Multiviewunet: A Deep Learning Surrogate for Wall Shear Stress Prediction in Aortic Aneurysmal Diseases DOI
Md. Ahasan Atick Faisal, Onur Mutlu, Sakib Mahmud

et al.

Published: Jan. 1, 2023

Computational Fluid Dynamics (CFD) analysis is widely used to simulate hemodynamics and investigate the biofluid mechanics of different tissue, whole organs, tissue–medical device interactions. However, CFD simulations are time-consuming computationally expensive; hence not readily available practical for patient-specific time-sensitive clinical applications prohibiting quick responses from clinicians. Disturbed known influence progression many cardiac conditions. Aorta main blood artery in body diseases this vessel very common. One such condition Abdominal Aortic Aneurysm (AAA), where abdominal aorta widens has risk rupture. Precise determination Wall Shear Stress (WSS) on aneurysmal wall essential assess rupture tissue. In study, we have proposed a Deep Learning (DL) surrogate estimating aortic WSS distribution. The DL model was created trained receive input output distributions directly, bypassing procedure. A novel way analyzing geometry-to-geometry problems also using domain transformation, which compatible with existing state-of-the-art Neural Networks (NN). framework, MultiViewUnet, 23 real 230 synthetic geometries. algorithm predicted stress an average Normalized Mean Absolute Error (NMAE) 0.362%. We believe our will open up new dimensions precise levels important.

Language: Английский

RHYTHMI: A Deep Learning-Based Mobile ECG Device for Heart Disease Prediction DOI Creative Commons
Alaa Eleyan, Ebrahim Alboghbaish,

Abdulwahab AlShatti

et al.

Applied System Innovation, Journal Year: 2024, Volume and Issue: 7(5), P. 77 - 77

Published: Aug. 29, 2024

Heart disease, a global killer with many variations like arrhythmia and heart failure, remains major health concern. Traditional risk factors include age, cholesterol, diabetes, blood pressure. Fortunately, artificial intelligence (AI) offers promising solution. We have harnessed the power of AI, specifically deep learning convolutional neural networks (CNNs), to develop Rhythmi, an innovative mobile ECG diagnosis device for disease detection. Rhythmi leverages extensive medical data from databases MIT-BIH BIDMC. These empower training testing developed model analyze signals accuracy, precision, sensitivity, specificity, F1-score in identifying arrhythmias other conditions, performances reaching 98.52%, 98.55%, 99.26%, respectively. Moreover, we tested real time using single-lead sensor. This user-friendly prototype captures signal, transmits it Rhythmi’s dedicated website, provides instant feedback on patient’s health. The addresses main problems traditional diagnostic devices such as accessibility, cost, mobility, complexity, integration. However, believe that despite results, our system will still need intensive clinical validation future.

Language: Английский

Citations

2
ImageLM: Interpretable image-based learner modelling for classifying learners’ computational thinking DOI Creative Commons
Danial Hooshyar, Yeongwook Yang

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 238, P. 122283 - 122283

Published: Oct. 21, 2023

Predictive learner modelling is crucial for personalized education. While convolutional neural networks (CNNs) have shown great success in education, their potential via image data unexplored. This research introduces a novel and interpretable approach Image-based Learner Modelling (ImageLM) using CNNs transfer learning to model learners' performance accordingly classify computational thinking solutions. The integrates Grad-CAM, enabling it provide insights into its decision-making process. Findings show that our custom CNN outperforms other models (namely ResNet, VGG, Inception), with 83% accuracy predicting solution correctness. More importantly, the ImageLM identifies regions contribute most predictions, shedding light on knowledge advancing toward trustworthy AI These results underline of utilizing imagery from activities during process predict performance, especially challenging environments like programming where traditional feature extraction might struggle.

Language: Английский

Citations

6
Adaptive Learning Based on Artificial Intelligence to Overcome Student Academic Inequalities DOI Creative Commons

Faridul Ansor,

Nur Aisyah Zulkifli,

Dwi Susi Miftakhul Jannah

et al.

Journal of Social Science Utilizing Technology, Journal Year: 2023, Volume and Issue: 1(4), P. 202 - 213

Published: Dec. 14, 2023

Background. In the context of higher education, academic inequality is a serious obstacle in achieving equitable learning outcomes among students. Factors such as educational background, styles, and differences mastery material are main triggers for this inequality. To overcome challenge, innovative approaches adaptive based on Artificial Intelligence (AI) have emerged potential solution. Purpose. This research aims to investigate AI-based overcoming By combining AI technology, seeks provide personalized solutions tailored each student's needs. Method. uses quantitative methods with survey model. A total 20 respondents were selected representatively their views experiences, preferences regarding learning. provides relevant data understand whether implementation can be considered an effective measure reduce Results. The results show that majority face difficulties understanding course general. However, most also expressed openness use positive perception indication success implementing technology solution Conclusion. Taking into account results, promising align needs individual Although challenges remain, initial impetus further exploration application technologies equity education settings.

Language: Английский

Citations

5
Assessing students’ handwritten text productions: A two-decades literature review DOI
Lenardo Chaves e Silva, Álvaro Sobrinho, Thiago Cordeiro

et al.

Expert Systems with Applications, Journal Year: 2024, Volume and Issue: 250, P. 123780 - 123780

Published: March 21, 2024

Language: Английский

Citations

1
The educational resource management based on image data visualization and deep learning DOI Creative Commons

Xudong Liu

Heliyon, Journal Year: 2024, Volume and Issue: 10(13), P. e32972 - e32972

Published: June 20, 2024

In order to address issues such as inaccurate education resource positioning and inefficient utilization, this study optimizes the Educational Resource Management System (ERMS) by combining image data visualization techniques with convolutional neural networks (CNNs) technology in deep learning. Firstly, crucial role of ERMS teaching is analyzed. Secondly, application CNNs system explained, along associated challenges. Finally, optimizing model architecture validating experimental data, rationality proposed confirmed. Experimental results indicate a significant improvement various performance metrics compared traditional models. The recognition accuracy on Mnist dataset reaches 98.1%, notably, cifar-10 dataset, optimized achieves an close 98.3% improved runtime reduced only 640.4 seconds. Additionally, through systematic simulation experiments, designed shown fully meet earlier requirements for functionality, feasibility study. Therefore, holds high practical value provides meaningful insights into optimization.

Language: Английский

Citations

1
A Plug-In for Cognitive Diagnosis Method Based on Correlation Representation Under Long-Tailed Distribution DOI
Yuhong Zhang, Mian Wang, Tiancheng He

et al.

Published: Jan. 1, 2024

Cognitive diagnosis is a fundamental task in intelligence education, which aims to discover students' proficiency for specific knowledge concepts. Existing cognitive models are trained based on sufficient answering records. In applications, however, these records usually follow long-tailed distribution, i.e., there only few students with but large number of handful The sparsity poses challenge diagnosis. To this end, plug-in correlation representation proposed address under which, the between head and tail learned Specially, representations view both state learning mode, node sub-graph respectively. Then used as enhance their related exercise With enhanced representations, performance improved. Extensive experiments evaluate improvement good compatibility our component. Our code available at https://github.com/joyce99/Wangmian.

Language: Английский

Citations

0
A Tertiary Study for Process Mining DOI Creative Commons
Elia Kouzari, Ioannis Stamelos

Algorithms, Journal Year: 2024, Volume and Issue: 17(12), P. 548 - 548

Published: Dec. 2, 2024

Background: This tertiary study lists the secondary studies published in process mining domain and provides an analysis related to a set of research questions. It is first this area. The objective provide information about available mining, respond questions relating thematic areas covered studies, as well trends regarding their quality, report on findings for publication venues, citations, guidelines used, demographics. Method: A based systematic up March 2023. total 25 have been identified following application inclusion/exclusion criteria quality assessment. Results: most popular addressed are technologies applications healthcare. medium score 3.5. introduced by Kitchenham over years preferred field. There no trend number primary included mining. Conclusion: Although numerous exist there still room more research, specifically highlighted study. Future researchers can use reference, they also listed topics dive deep into issues identified.

Language: Английский

Citations

0
К вопросу об интеграции искусственного интеллекта в систему высшего образования: мнение преподавателей DOI Creative Commons
Yelena SPIRINA

Bulletin of the Karaganda University Pedagogy series, Journal Year: 2024, Volume and Issue: 11629(4), P. 136 - 145

Published: Dec. 30, 2024

В статье рассмотрены вопросы внедрения и применения искусственного интеллекта (ИИ) в образовательном процессе системы высшего образования. Проведен анализ психолого- педагогической научной литературы для уточнения понятий «искусственный интеллект», «адаптивное обучение», «персонализированное обучение». Выделены преимущества преподавании обучении студентов, например, предоставления оперативной обратной связи, разработки адаптивной траектории обучения, прогнозирования успеваемости, а также помощи преподавателям при разработке учебного контента оценочных материалов с учетом индивидуальных особенностей обучающихся. Целью исследования является выявление интереса доверия преподавателей к технологиям инструментам использования их процессе. Анализ текущей ситуации результаты опроса 111 педагогов 6 ведущих университетов Республики Казахстан выявили ряд проблем. Результаты показали, что респонденты понимают положительно оценивают потенциал инструментов интеллекта, частности адаптивного но большинство (более 50 %) сталкиваются ограничениями приложений своей деятельности, так как не обладают достаточными знаниями области технологий интеллекта. этой связи авторы приходят выводу о необходимости повышения квалификации целью эффективного подтверждают необходимость цифровой платформы реализации персонализированного обучения системе

Language: Русский

Citations

0
Advancing path of digital empowerment in the reform of intelligent teaching design DOI
Jing Wang,

Fan Zhang,

Chang Shu

et al.

Published: Nov. 22, 2024

Language: Английский

Citations

0
IMPROVING THE IMAGE-TO-SPEECH SYSTEM ACCURACY THROUGH INTEGRATION OF OPTICAL CHARACTER RECOGNITION AND LANGUAGE PROCESSING TECHNIQUES DOI Open Access

K.H. Nikoghosyan,

E.A. Harutyunyan,

D.M. Galstyan

et al.

Published: Jan. 1, 2023

Image-to-speech systems are a type of technology allowing for the conversion visual information, such as images or videos, into auditory output. These use complex algorithms and machine learning techniques to recognize describe content, individuals who visually impaired blind access in-formation that would otherwise be inaccessible them. becoming increasingly sophisticated can integrated variety devices, from smartphones smart glasses. This article presents an approach improving accuracy image-to-speech system by incorporating multiple techniques. The proposed begins using Tesseract, optical character recognition (OCR) engine, extract text infor-mation images. However, OCR is often imperfect produces errors, which impact models. To address this issue, Text-Davinci-002 engine was applied post-processing output, help correct errors improve extracted text. Finally, Microsoft Speech API employed in order generate speech By integrating these three techniques, significantly improved. An example generated synthetic dataset showed both on word levels, also perform punctuation error correction. useful various applications, including reading images, translating written speech, assisting people with im-pairments.

Language: Английский

Citations

0