Cause Of Fatalities Research Articles

Optimized Federated Learning Algorithm for Breast Cancer Detection Using the Marine Predators Algorithm

Breast cancer stands as the leading cause of cancer-related fatalities among women worldwide, necessitating effective detection methods. However, the current landscape of breast cancer detection poses several challenges, including high costs, complexity, and suboptimal accuracy rates. In response, this paper presents a pioneering approach to breast cancer detection leveraging federated learning, a paradigm that allows model training across distributed data sources without centralized data sharing. Central to this novel detection model is the integration of a Convolutional Neural Network (CNN) architecture and a meta-heuristic optimization algorithm known as the marine predator's algorithm (MPA). CNNs have emerged as powerful tools for biomedical image analysis, enabling automatic feature extraction crucial for detecting abnormalities in medical images. However, optimizing CNNs for detection tasks requires meticulous tuning of hyperparameters across different layers, a process impractical to perform manually due to its complexity and time-intensive nature. To address this challenge, the paper proposes the utilization of the MPA as an optimization algorithm within the federated learning framework. The MPA, inspired by the foraging behaviour of marine predators, enhances the exploration-exploitation trade-off to prevent the model from getting stuck in local optima during training. By incorporating the MPA into the CNN architecture, the proposed model achieves remarkable testing accuracy, surpassing the performance of traditional DCNN models. During testing, the proposed model attained an impressive accuracy rate of 98.32%, showcasing its superior performance compared to the DCNN model, which achieved a 95.95% accuracy. Moreover, the proposed approach outperforms existing algorithms, achieving an accuracy rate of 98%. These results underscore the effectiveness of the federated learning-based approach in enhancing breast cancer detection accuracy. the proposed breast cancer detection model represents a significant advancement in the field, offering a highly accurate and efficient solution to overcome the challenges associated with traditional detection methods. By leveraging federated learning and innovative optimization techniques, this model holds promise for improving early detection rates, ultimately contributing to better patient outcomes in the fight against breast cancer

Primary angiosarcoma of the heart: an unexpected and fatal cause of cardiac tamponade and right atrial rupture.

Primary angiosarcoma of the heart: an unexpected and fatal cause of cardiac tamponade and right atrial rupture.

Characteristics of Work-related Fatal Injuries Among Aged Workers in Republic of Korea

ObjectivesThe present paper aimed to examine whether an aging workforce is associated with an increase in work-related fatal injuries and to explore the underlying reasons for this potential increase. Material and methodsAged workers were defined as those who were at least 55 years old. Work-related fatalities were assessed in aged and young workers who were registered with the workers' compensation system in 2021 in the Republic of Korea. Total waged workers, based on raw data from the Local Area Labor Force Survey in 2021, were used as the denominator to estimate the work-related fatality rates. ResultsMost work-related fatalities in the aged workers occurred among individuals working in the “construction sector” (58.9%), those with “elementary occupations (unskilled workers)” (46.1%), and those with the employment status of “daily worker” (60.8%). The estimated incidence (0.973/10,000) of work-related fatalities among aged workers was about four times higher than that (0.239/10,000) among younger workers. “Falling,” “collision,” “struck by an object,” and “trip and slip” were more frequent types of work-related fatalities among aged workers relative to young workers. The category of “buildings, structures, and surfaces” was a more frequent cause of work-related fatalities among aged workers than among young workers. ConclusionsAged workers had a higher incidence of work-related fatalities than young workers. Frequent engagement in precarious employment and jobs, coupled with the greater physical vulnerability of aged workers, were likely causes of their higher level of work-related fatal injuries.

Abstract 1482: FKBP52 drives diverse HCC phenotypes through context-dependent targeting of p53

Abstract p53 is notably the most frequently mutated protein in hepatocellular carcinoma (HCC), one of the major causes of cancer-related fatalities globally. These mutations often result in some of the most adverse outcomes in HCC, making p53 a prime target for therapeutic intervention. The concept of “hyperstability”of p53 is essential in understanding its gain-of-function and dominant-negative effects in tumorigenesis. However, the mechanisms driving this hyperstability are not completely understood. In this study, we have identified FKBP52, a member of the FK506-binding protein family, as a crucial factor in stabilizing and accumulating p53, irrespective of the type of mutation. Furthermore, our study suggests that FKBP52 could be a valuable biomarker for predicting HCC outcomes. We observed a notable link between FKBP52 levels and the histological differentiation in HCC samples. To corroborate these observations, we performed a series of in vitro and in vivo experiments using liver cancer models. These experiments highlighted the dual nature of FKBP52’s impact on hepatocytes, which varies according to cell type. Given its connection to HCC severity, FKBP52 seems to contribute negatively to the disease’s progression. The insights gained from this study on how FKBP52 influences p53 stabilization are critical. We suggest that targeting the processes upstream of FKBP52 could lead to more effective HCC treatments. These findings emphasize the importance of further investigating this pathway, potentially paving the way for new strategies in liver cancer therapy. Citation Format: Xing Wang, Jingfeng Wu, Chengyan Tang, Xiang Li, Li Zhu, Yaoyu Liu. FKBP52 drives diverse HCC phenotypes through context-dependent targeting of p53 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1482.

The impact of general anesthesia on the outcomes of preterm infants with gestational age less than 32 weeks delivered via cesarean section.

Recent advancements in China's perinatal and neonatal intensive care have significantly reduced neonatal mortality, yet preterm births before 32weeks remain the primary cause of neonatal fatalities and contribute to long-term disabilities. The prognosis of very preterm infants (VPIs) is significantly affected by factors including the intrauterine environment, delivery method and neonatal intensive care. Cesarean section which often used for preterm births has implications that are not fully understood, particularly concerning the type of anesthesia used. This study examines the impact of general anesthesia (GA) during cesarean delivery on VPI outcomes, aiming to identify strategies for mitigating GA-associated risks. This cohort study analyzed 1,029 VPIs born via cesarean section under 32 weeks' gestation at our single-center from 1 January 2018, to 31 December 2022. Detailed medical records, encompassing perioperative information, maternal data and neonatal outcomes were meticulously examined. The primary aim of this investigation was to compare maternal characteristics and neonatal outcomes between VPIs delivered under GA and neuraxial anesthesia (NA). A significance level of p < 0.05 was established. Of the 1,029 VPIs analyzed, 87.95% (n = 905) were delivered via NA and 12.05% (n = 124) via GA. Mothers with hypertensive pregnancy diseases and emergency operations were more inclined to choose GA. VPIs delivered under GA showed a lower Apgar score at one and 5minutes (p < 0.01), increased need for tracheal intubation resuscitation (32.2% vs. 12.2%, p < 0.01) and a greater incidence of severe neurological injury (SNI) (14.5% vs. 5%, p < 0.01). Multivariable analysis revealed GA was significantly associated with lower Apgar scores at one (OR 6.321, 95% CI 3.729-10.714; p < 0.01) and 5minutes (OR 4.535, 95% CI 2.975-6.913; p < 0.01), higher risk of tracheal intubation resuscitation (OR = 3.133, 95% CI = 1.939-5.061; p < 0.01) and SNI (OR = 3.019, 95% CI = 1.615-5.643; p < 0.01). Furthermore, for VPIs delivered under GA, a prolonged interval from skin incision to fetus delivery was associated with a lower 5-min Apgar score (p < 0.01). This study revealed the significant impact of GA on adverse outcomes among VPIs. In cases when GA is required, proactive measures should be instituted for the care of VPIs such as expediting the interval from skin incision to fetal delivery.

Identification of root cause of fatal confined space incidents in Malaysia and strategies for prevention

BackgroundThe high number of confined space incidents in Malaysia suggests that working in a confined space exposes workers to a substantial risk of serious injury or fatality. The rapid development of industry and technology has increased the need to work in confined spaces and significantly impacted the risk management of confined spaces. Utilizing past incident data through investigation will help develop risk management and preventive measures related to confined spaces. As a result, this paper will conduct a fatal incident investigation associated with confined spaces from 2010 to 2020 in Malaysia.MethodologyThe incident investigation uses the improved systematic casual analysis technique (SCAT) to identify the immediate cause and the direct cause. From the identified fatal incidents immediate causes, the study will highlight the main hazard that causes fatal incidents in confined spaces. The preventive measure is developed based on the main hazard of confined spaces that cause fatalities.ResultThe result of the study has identified that physical and atmospheric hazards are the main factors that cause fatal incidents associated with confined spaces in Malaysia. Moreover, from the SCAT checklist, the study has revealed that failing to use PPE properly and hazardous environmental conditions are the main immediate cause of fatal incidents in confined spaces.ConclusionEffective risk management of confined spaces, with a particular focus on addressing atmospheric and physical hazards, is crucial for preventing accidents, especially those incidents resulting in fatalities. Future studies should expand their focus to include injuries and near-miss events in confined spaces, aiming to develop more effective and inclusive safety measures by analyzing a broader range of incidents.

Awareness of lung cancer risk factors and symptoms in Syria: an online cross-sectional study

Background Globally, lung cancer is the leading cause of cancer fatalities and the second most frequent cancer. Population knowledge of the features of lung cancer is a crucial strategy for early diagnosis and decreasing the mortality rate of lung cancer patients. In this study, we aim to assess the Syrian population’s knowledge of lung cancer and its risk factors and to measure awareness of symptoms related to lung cancer. Methods This national cross-sectional study was conducted between October 12 to November 21, 2022, in Syria. We included Syrian people above 18 years from all Syrian governorates. The questionnaire consisted of three categories of questions: sociodemographic information, awareness of lung cancer symptoms, and awareness of lung cancer risk factors. Results Overall, 2251 participants were involved in this research; almost half of them (47.3%) were aged between 21-30 years, and 30.9% indicated they are smoking cigarettes. The overall mean score of knowledge regarding closed questions about risk factors of lung cancer was 4.29; however, the mean score of knowledge regarding open questions about symptoms of lung cancer was 1.52. About half of the study sample (51.3%) indicated that unexplained weight loss is a possible symptom of lung cancer. Our findings showed that cigarette smokers have a lower probability of having adequate knowledge toward lung cancer risk factors than a non-smoker (AOR=0.73, COR=0.68, P-value&lt;0.05). We also defined that females have higher statistically significant odds (AOR=1.3, COR=1.38, P-value&lt;0.05) for being knowledgeable about the symptoms of lung cancer compared to the male sample study. Conclusion According to our findings, there is inadequate knowledge toward lung cancer risk factors and moderate knowledge of lung cancer symptoms. Along with educational programs to raise public knowledge of the dangers of smoking and other LC risk factors, effective tobacco control policy execution is crucial.

Integration of a CNN-based model and ensemble learning for detecting post-earthquake road cracks with deep features

A safe and healthy infrastructure is essential for humanity. Maintenance and repair of roads, which are of great importance, especially in transportation, is essential. Roads can sometimes be damaged due to natural disasters or human factors. In this study, roads damaged as a result of the natural disaster and the 6 February 2023 Gaziantep-Kahramanmaras earthquakes, which had devastating/fatal consequences, were studied. In recent years, the use of machine learning methods in the automatic detection process of road cracks has become widespread. Here, deep learning architectures generally based on multilayer convolutional neural networks have become a popular solution, mainly thanks to their high discrimination power in visual data. In this study, it was aimed to automatically detect cracks on the roads in Gaziantep (Nurdagi) and Adiyaman provinces after the earthquake. The data set used in the experimental process consists of two categories: uncracked and cracked, and was collected by the researchers. ResNet152 deep learning architecture, traditional machine learning algorithms, and an ensemble learning (EL) model were used to classify cracks. Among the machine learning algorithms used in the experiments are support vector machine (SVM), K-nearest neighbors, Adaptive Boosting, and Naive Bayes. In the experimental process, ResNet152 architecture was used for feature extraction, and classification was carried out with machine learning algorithms. In addition, the EL method has been used to improve the success of machine learning algorithms. According to experimental results, the SVM algorithm and EL method gave the most successful results with 98.68% accuracy values. The proposed method can benefit experts in the automatic classification of road cracks. The study can contribute to activating rapid coordination and transportation networks in natural disasters that cause fatal and destructive causes, such as earthquakes.

Retrospective Risk Assessment of Injuries and Fatalities in the Forestry and Logging Workforce in the United States, 2003-2019.

The objectives of this study on the forestry and logging workforce are to: 1) Analyze causes of injuries/fatalities to inform future intervention studies focused on risk mitigation, 2) determine whether there are any trends or associations between work-related risk factors and workplace injuries/fatalities over a 16-year period (2003-2019), and 3) identify knowledge gaps related to injuries and fatalities for future studies to address. Data on fatalities, injuries, and illnesses of the forestry and logging workforce from the United States Bureau of Labor Statistics were analyzed. Correlation analysis (p < .05) was conducted to assess the relationship between causes of forestry and logging workforce fatalities by cause of fatality in the United States. Injury and fatality rates were calculated for each year (fatalities: 2003-2018; injuries: 2005-2019) and time span-specific incidence rates were calculated by cause. Contact with objects and equipment was the primary cause of injuries and fatalities in the forestry and logging workforce during the study period. Transportation-related incidents ranked second as the cause of fatalities, while the category of falls, slips, and trips was the second leading cause of injuries. Gaps in occupational health and safety identified by this study should be collaboratively addressed by researchers and the forestry industry.

Barriers: Location, Functionality, and Method of Access in Childhood Pool/Spa Submersion Incidents, United States, 2000-2017

Drowning is a leading cause of fatality among children in the United States, and residential pools/spas currently account for as much as 80% of these submersion incidents. This study reviewed narrative case reports obtained from the United States Consumer Product Safety Commission (CPSC) to understand barrier location, type, and functionality as methods of pool/spa access for childhood submersion incidents. Retroactive analysis of 1,523 fatal and non-fatal submersion incidents among children aged 13 years old and younger was conducted using the CPSC in depth investigation dataset from 2000-2017. Narrative descriptions were coded according to the attributes of barrier location, functionality, and method of access; to our knowledge, these characteristics of barriers have not been previously investigated to this depth. Outcomes indicate that consumer education campaigns should target households with swimming pools/spas and reiterate key findings that (1) submersion incidents occur even when barriers are present and (2) the importance of multiple layers of protection against unauthorized pool/spa access. Additionally, improved documentation of barriers is imperative to better understand and prevent drowning among children.

Role of non-invasive coronary imaging in stable angina.

Chest pain represents a symptom of significant clinical concern due to the potential for lethal etiologies. Accordingly, it is critical to ascertain the presence of stable angina through various diagnostic tests to inform subsequent therapeutic strategies. Stable angina, while potentially progressing to more severe conditions if left untreated, suffers from a paucity of research regarding its management compared to other more fatal causes of chest pain. Recent advancements in radiological imaging necessitate a re-evaluation of the array and functionality of diagnostic tests, with particular emphasis on prioritizing non-invasive methods such as electrocardiography and echocardiography. This study undertakes a comprehensive review of the literature pertaining to various diagnostic tests for stable angina. We conclude that the management of a patient presenting with chest pain encompasses a continuum of care, beginning with a detailed patient history to estimate pre-test probability and culminating in computed tomography coronary angiography. This continuum is highly individualized, taking into account patient-specific variables, disease burden, and test indications. In an era of rapid research advancement, our findings delineate the optimal sequence of initial diagnostic tests, emphasizing the role of current non-invasive imaging modalities as outlined in standard clinical guidelines.

ERNet: Edge Regularization Network for Cerebral Vessel Segmentation in Digital Subtraction Angiography Images.

Stroke is a leading cause of disability and fatality in the world, with ischemic stroke being the most common type. Digital Subtraction Angiography images, the gold standard in the operation process, can accurately show the contours and blood flow of cerebral vessels. The segmentation of cerebral vessels in DSA images can effectively help physicians assess the lesions. However, due to the disturbances in imaging parameters and changes in imaging scale, accurate cerebral vessel segmentation in DSA images is still a challenging task. In this paper, we propose a novel Edge Regularization Network (ERNet) to segment cerebral vessels in DSA images. Specifically, ERNet employs the erosion and dilation processes on the original binary vessel annotation to generate pseudo-ground truths of False Negative and False Positive, which serve as constraints to refine the coarse predictions based on their mapping relationship with the original vessels. In addition, we exploit a Hybrid Fusion Module based on convolution and transformers to extract local features and build long-range dependencies. Moreover, to support and advance the open research in the field of ischemic stroke, we introduce FPDSA, the first pixel-level semantic segmentation dataset for cerebral vessels. Extensive experiments on FPDSA illustrate the leading performance of our ERNet.

Lethal intimate partner violence and gendered dimensions of the COVID-19 lockdown in Nigeria: evidence from a descriptive analysis of secondary data

IntroductionConcerns emerged over the escalation of intimate partner violence (IPV) as many governments imposed COVID-19 lockdown measures. This paper examined the lethality trends, gender contexts and sources of fatal IPV...

A deep learning approach for robust traffic accident information extraction from online chinese news

AbstractRoad traffic accidents are the leading causes of injuries and fatalities. Understanding the traffic accident occurrence pattern and its contributing factors are prerequisites for effective traffic safety management. The paper proposes a deep learning approach for traffic accident recognition and information extraction from online Chinese news to extract and organize traffic accidents automatically. The approach consists of three modules, including automated news collection, news classification, and traffic accident information extraction. The automated news collection module crawls news from online sources, cleans and organizes it into a general news database with different categories of news. The news classification module robustly recognizes the traffic accident news from all types of news by fusing the sentence‐wise and context‐wise semantic news information. The accident information extraction module extracts the key attributes of traffic accidents (e.g. causes, times, locations) from news text using the SoftLexicon‐BiLSTM‐CRF method. The proposed approach is validated by comparing it with state‐of‐the‐art text mining methods using Chinese news data crawled online. The results show that the approach can achieve a high information extraction performance in terms of precision, recall, and F1‐score. It improves the performance of the best benchmark model (BiLSTM‐CRF) by 18.8% in precision and 12.08% in F1‐score. In addition, the potential value of the automatically extracted accident data is illustrated from online news in complementing traditional authority accident data to drive more effective traffic safety management in practice.

Assessment of Factors Associated with Alcohol Abuse among Youth Aged (15-25) Years in Starch Factory at Lira Municipality Lira District, Uganda

In Uganda, alcohol consumption has been on the rise, with WHO ranking Uganda as the highest consumer in the world. Studies conducted among internally displaced populations in war-torn areas of northern Uganda have revealed that alcoholism is a prevalent issue, particularly among men. This has become a major cause of domestic violence, accidents, and other fatalities within homes. The aim of this study was to assess the factors contributing to alcohol abuse among young people aged 15-25 at Starch Factory at Lira district Uganda. Questionnaires were used to gather data, which was then analyzed manually using pens, papers, and calculators, and later presented in numerical tables using a computer. The study revealed that the main contributing factors to alcohol abuse among young people were environmental, such as the availability of alcohol, local breweries, and persuasive advertisements. Socio-economic factors, including peer pressure, social norms, unemployment stress, and feelings of boredom, also played a significant role. The study also identified the major repercussions of alcohol abuse on health and the economy. In conclusion, the availability of alcohol through numerous breweries and persuasive advertisements, along with social norms, unemployment, boredom, and peer pressure, were found to be the primary drivers of alcohol abuse among young people in this village. The government should therefore organize and fund health talk programs and seminars to raise awareness within the community about the factors contributing to alcohol abuse and its impact on their lives. Additionally, the government should implement strict alcohol policies that regulate the age, level, and time of alcohol consumption. Keywords: Alcohol consumption, Alcohol abuse, Youths, breweries, Persuasive advertisements.

Cytotoxic Effects of Ardisiacrispin A from Labisia pumila on A549 Human Lung Cancer Cells.

Lung cancer is the predominant cause of cancer-related fatalities. This prompted our exploration into the anti-lung cancer efficacy of Labisia pumila, a species meticulously selected from the preliminary screening of 600 plants. Through the strategic implementation of activity-guided fractionation, ardisiacrispin A (1) was isolated utilizing sequential column chromatography. Structural characterization was achieved employing various spectroscopic methods, including nuclear magnetic resonance (NMR), mass spectrometry (MS), and infrared spectroscopy (IR). L. pumila 70% EtOH extract showed significant toxicity in A549 lung cancer cells, with an IC50 value of 57.04 ± 10.28 µg/mL, as well as decreased expression of oncogenes and induced apoptosis. Compound 1, ardisiacrispin A, induced a 50% cell death response in A549 cells at a concentration of 11.94 ± 1.14 µg/mL. The present study successfully investigated ardisiacrispin A extracted from L. pumila leaves, employing a comprehensive spectroscopic approach encompassing NMR, IR, and MS analyses. The anti-lung cancer efficacy of ardisiacrispin A and L. pumila extract was successfully demonstrated for the first time, to the best of our knowledge.

Navigating Lung Cancer: Exploring Progress and Obstacles - A Comprehensive Review

Lung cancer persists as a significant global health dilemma, characterized by elevated incidence and mortality rates worldwide. Despite recent strides in diagnostic methodologies, screening approaches, and therapeutic modalities, lung cancer retains its status as the foremost cause of cancer-related fatalities internationally. The disease is predominantly categorized into two primary forms: small-cell carcinomas and the notably more prevalent non-small-cell carcinomas, constituting approximately 85% of cases. While smoking stands as the primary risk factor for lung cancer, accounting for 80-90% of related deaths, other factors such as dietary habits, air pollution, genetic predisposition, and prior lung conditions play substantial roles. Diagnostic procedures encompass a range of techniques, including radiography, sputum cytology, bronchoscopy, biomarker analysis, among others. The imperative for progress and refinement in diagnostic tools, especially for early detection and monitoring in high-risk demographics, underscores the urgency. Continual research and enhancement of surgical, systemic, and localized treatments are indispensable in the battle against lung cancer. Despite advancements, challenges persist in achieving timely diagnoses, particularly in high-risk groups, and augmenting overall survival rates. This comprehensive review delves into current research advancements in lung cancer, with specific focus on classification, patient well-being, risk factors, diagnostic methodologies, and treatment options.

Repression of YEATS2 induces cellular senescence in hepatocellular carcinoma and inhibits tumor growth

ABSTRACT Hepatocellular carcinoma (HCC) stands as the third leading cause of cancer-related fatalities globally. In this study, we observed a significant increase in the expression level of the YEATS2 gene in HCC patients, and it is negatively correlated with the patients’ survival rate. While we have previously identified the association between YEATS2 and the survival of pancreatic cancer cells, the regulatory mechanisms and significance in HCC are still to be fully elucidated. Our study shows that knockdown (KD) of YEATS2 expression leads to DNA damage, which in turn results in an upregulation of γ-H2A.X expression and activation of the canonical senescence-related pathway p53/p21Cip1. Moreover, our transcriptomic analysis reveals that YEATS2 KD cells can enhance the expression of p21Cip1 via the c-Myc/miR-93-5p pathway, consequently fostering the senescence of HCC cells. The initiation of cellular senescence through dual-channel activation suggests that YEATS2 plays a pivotal regulatory role in the process of cell proliferation. Ultimately, our in vivo research utilizing a nude mouse tumor model revealed a notable decrease in both tumor volume and weight after the suppression of YEATS2 expression. This phenomenon is likely attributable to the attenuation of proliferative cell activity, coupled with a concurrent augmentation in the population of natural killer (NK) cells. In summary, our research results have supplemented the understanding of the regulatory mechanisms of HCC cell proliferation and indicated that targeting YEATS2 may potentially inhibit liver tumor growth.

Pattern of injuries in Two-Wheeler riders involved in Road Traffic Occurrence –an Autopsy Study

Background: Road traffic injuries are India’s sixth leading cause of death, affecting youth and middle-aged individuals. Rapid motorized traffic growth, particularly two-wheelers (72% of vehicles), has increased accidents. Inadequate and non-uniform helmet regulations worsen the situation. Road injuries strain healthcare, demanding hospitalization, and rehabilitation. Kerala experiences significant two-wheeler accidents, often fatal due to head injuries. Studying injury patterns and causes among riders in fatal accidents is vital. Methods: A cross-sectional study was conducted at Govt TD Medical College, Vandanam, Kerala, following approval from the Institutional Ethics Committee. The study included all deaths of two-wheeler riders involved in road traffic accidents with clear vehicle information, excluding decomposed or unidentified bodies. A total of 101 cases were analyzed, recording external and internal injuries directly onto the Proforma during autopsy. Data were summarized using frequencies and proportions, while the chi-square test assessed relationships between categorical variables. Conclusion:The study found that most accidents occurred between 6-9 pm, involving mainly male (91.1%) two-wheeler riders (80.8%). Few wore helmets (5%). Head injuries (97%) were common, often caused by ejection from vehicles (87%). Non-helmet use and ejection related significantly to head injuries, a leading cause of twowheeler accident fatalities

Fatal Crash Occurrence Prediction and Pattern Evaluation by Applying Machine Learning Techniques

Background Highway safety remains a significant issue, with road crashes being a leading cause of fatalities and injuries. While several studies have been conducted on crash severity, few have analyzed and predicted specific types of crashes, such as fatal crashes. Identifying the key factors associated with fatal crashes and predicting their occurrence can help develop effective preventative measures. Objective This study intended to develop cluster analysis and ML-based models using crash data to extract the prominent factors behind fatal crash occurrences and analyze the inherent pattern of variables contributing to fatal crashes. Methods Several branches and categories of supervised ML models have been implemented for fatality prediction and their results have been compared. SHAP analysis was conducted using the ML model to explore the contributing factors of fatal crashes. Additionally, the underlying hidden patterns of fatal crashes have been evaluated using K-means clustering, and specific fatal crash scenarios have been extracted. Results The deep neural networks model achieved 85% accuracy in predicting fatal crashes in Kansas. Factors, such as speed limits, nighttime, darker road conditions, two-lane highways, highway interchange areas, motorcycle and tractor-trailer involvement, and head-on collisions were found to be influential. Moreover, the clusters were able to discern certain scenarios of fatal crashes. Conclusion The study can provide a clear image of the important factors related to fatal crashes, which can be utilized to create new safety protocols and countermeasures to reduce fatal crashes. The results from cluster analysis can facilitate transportation professionals with representative scenarios, which will benefit in identifying potential fatal crash conditions.