—Early detection and prevention of Cataract disease can effectively contribute in reducing the impact of cataracts. In this study, we explore the effectiveness of deep learning algorithms implemented with three pre-trained models —MobileNet VGG19, and ResNet50— for cataract disease detection. These algorithms leverage image processing techniques and have shown promise in various computer vision tasks. Our objective is to predict which algorithm performs best in cataract detection. We use a dataset of retinal fundus images to train and evaluate the models. The results demonstrate the potential of deep learning in early cataract diagnosis, which can significantly improve patient outcomes. Our model was able to achieve an accuracy of 96.33%.
{"title":"Cataract Disease Detection Using Pre-trained Models","authors":"Merna Youssef, Kareem Hassan, Mohanad Deif, Rania Elgohary","doi":"10.21608/iiis.2024.357771","DOIUrl":"https://doi.org/10.21608/iiis.2024.357771","url":null,"abstract":"—Early detection and prevention of Cataract disease can effectively contribute in reducing the impact of cataracts. In this study, we explore the effectiveness of deep learning algorithms implemented with three pre-trained models —MobileNet VGG19, and ResNet50— for cataract disease detection. These algorithms leverage image processing techniques and have shown promise in various computer vision tasks. Our objective is to predict which algorithm performs best in cataract detection. We use a dataset of retinal fundus images to train and evaluate the models. The results demonstrate the potential of deep learning in early cataract diagnosis, which can significantly improve patient outcomes. Our model was able to achieve an accuracy of 96.33%.","PeriodicalId":518706,"journal":{"name":"International Integrated Intelligent Systems","volume":"35 20","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141276290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.21608/iiis.2024.357785
Daniel Halim, Mariam Hanafy, Youssef Lotfy, Mohanad Deif, Rania Elgohary
—Abstract: This paper presents a driver drowsiness detection for accident prevention which is based on the curvature of the eye. Our attempt is to develop a deep learning model that can use the input from a camera in real time by extracting the eyes to detect the drowsiness of the drivers.This paper helps to resolve the problem of drowsiness detection with an accuracy of 96% for test and 99% for validation
{"title":"Real-time Driver Drowsiness Detection Using Deep Neural Networks","authors":"Daniel Halim, Mariam Hanafy, Youssef Lotfy, Mohanad Deif, Rania Elgohary","doi":"10.21608/iiis.2024.357785","DOIUrl":"https://doi.org/10.21608/iiis.2024.357785","url":null,"abstract":"—Abstract: This paper presents a driver drowsiness detection for accident prevention which is based on the curvature of the eye. Our attempt is to develop a deep learning model that can use the input from a camera in real time by extracting the eyes to detect the drowsiness of the drivers.This paper helps to resolve the problem of drowsiness detection with an accuracy of 96% for test and 99% for validation","PeriodicalId":518706,"journal":{"name":"International Integrated Intelligent Systems","volume":"50 18","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141277505","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.21608/iiis.2024.357817
Yara tarek, Rania Elgohary, Mohanad Deif
— automated recognition of medical images poses a significant challenge in the field of medical image processing. These images are obtained from various modalities such as Computed Tomography (CT), Magnetic Resonance Imaging (MRI), etc., and are crucial for diagnosis purposes. In the medical field, brain tumor classification is very important phase for the further treatment. Human interpretation of large number of MRI slices (Normal or Abnormal) may leads to misclassification hence there is need of such a automated recognition system, which can classify the type of the brain tumor. The aim of this study is to detect brain tumor so we identify various features within an image. We extract the feature data from an image Using GLCM , LBP and other filters like Gaussian Filter, Sobel Filter, Laplace Filter, Gabor Filter, Hessian, Prewitt and create a data frame that can be fed into binary classification algorithms like Logistic Regression, KNN and decision tree . The accuracy achieved by Logistic Regression was 72%, KNN was 65% and decision tree was 80%.
{"title":"Brain Tumor Detection Using GLCM and Machine learning Techniques","authors":"Yara tarek, Rania Elgohary, Mohanad Deif","doi":"10.21608/iiis.2024.357817","DOIUrl":"https://doi.org/10.21608/iiis.2024.357817","url":null,"abstract":"— automated recognition of medical images poses a significant challenge in the field of medical image processing. These images are obtained from various modalities such as Computed Tomography (CT), Magnetic Resonance Imaging (MRI), etc., and are crucial for diagnosis purposes. In the medical field, brain tumor classification is very important phase for the further treatment. Human interpretation of large number of MRI slices (Normal or Abnormal) may leads to misclassification hence there is need of such a automated recognition system, which can classify the type of the brain tumor. The aim of this study is to detect brain tumor so we identify various features within an image. We extract the feature data from an image Using GLCM , LBP and other filters like Gaussian Filter, Sobel Filter, Laplace Filter, Gabor Filter, Hessian, Prewitt and create a data frame that can be fed into binary classification algorithms like Logistic Regression, KNN and decision tree . The accuracy achieved by Logistic Regression was 72%, KNN was 65% and decision tree was 80%.","PeriodicalId":518706,"journal":{"name":"International Integrated Intelligent Systems","volume":"36 8","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141274534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-06-01DOI: 10.21608/iiis.2024.357780
Ahmed Gamal, Mohammed El Saeed, Mohanad Deif, Rania Elgohary
:This study addresses the ongoing pursuit of achieving optimal performance in digit recognition tasks, focusing on the widely studied MNIST dataset. Our motivation stems from the challenge of accurately classifying the remaining 1% of images, despite the relatively high 99% accuracy achieved by existing models. In this work, we present a simplified approach to convolutional neural network (CNN) architecture, aiming to streamline model complexity while maintaining or even enhancing performance. Unlike previous approaches, our methodology involves utilizing only two CNN layers with fewer filters, resulting in a reduction in model parameters and learning time. Through rigorous experimentation and evaluation, we demonstrate that our streamlined CNN architecture yields competitive results. Our findings underscore the importance of exploring alternative model architectures and optimization techniques to achieve state-of-the-art performance in digit recognition tasks.
{"title":"Enhanced Convolutional Neural Networks for MNIST Digit Recognition","authors":"Ahmed Gamal, Mohammed El Saeed, Mohanad Deif, Rania Elgohary","doi":"10.21608/iiis.2024.357780","DOIUrl":"https://doi.org/10.21608/iiis.2024.357780","url":null,"abstract":":This study addresses the ongoing pursuit of achieving optimal performance in digit recognition tasks, focusing on the widely studied MNIST dataset. Our motivation stems from the challenge of accurately classifying the remaining 1% of images, despite the relatively high 99% accuracy achieved by existing models. In this work, we present a simplified approach to convolutional neural network (CNN) architecture, aiming to streamline model complexity while maintaining or even enhancing performance. Unlike previous approaches, our methodology involves utilizing only two CNN layers with fewer filters, resulting in a reduction in model parameters and learning time. Through rigorous experimentation and evaluation, we demonstrate that our streamlined CNN architecture yields competitive results. Our findings underscore the importance of exploring alternative model architectures and optimization techniques to achieve state-of-the-art performance in digit recognition tasks.","PeriodicalId":518706,"journal":{"name":"International Integrated Intelligent Systems","volume":"50 21","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141277473","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-02-01DOI: 10.21608/iiis.2024.342001
Alaa Zaghloul, Omar Khaled, Rania Elgohary
: This research study aims to provide a depression detection project that uses text analysis and natural language processing (NLP) to identify symptoms of depression. In order to conduct sentiment analysis on big datasets of tweets, this project will employ a deep learning model. Social media platforms have evolved into places where individuals express their ideas and feelings. Our objective is to create a chat platform that enables users to interact with friends, coworkers, or complete strangers while using text analysis to identify sadness. There are several browsers that can be used to visit the website and guidance on interacting with it. The significance of early depression detection and its possible effects on community well-being—including detrimental effects on local company productivity and healthcare costs—will be emphasized in our research. The purpose of this project is to increase public awareness of the advantages of early identification and to offer a deep learning-based approach to assist people in identifying depression and obtaining the necessary assistance.
{"title":"Depression Detection using Deep Learning Algorithms","authors":"Alaa Zaghloul, Omar Khaled, Rania Elgohary","doi":"10.21608/iiis.2024.342001","DOIUrl":"https://doi.org/10.21608/iiis.2024.342001","url":null,"abstract":": This research study aims to provide a depression detection project that uses text analysis and natural language processing (NLP) to identify symptoms of depression. In order to conduct sentiment analysis on big datasets of tweets, this project will employ a deep learning model. Social media platforms have evolved into places where individuals express their ideas and feelings. Our objective is to create a chat platform that enables users to interact with friends, coworkers, or complete strangers while using text analysis to identify sadness. There are several browsers that can be used to visit the website and guidance on interacting with it. The significance of early depression detection and its possible effects on community well-being—including detrimental effects on local company productivity and healthcare costs—will be emphasized in our research. The purpose of this project is to increase public awareness of the advantages of early identification and to offer a deep learning-based approach to assist people in identifying depression and obtaining the necessary assistance.","PeriodicalId":518706,"journal":{"name":"International Integrated Intelligent Systems","volume":"197 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140529172","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-02-01DOI: 10.21608/iiis.2024.342003
Ahmed El-Gabry, Antonious Atef Saleh, Omar El Saeed
— This study investigates antimicrobial peptides (AMPs), pivotal in combating infections, using accessible machine learning methods. We examined long, medium, and short peptides, focusing on specific features. Initially, supervised classification, guided by a reference paper from fellow researchers in our department, was employed to analyze peptides across several features. This approach provided insights into the effectiveness of these peptides. Subsequently, we adopted unsupervised learning techniques, utilizing tools such as SVM (Support Vector Machines), RF (Random Forest), and KNN (K-Nearest Neighbors). Our findings unveil new insights into the peptides, revealing both anticipated and unexpected patterns. While the supervised approach helped us understand the known characteristics, unsupervised learning allowed for the discovery of hidden analogies and patterns not considered by traditional chemical analysis. This work is significant as it deepens our comprehension of AMPs, paving the way for improved treatments for infections. The study underscores the synergy between machine learning and biochemical insights in the exploration of peptide functionality.
{"title":"Comprehensive Machine Learning Analysis of Long and Middle Peptides: Supervised and Unsupervised Approaches","authors":"Ahmed El-Gabry, Antonious Atef Saleh, Omar El Saeed","doi":"10.21608/iiis.2024.342003","DOIUrl":"https://doi.org/10.21608/iiis.2024.342003","url":null,"abstract":"— This study investigates antimicrobial peptides (AMPs), pivotal in combating infections, using accessible machine learning methods. We examined long, medium, and short peptides, focusing on specific features. Initially, supervised classification, guided by a reference paper from fellow researchers in our department, was employed to analyze peptides across several features. This approach provided insights into the effectiveness of these peptides. Subsequently, we adopted unsupervised learning techniques, utilizing tools such as SVM (Support Vector Machines), RF (Random Forest), and KNN (K-Nearest Neighbors). Our findings unveil new insights into the peptides, revealing both anticipated and unexpected patterns. While the supervised approach helped us understand the known characteristics, unsupervised learning allowed for the discovery of hidden analogies and patterns not considered by traditional chemical analysis. This work is significant as it deepens our comprehension of AMPs, paving the way for improved treatments for infections. The study underscores the synergy between machine learning and biochemical insights in the exploration of peptide functionality.","PeriodicalId":518706,"journal":{"name":"International Integrated Intelligent Systems","volume":"567 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140528559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：