An Aseptic Approach Towards Skin Lesion Localization and Grading using Deep Learning and Harris Hawks Optimization DOI Creative Commons
Hossam Magdy Balaha, Asmaa El-Sayed Hassan, Eman M. El-Gendy

и другие.

Research Square (Research Square), Год журнала: 2022, Номер unknown

Опубликована: Авг. 9, 2022

Abstract Skin cancer is the most common form of cancer. Hence, lives millions people are affected by this every year. Approximately, it predicted that total number cases will double in next fifty years. It an expensive procedure to discover skin types early stages. Additionally, survival rate reduces as progresses. The current study proposes aseptic approach toward lesion detection, classification, and segmentation using deep learning a meta-heuristic optimizer called Harris Hawks Optimization Algorithm (HHO). utilized manual automatic approaches. used when dataset has no masks use while used, U-Net models, build adaptive model. HHO achieve optimization hyperparameters 5 pre-trained CNN models (i.e., VGG16, VGG19, DenseNet169, DenseNet201, MobileNet). Two collected "Melanoma Cancer Dataset 10000 Images" "Skin ISIC" dataset) from two publically available sources. For segmentation, best-reported scores 0.15908, 91.95%, 0.08864, 0.04313, 0.02072, 0.20767 terms loss, accuracy, Mean Absolute Error, Squared Logarithmic Root respectively. dataset, applied experiments, best reported overall accuracy 97.08% DenseNet169 96.06% MobileNet After computing results, suggested compared with 9 related studies.

Язык: Английский

Skin cancer diagnosis based on deep transfer learning and sparrow search algorithm DOI Creative Commons
Hossam Magdy Balaha, Asmaa El-Sayed Hassan

Neural Computing and Applications, Год журнала: 2022, Номер 35(1), С. 815 - 853

Опубликована: Сен. 23, 2022

Abstract Skin cancer affects the lives of millions people every year, as it is considered most popular form cancer. In USA alone, approximately three and a half million are diagnosed with skin annually. The survival rate diminishes steeply progresses. Despite this, an expensive difficult procedure to discover this type in early stages. study, threshold-based automatic approach for detection, classification, segmentation utilizing meta-heuristic optimizer named sparrow search algorithm (SpaSA) proposed. Five U-Net models (i.e., U-Net, U-Net++, Attention V-net, Swin U-Net) different configurations utilized perform process. Besides SpaSA used optimization hyperparameters using eight pre-trained CNN VGG16, VGG19, MobileNet, MobileNetV2, MobileNetV3Large, MobileNetV3Small, NASNetMobile, NASNetLarge). dataset gathered from five public sources which two types datasets generated 2-classes 10-classes). For segmentation, concerning “skin classification” dataset, best reported scores by U-Net++ DenseNet201 backbone architecture 0.104, $$94.16\%$$ 94.16 % , $$91.39\%$$ 91.39 $$99.03\%$$ 99.03 $$96.08\%$$ 96.08 $$96.41\%$$ 96.41 $$77.19\%$$ 77.19 $$75.47\%$$ 75.47 terms loss, accuracy, F1-score, AUC, IoU, dice, hinge, squared respectively, while “PH2” 0.137, $$94.75\%$$ 94.75 $$92.65\%$$ 92.65 $$92.56\%$$ 92.56 $$92.74\%$$ 92.74 $$96.20\%$$ 96.20 $$86.30\%$$ 86.30 $$69.28\%$$ 69.28 $$68.04\%$$ 68.04 precision, sensitivity, specificity, respectively. “ISIC 2019 2020 Melanoma” overall accuracy applied experiments $$98.27\%$$ 98.27 MobileNet model. Similarly, “Melanoma Classification (HAM10K)” $$98.83\%$$ 98.83 diseases image” $$85.87\%$$ 85.87 MobileNetV2 After computing results, suggested compared 13 related studies.

Язык: Английский

Процитировано

86

Prostate cancer grading framework based on deep transfer learning and Aquila optimizer DOI Creative Commons
Hossam Magdy Balaha,

Ahmed Osama Shaban,

Eman M. El-Gendy

и другие.

Neural Computing and Applications, Год журнала: 2024, Номер 36(14), С. 7877 - 7902

Опубликована: Фев. 22, 2024

Abstract Prostate cancer is the one of most dominant among males. It represents leading death causes worldwide. Due to current evolution artificial intelligence in medical imaging, deep learning has been successfully applied diseases diagnosis. However, recent studies prostate classification suffers from either low accuracy or lack data. Therefore, present work introduces a hybrid framework for early and accurate segmentation using learning. The proposed consists two stages, namely stage stage. In stage, 8 pretrained convolutional neural networks were fine-tuned Aquila optimizer used classify patients normal ones. If patient diagnosed with cancer, segmenting cancerous spot overall image U-Net can help diagnosis, here comes importance trained on 3 different datasets order generalize framework. best reported accuracies are 88.91% MobileNet “ISUP Grade-wise Cancer” dataset 100% ResNet152 “Transverse Plane Dataset” precisions 89.22% 100%, respectively. model gives an average AUC 98.46% 0.9778, respectively, “PANDA: Resized Train Data (512 × 512)” dataset. results give indicator acceptable performance

Язык: Английский

Процитировано

19

A vision-based deep learning approach for independent-users Arabic sign language interpretation DOI Creative Commons

Mostafa Magdy Balaha,

Sara El-Kady,

Hossam Magdy Balaha

и другие.

Multimedia Tools and Applications, Год журнала: 2022, Номер 82(5), С. 6807 - 6826

Опубликована: Авг. 10, 2022

Abstract More than 5% of the people around world are deaf and have severe difficulties in communicating with normal according to World Health Organization (WHO). They face a real challenge express anything without an interpreter for their signs. Nowadays, there lot studies related Sign Language Recognition (SLR) that aims reduce this gap between as it can replace need interpreter. However, challenges facing sign recognition systems such low accuracy, complicated gestures, high-level noise, ability operate under variant circumstances generalize or be locked limitations. Hence, many researchers proposed different solutions overcome these problems. Each language has its signs very challenging cover all languages’ The current study objectives: (i) presenting dataset 20 Arabic words, (ii) proposing deep learning (DL) architecture by combining convolutional neural network (CNN) recurrent (RNN). suggested reported 98% accuracy on presented dataset. It also 93.4% 98.8% top-1 top-5 accuracies UCF-101

Язык: Английский

Процитировано

55

Comprehensive machine and deep learning analysis of sensor-based human activity recognition DOI
Hossam Magdy Balaha, Asmaa El-Sayed Hassan

Neural Computing and Applications, Год журнала: 2023, Номер 35(17), С. 12793 - 12831

Опубликована: Март 8, 2023

Язык: Английский

Процитировано

32

A survey on deep learning models for detection of COVID-19 DOI Open Access
Javad Mozaffari, Abdollah Amirkhani, Shahriar B. Shokouhi

и другие.

Neural Computing and Applications, Год журнала: 2023, Номер 35(23), С. 16945 - 16973

Опубликована: Май 27, 2023

Язык: Английский

Процитировано

25

A concentrated machine learning-based classification system for age-related macular degeneration (AMD) diagnosis using fundus images DOI Creative Commons

Aya A. Abd El-Khalek,

Hossam Magdy Balaha,

Norah Saleh Alghamdi

и другие.

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Янв. 29, 2024

Abstract The increase in eye disorders among older individuals has raised concerns, necessitating early detection through regular examinations. Age-related macular degeneration (AMD), a prevalent condition over 45, is leading cause of vision impairment the elderly. This paper presents comprehensive computer-aided diagnosis (CAD) framework to categorize fundus images into geographic atrophy (GA), intermediate AMD, normal, and wet AMD categories. crucial for precise age-related enabling timely intervention personalized treatment strategies. We have developed novel system that extracts both local global appearance markers from images. These are obtained entire retina iso-regions aligned with optical disc. Applying weighted majority voting on best classifiers improves performance, resulting an accuracy 96.85%, sensitivity 93.72%, specificity 97.89%, precision 93.86%, F1 ROC 95.85%, balanced 95.81%, sum 95.38%. not only achieves high but also provides detailed assessment severity each retinal region. approach ensures final aligns physician’s understanding aiding them ongoing follow-up patients.

Язык: Английский

Процитировано

14

A multi-variate heart disease optimization and recognition framework DOI Creative Commons
Hossam Magdy Balaha,

Ahmed Osama Shaban,

Eman M. El-Gendy

и другие.

Neural Computing and Applications, Год журнала: 2022, Номер 34(18), С. 15907 - 15944

Опубликована: Май 2, 2022

Abstract Cardiovascular diseases (CVD) are the most widely spread all over world among common chronic diseases. CVD represents one of main causes morbidity and mortality. Therefore, it is vital to accurately detect existence heart help save patient life prescribe a suitable treatment. The current evolution in artificial intelligence plays an important role helping physicians diagnose different In present work, hybrid framework for detection using medical voice records suggested. A that consists four layers, namely “Segmentation” Layer, “Features Extraction” “Learning Optimization” “Export Statistics” Layer proposed. first layer, novel segmentation technique based on variable durations directions (i.e., forward backward) Using proposed technique, 11 datasets with 14,416 numerical features generated. second layer responsible feature extraction. Numerical graphical extracted from resulting datasets. third passed 5 Machine Learning (ML) algorithms, while 8 Convolutional Neural Networks (CNN) transfer learning select configurations. Grid Search Aquila Optimizer (AO) used optimize hyperparameters ML CNN configurations, respectively. last output validated performance metrics. best-reported metrics (1) 100% accuracy algorithms including Extra Tree Classifier (ETC) Random Forest (RFC) (2) 99.17% CNN.

Язык: Английский

Процитировано

35

AutYOLO-ATT: an attention-based YOLOv8 algorithm for early autism diagnosis through facial expression recognition DOI Creative Commons

Reham Hosney,

Fatma M. Talaat, Eman M. El-Gendy

и другие.

Neural Computing and Applications, Год журнала: 2024, Номер 36(27), С. 17199 - 17219

Опубликована: Июнь 6, 2024

Abstract Autism Spectrum Disorder (ASD) is a developmental condition resulting from abnormalities in brain structure and function, which can manifest as communication social interaction difficulties. Conventional methods for diagnosing ASD may not be effective the early stages of disorder. Hence, diagnosis crucial to improving patient's overall health well-being. One alternative method autism facial expression recognition since autistic children typically exhibit distinct expressions that aid distinguishing them other children. This paper provides deep convolutional neural network (DCNN)-based real-time emotion system kids. The proposed designed identify six emotions, including surprise, delight, sadness, fear, joy, natural, assist medical professionals families recognizing intervention. In this study, an attention-based YOLOv8 (AutYOLO-ATT) algorithm proposed, enhances model's performance by integrating attention mechanism. outperforms all classifiers metrics, achieving precision 93.97%, recall 97.5%, F1-score 92.99%, accuracy 97.2%. These results highlight potential real-world applications, particularly fields where high essential.

Язык: Английский

Процитировано

7

A Comprehensive Review of Machine Learning Used to Combat COVID-19 DOI Creative Commons
Rahul Gomes, Connor Kamrowski, Jordan Langlois

и другие.

Diagnostics, Год журнала: 2022, Номер 12(8), С. 1853 - 1853

Опубликована: Июль 31, 2022

Coronavirus disease (COVID-19) has had a significant impact on global health since the start of pandemic in 2019. As June 2022, over 539 million cases have been confirmed worldwide with 6.3 deaths as result. Artificial Intelligence (AI) solutions such machine learning and deep played major part this for diagnosis treatment COVID-19. In research, we review these modern tools deployed to solve variety complex problems. We explore research that focused analyzing medical images using AI models identification, classification, tissue segmentation disease. also prognostic were developed predict outcomes optimize allocation scarce resources. Longitudinal studies conducted better understand COVID-19 its effects patients period time. This comprehensive different methods modeling efforts will shed light role what path it intends take fight against

Язык: Английский

Процитировано

27

A variate brain tumor segmentation, optimization, and recognition framework DOI
Hossam Magdy Balaha, Asmaa El-Sayed Hassan

Artificial Intelligence Review, Год журнала: 2022, Номер 56(7), С. 7403 - 7456

Опубликована: Дек. 15, 2022

Язык: Английский

Процитировано

22