Investigating Injury Outcomes of Horse-and-Buggy Crashes in Rural Michigan by Mining Crash Reports Using NLP and CNN Algorithms DOI Creative Commons
Baraah Qawasmeh, Jun-Seok Oh, Valerian Kwigizile

и другие.

Safety, Год журнала: 2024, Номер 11(1), С. 1 - 1

Опубликована: Дек. 30, 2024

Horse-and-buggy transportation, vital for many rural communities and the Amish population, has been largely overlooked in safety research. This study examines characteristics injury severity of horse-and-buggy roadway crashes Michigan’s areas. Detailed crash data are essential studies, as scene descriptions mainly found narratives diagrams. However, extracting utilizing this information from traffic reports is challenging. research tackles these challenges using image-processing text-mining techniques to analyze diagrams narratives. The employs AlexNet convolutional neural network (CNN) identify extract crashes, analyzing (2020–2023) Michigan UD-10 reports. Natural Language Processing (NLP) also identified primary risk factors narratives, single-word patterns (“unigrams”) sequences three consecutive words (“trigrams”). findings emphasize risks involved interactions on roadways highlight various contributing including distracted or careless actions by motorists, nighttime visibility issues, failure yield, especially elderly drivers. suggests prioritizing riders road public health programs recommends comprehensive measures that could significantly reduce incidence severity, improving overall areas, better signage, driver education, community outreach. Also, highlights potential advanced lead more precise actionable findings, enhancing all users.

Язык: Английский

Prospective Applications of Artificial Intelligence In Fetal Medicine: A Scoping Review of Recent Updates DOI Creative Commons
Elhadi Miskeen, Jaber Alfaifi,

Dalal Mohammed Alhuian

и другие.

International Journal of General Medicine, Год журнала: 2025, Номер Volume 18, С. 237 - 245

Опубликована: Янв. 1, 2025

With the incorporation of artificial intelligence (AI), significant advancements have occurred in field fetal medicine, holding potential to transform prenatal care and diagnostics, promising revolutionize diagnostics. This scoping review aims explore recent updates prospective application AI evaluating its current uses, benefits, limitations. Compiling literature concerning utilization medicine does not appear modify subject or provide an exhaustive exploration electronic databases. Relevant studies, reviews, articles published years were incorporated ensure up-to-date data. The selected works analyzed for common themes, methodologies applied, scope AI's integration into practice. identified several key areas where applications are making strides including screening, diagnosis congenital anomalies, predicting pregnancy complications. AI-driven algorithms been developed analyze complex ultrasound data, enhancing image quality interpretative accuracy. monitoring has also explored, with systems designed identify patterns indicative distress. Despite these advancements, challenges related ethical use AI, data privacy, need extensive validation tools diverse populations noted. benefits immense, offering a brighter future our field. equips us enhanced diagnosis, monitoring, prognostic capabilities, way we approach optimistic outlook underscores further research interdisciplinary partnerships fully leverage driving forward practice medicine.

Язык: Английский

Процитировано

4
Micro-Mobility Safety Assessment: Analyzing Factors Influencing the Micro-Mobility Injuries in Michigan by Mining Crash Reports DOI Creative Commons
Baraah Qawasmeh, Jun-Seok Oh, Valerian Kwigizile

и другие.

Future Transportation, Год журнала: 2024, Номер 4(4), С. 1580 - 1601

Опубликована: Дек. 10, 2024

The emergence of micro-mobility transportation in urban areas has led to a transformative shift mobility options, yet it also brought about heightened traffic conflicts and crashes. This research addresses these challenges by pioneering the integration image-processing techniques with machine learning methodologies analyze crash diagrams. study aims extract latent features from data, specifically focusing on understanding factors influencing injury severity among vehicle crashes Michigan’s areas. Micro-mobility devices analyzed this are bicycles, e-wheelchairs, skateboards, e-scooters. AlexNet Convolutional Neural Network (CNN) was utilized identify various attributes diagrams, enabling recognition classification device collision locations into three categories: roadside, shoulder, bicycle lane. 2023 Michigan UD-10 reports comprising 1174 diverse Subsequently, Random Forest algorithm pinpoint primary their interactions that affect injuries. results suggest roads speed limits exceeding 40 mph most significant factor determining In addition, rider violations motorists left-turning maneuvers associated more severe outcomes. findings emphasize overall effect many different variables, such as improper lane use, violations, hazardous actions users. These demonstrate elevated rates prevalence younger users found be distracted motorists, elderly or those who ride during nighttime.

Язык: Английский

Процитировано

4
Comparative Analysis of AlexNet, ResNet-50, and VGG-19 Performance for Automated Feature Recognition in Pedestrian Crash Diagrams DOI Creative Commons
Baraah Qawasmeh, Jun-Seok Oh, Valerian Kwigizile

и другие.

Applied Sciences, Год журнала: 2025, Номер 15(6), С. 2928 - 2928

Опубликована: Март 8, 2025

Pedestrians, as the most vulnerable road users in traffic crashes, prompt transportation researchers and urban planners to prioritize pedestrian safety due elevated risk growing incidence of injuries fatalities. Thorough crash data are indispensable for research, detailed descriptions scenes actions typically found narratives diagrams. However, extracting analyzing this information from police reports poses significant challenges. This study tackles these issues by introducing innovative image-processing techniques analyze By employing cutting-edge technological methods, research aims uncover extract hidden features Michigan, thereby enhancing understanding prevention such incidents. evaluates effectiveness three Convolutional Neural Network (CNN) architectures—VGG-19, AlexNet, ResNet-50—in classifying multiple These include intersection type (three-leg or four-leg), (divided undivided), presence marked crosswalk (yes no), angle (skewed unskewed), Michigan left turn nearby residentials no). The utilizes 2020–2023 UD-10 reports, comprising 5437 diagrams large urbanized areas 609 rural areas. CNNs underwent comprehensive evaluation using various metrics, including accuracy F1-score, assess their capacity reliably features. results reveal that AlexNet consistently surpasses other models, attaining highest F1-score. highlights critical importance choosing appropriate architecture diagram analysis, particularly context safety. outcomes minimizing errors image classification, especially studies. In addition evaluating model performance, computational efficiency was also considered. regard, emerged efficient model. is precious situations where there limitations on computing resources. contributes novel insights leveraging processing technology, CNNs’ potential use detecting concealed patterns. lay groundwork future offer promise supporting initiatives facilitating countermeasures’ development researchers, planners, engineers, agencies.

Язык: Английский

Процитировано

0
Current State of Artificial Intelligence Model Development in Obstetrics DOI
Lawrence D. Devoe,

M. Muhanna,

James Mäher

и другие.

Obstetrics and Gynecology, Год журнала: 2025, Номер unknown

Опубликована: Июнь 5, 2025

Publications on artificial intelligence (AI) applications have dramatically increased for most medical specialties, including obstetrics. Here, we review the recent pertinent publications AI programs in obstetrics, describe trends specific obstetric problems, and assess AI's possible effects care. Searches were performed PubMed (MeSH), MEDLINE, Ovid, ClinicalTrials.gov, Google Scholar, Web of Science using a combination keywords text words related to “obstetrics,” “pregnancy,” “artificial intelligence,” “machine learning,” “deep “neural networks,” articles published between June 1, 2019, May 31, 2024. A total 1,768 met at least one search criterion. After eliminating reviews, duplicates, retractions, inactive research protocols, unspecified programs, non–English-language articles, 207 remained further review. Most studies conducted outside United States, nonobstetric journals, focused risk prediction. Study population sizes ranged widely from 10 953,909, model performance abilities also varied widely. Evidence quality was assessed by description construction, predictive accuracy, whether validation had been performed. patient groups differing considerably U.S. populations, rendering their generalizability patients uncertain. Artificial ultrasound imaging issues are those likely influence current Other promising models include early screening spontaneous preterm birth, preeclampsia, gestational diabetes mellitus. The rate which being virtually guarantees that numerous will eventually be introduced into future practice. Very few deployed practice, more high-quality needed with high accuracy generalizability. Assuming these conditions met, there an urgent need educate students, postgraduate trainees practicing physicians understand how effectively safely implement this technology.

Язык: Английский

Процитировано

0
Investigating Injury Outcomes of Horse-and-Buggy Crashes in Rural Michigan by Mining Crash Reports Using NLP and CNN Algorithms DOI Creative Commons
Baraah Qawasmeh, Jun-Seok Oh, Valerian Kwigizile

и другие.

Safety, Год журнала: 2024, Номер 11(1), С. 1 - 1

Опубликована: Дек. 30, 2024

Horse-and-buggy transportation, vital for many rural communities and the Amish population, has been largely overlooked in safety research. This study examines characteristics injury severity of horse-and-buggy roadway crashes Michigan’s areas. Detailed crash data are essential studies, as scene descriptions mainly found narratives diagrams. However, extracting utilizing this information from traffic reports is challenging. research tackles these challenges using image-processing text-mining techniques to analyze diagrams narratives. The employs AlexNet convolutional neural network (CNN) identify extract crashes, analyzing (2020–2023) Michigan UD-10 reports. Natural Language Processing (NLP) also identified primary risk factors narratives, single-word patterns (“unigrams”) sequences three consecutive words (“trigrams”). findings emphasize risks involved interactions on roadways highlight various contributing including distracted or careless actions by motorists, nighttime visibility issues, failure yield, especially elderly drivers. suggests prioritizing riders road public health programs recommends comprehensive measures that could significantly reduce incidence severity, improving overall areas, better signage, driver education, community outreach. Also, highlights potential advanced lead more precise actionable findings, enhancing all users.

Язык: Английский

Процитировано

1