71), moderate (n=39), and severe (n=23) groups. Serum MDA levels significantly increased with the severity of CAD. Specifically, MDA levels were 116.61 ± 41.95in the mild group, 253.45 ± 180.29in the moderate group, and peaked at 459.91 ± 149.80in the severe group. The differences in MDA levels among these groups were statistically significant (p < 0.01), supporting the association between higher MDA levels and increased CAD severity. Factors such as BMI, heart rate, blood pressure, and smoking status also significantly influenced MDA levels. Receiver operating characteristic (ROC) curve analysis demonstrated high diagnostic accuracy of MDA for assessing CAD severity, with area under the curve (AUC) values of 0.81 for moderate and 0.94 for severe CAD. Comorbid conditions such as diabetes mellitus were associated with elevated MDA levels. Conclusion Elevated serum MDA serves as a reliable, non-invasive biomarker for predicting CAD severity, with potential applications in clinical risk assessment and management strategies. By identifying patients with elevated oxidative stress early, clinicians can implement timely interventions, potentially slowing disease progression and improving outcomes.

Multimodal Social Sensing for the Spatio-Temporal Evolution and Assessment of Nature Disasters.

Social sensing, using humans as sensors to collect disaster data, has emerged as a timely, cost-effective, and reliable data source. However, research has focused on the textual data. With advances in information technology, multimodal data such as images and videos are now shared on media platforms, aiding in-depth analysis of social sensing systems. This study proposed an analytical framework to extract disaster-related spatiotemporal information from multimodal social media data. Using a pre-trained multimodal neural network and a location entity recognition model, the framework integrates disaster semantics with spatiotemporal information, enhancing situational awareness. A case study of the April 2024 heavy rain event in Guangdong, China, using Weibo data, demonstrates that multimodal content correlates more strongly with rainfall patterns than textual data alone, offering a dynamic perception of disasters. These findings confirm the utility of multimodal social media data and offer a foundation for future research. The proposed framework offers valuable applications for emergency response, disaster relief, risk assessment, and witness discovery, and presents a viable approach for safety risk monitoring and early warning systems.

OS01-04 Review of New Approach Methodologies for application in risk assessment of nanoparticles in the food and feed sector: status and challenges

OS01-04 Review of New Approach Methodologies for application in risk assessment of nanoparticles in the food and feed sector: status and challenges

Satellite-Based Census of Residential Buildings: Application for Climate Risk Assessment

Satellite-Based Census of Residential Buildings: Application for Climate Risk Assessment