Nahed Tawfik
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Nahed Tawfik has a total of 35 co-authors. Here are the top co-authors by number of shared publications:
Fathi E. Abd El-Samie
Department of Information Technology, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia
4 shared publicationsAbeer D. Algarni
Department of Information Technology, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia
2 shared publications- Alok Kumar
School of Engineering and Technology, Chhatrapati Shahu Ji Maharaj University, Kanpur, India
2 shared publications Ankur Prakash
MNNIT Allahabad, India
2 shared publications- Deepak Kumar Verma
School of Engineering and Technology, CSJM University, Kanpur, India
2 shared publications Heba A. Elnemr
Department of Computer and Software Engineering, Misr University for Science and Technology, Giza, Egypt
2 shared publications
Latest Publications
- Open Access
- 19 August 2026
- Published by Elsevier in Intelligent Systems with Applications
- Vol. 31
Trustworthy fall detection: A responsible AI approach for real-time video monitoring of the elderly
- 19 August 2026
- Published by Elsevier in Intelligent Systems with Applications
- Vol. 31
Open AccessTrustworthy fall detection: A responsible AI approach for real-time video monitoring of the elderly
- Open Access
- 9 February 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
PCSA-Net: pyramid channel and spatial attention network for multiclass renal disease diagnosis using CT images
Kidney failure represents a pressing global health concern, further exacerbated by the widespread shortage of nephrologists, thereby necessitating the development of ِِِArtificial Intelligence (AI)-driven systems for automated renal disease diagnosis. This study focuses on the diagnosis of three major renal conditions: kidney stones, tumors, and cysts. Recent advancements in Deep Learning (DL) have highlighted the potential of attention mechanisms in enhancing the performance of Convolutional Neural Networks (CNNs), particularly in medical image analysis. In this context, we propose a novel method termed Pyramid Channel and Spatial Attention (PCSA), which depends on pyramidal multiscale convolution to reconstruct feature representations by extracting both spatial and channel attention weights. This dual-weight extraction facilitates the precise integration of multiscale contextual information, thereby improving the model capability to localize and focus on complex regions within medical images. The PCSA module is designed as a plug-and-play component that can be seamlessly integrated into various CNN backbone architectures to enhance diagnostic accuracy. To validate its effectiveness, we incorporate the PCSA module into several backbone networks and evaluate its performance. Experimental results demonstrate that PCSA-enhanced networks outperform multiple state-of-the-art image classification methods, achieving superior accuracy in renal disease classification. Although the current study focuses on three specific renal conditions, the modular architecture of PCSA-Net allows for future adaptation to a broader spectrum of renal pathologies. These findings underscore the potential of the proposed PCSA module to support automated, accurate, and scalable kidney disease diagnosis in clinical settings. The modular design also enhances the model suitability for real-world deployment, enabling integration into diverse diagnostic workflows.
- 9 February 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessPCSA-Net: pyramid channel and spatial attention network for multiclass renal disease diagnosis using CT images
Kidney failure represents a pressing global health concern, further exacerbated by the widespread shortage of nephrologists, thereby necessitating the development of ِِِArtificial Intelligence (AI)-driven systems for automated renal disease diagnosis. This study focuses on the diagnosis of three major renal conditions: kidney stones, tumors, and cysts. Recent advancements in Deep Learning (DL) have highlighted the potential of attention mechanisms in enhancing the performance of Convolutional Neural Networks (CNNs), particularly in medical image analysis. In this context, we propose a novel method termed Pyramid Channel and Spatial Attention (PCSA), which depends on pyramidal multiscale convolution to reconstruct feature representations by extracting both spatial and channel attention weights. This dual-weight extraction facilitates the precise integration of multiscale contextual information, thereby improving the model capability to localize and focus on complex regions within medical images. The PCSA module is designed as a plug-and-play component that can be seamlessly integrated into various CNN backbone architectures to enhance diagnostic accuracy. To validate its effectiveness, we incorporate the PCSA module into several backbone networks and evaluate its performance. Experimental results demonstrate that PCSA-enhanced networks outperform multiple state-of-the-art image classification methods, achieving superior accuracy in renal disease classification. Although the current study focuses on three specific renal conditions, the modular architecture of PCSA-Net allows for future adaptation to a broader spectrum of renal pathologies. These findings underscore the potential of the proposed PCSA module to support automated, accurate, and scalable kidney disease diagnosis in clinical settings. The modular design also enhances the model suitability for real-world deployment, enabling integration into diverse diagnostic workflows.
- 17 January 2025
- Published by IGI Global Scientific Publishing
- p. 73-102
Empowering Education Through Data Mining Tools
In educational field due to recent advancements, digitization tools and worldwide epidemic scenarios, there is evolution of large number of online learning platforms as well as educational institutions starting from primary level to high education which is ultimately leading to the large amount of data and its handling that can be successfully achieved through Data Mining tools and techniques for extraction of relevant information. In this chapter it have been enlisted the ways utilized by researchers previously and currently for examining the way data mining was handled, moreover most updated trends for data mining in education field, as well as for evaluating the possibility for utilization of machine learning techniques in educational field. This work also explores the application of different Data Mining tools as well as techniques which may be utilized in an effective manner for resolving the problems of predictions of student's performance as well as their profiling. The different current researches are evaluated and recommendations have been enlisted for future research.
- 17 January 2025
- Published by IGI Global Scientific Publishing
- p. 73-102
Empowering Education Through Data Mining Tools
In educational field due to recent advancements, digitization tools and worldwide epidemic scenarios, there is evolution of large number of online learning platforms as well as educational institutions starting from primary level to high education which is ultimately leading to the large amount of data and its handling that can be successfully achieved through Data Mining tools and techniques for extraction of relevant information. In this chapter it have been enlisted the ways utilized by researchers previously and currently for examining the way data mining was handled, moreover most updated trends for data mining in education field, as well as for evaluating the possibility for utilization of machine learning techniques in educational field. This work also explores the application of different Data Mining tools as well as techniques which may be utilized in an effective manner for resolving the problems of predictions of student's performance as well as their profiling. The different current researches are evaluated and recommendations have been enlisted for future research.
- Open Access
Accelerating Deep Learning-Based Morphological Biometric Recognition with Field-Programmable Gate Arrays
Convolutional neural networks (CNNs) are increasingly recognized as an important and potent artificial intelligence approach, widely employed in many computer vision applications, such as facial recognition. Their importance resides in their capacity to acquire hierarchical features, which is essential for recognizing complex patterns. Nevertheless, the intricate architectural design of CNNs leads to significant computing requirements. To tackle these issues, it is essential to construct a system based on field-programmable gate arrays (FPGAs) to speed up CNNs. FPGAs provide fast development capabilities, energy efficiency, decreased latency, and advanced reconfigurability. A facial recognition solution by leveraging deep learning and subsequently deploying it on an FPGA platform is suggested. The system detects whether a person has the necessary authorization to enter/access a place. The FPGA is responsible for processing this system with utmost security and without any internet connectivity. Various facial recognition networks are accomplished, including AlexNet, ResNet, and VGG-16 networks. The findings of the proposed method prove that the GoogLeNet network is the best fit due to its lower computational resource requirements, speed, and accuracy. The system was deployed on three hardware kits to appraise the performance of different programming approaches in terms of accuracy, latency, cost, and power consumption. The software programming on the Raspberry Pi-3B kit had a recognition accuracy of around 70–75% and relied on a stable internet connection for processing. This dependency on internet connectivity increases bandwidth consumption and fails to meet the required security criteria, contrary to ZYBO-Z7 board hardware programming. Nevertheless, the hardware/software co-design on the PYNQ-Z2 board achieved an accuracy rate of 85% to 87%. It operates independently of an internet connection, making it a standalone system and saving costs.
Open AccessAccelerating Deep Learning-Based Morphological Biometric Recognition with Field-Programmable Gate Arrays
Convolutional neural networks (CNNs) are increasingly recognized as an important and potent artificial intelligence approach, widely employed in many computer vision applications, such as facial recognition. Their importance resides in their capacity to acquire hierarchical features, which is essential for recognizing complex patterns. Nevertheless, the intricate architectural design of CNNs leads to significant computing requirements. To tackle these issues, it is essential to construct a system based on field-programmable gate arrays (FPGAs) to speed up CNNs. FPGAs provide fast development capabilities, energy efficiency, decreased latency, and advanced reconfigurability. A facial recognition solution by leveraging deep learning and subsequently deploying it on an FPGA platform is suggested. The system detects whether a person has the necessary authorization to enter/access a place. The FPGA is responsible for processing this system with utmost security and without any internet connectivity. Various facial recognition networks are accomplished, including AlexNet, ResNet, and VGG-16 networks. The findings of the proposed method prove that the GoogLeNet network is the best fit due to its lower computational resource requirements, speed, and accuracy. The system was deployed on three hardware kits to appraise the performance of different programming approaches in terms of accuracy, latency, cost, and power consumption. The software programming on the Raspberry Pi-3B kit had a recognition accuracy of around 70–75% and relied on a stable internet connection for processing. This dependency on internet connectivity increases bandwidth consumption and fails to meet the required security criteria, contrary to ZYBO-Z7 board hardware programming. Nevertheless, the hardware/software co-design on the PYNQ-Z2 board achieved an accuracy rate of 85% to 87%. It operates independently of an internet connection, making it a standalone system and saving costs.
- 29 November 2024
- Published by IGI Global Scientific Publishing
- p. 229-248
Navigating Remote Work Evolution Generational Perspectives on Digital Transformation
In 2020, the COVID-19 pandemic came out to be a massive destruction for human-being, in terms of life and economy. Lockdown in most of the countries invoked the teleworking as a solution to boost economy and social distancing for combating the highly transmissible virus, but teleworking can't be a full replacement. Traditional methods and working conditions are slowly replaced by digitization and by automating the processes of work. Moreover, work from home conditions have facilitated the houses as offices for the individuals in the world ultimately leading to digital transformation acceleration, replacing traditional work cultures. Information and communication technologies facilitated the digitization process. This chapter explores the journey of working conditions before pandemic, during pandemic, and possible scenario post pandemic after incorporating digitization has been done, involving also about mental/physical health effects on employees/students during pandemic and how the people and organizations collectively paced up the digitization for fulfilling the work assignments/projects.
- 29 November 2024
- Published by IGI Global Scientific Publishing
- p. 229-248
Navigating Remote Work Evolution Generational Perspectives on Digital Transformation
In 2020, the COVID-19 pandemic came out to be a massive destruction for human-being, in terms of life and economy. Lockdown in most of the countries invoked the teleworking as a solution to boost economy and social distancing for combating the highly transmissible virus, but teleworking can't be a full replacement. Traditional methods and working conditions are slowly replaced by digitization and by automating the processes of work. Moreover, work from home conditions have facilitated the houses as offices for the individuals in the world ultimately leading to digital transformation acceleration, replacing traditional work cultures. Information and communication technologies facilitated the digitization process. This chapter explores the journey of working conditions before pandemic, during pandemic, and possible scenario post pandemic after incorporating digitization has been done, involving also about mental/physical health effects on employees/students during pandemic and how the people and organizations collectively paced up the digitization for fulfilling the work assignments/projects.
- Open Access
- 13 August 2024
- Published by MDPI AG in Bioengineering
- Vol. 11 (8) , 827
Automated Multi-Class Facial Syndrome Classification Using Transfer Learning Techniques
Genetic disorders affect over 6% of the global population and pose substantial obstacles to healthcare systems. Early identification of these rare facial genetic disorders is essential for managing related medical complexities and health issues. Many people consider the existing screening techniques inadequate, often leading to a diagnosis several years after birth. This study evaluated the efficacy of deep learning-based classifier models for accurately recognizing dysmorphic characteristics using facial photos. This study proposes a multi-class facial syndrome classification framework that encompasses a unique combination of diseases not previously examined together. The study focused on distinguishing between individuals with four specific genetic disorders (Down syndrome, Noonan syndrome, Turner syndrome, and Williams syndrome) and healthy controls. We investigated how well fine-tuning a few well-known convolutional neural network (CNN)-based pre-trained models—including VGG16, ResNet-50, ResNet152, and VGG-Face—worked for the multi-class facial syndrome classification task. We obtained the most encouraging results by adjusting the VGG-Face model. The proposed fine-tuned VGG-Face model not only demonstrated the best performance in this study, but it also performed better than other state-of-the-art pre-trained CNN models for the multi-class facial syndrome classification task. The fine-tuned model achieved both accuracy and an F1-Score of 90%, indicating significant progress in accurately detecting the specified genetic disorders.
- 13 August 2024
- Published by MDPI AG in Bioengineering
- Vol. 11 (8) , 827
Open AccessAutomated Multi-Class Facial Syndrome Classification Using Transfer Learning Techniques
Genetic disorders affect over 6% of the global population and pose substantial obstacles to healthcare systems. Early identification of these rare facial genetic disorders is essential for managing related medical complexities and health issues. Many people consider the existing screening techniques inadequate, often leading to a diagnosis several years after birth. This study evaluated the efficacy of deep learning-based classifier models for accurately recognizing dysmorphic characteristics using facial photos. This study proposes a multi-class facial syndrome classification framework that encompasses a unique combination of diseases not previously examined together. The study focused on distinguishing between individuals with four specific genetic disorders (Down syndrome, Noonan syndrome, Turner syndrome, and Williams syndrome) and healthy controls. We investigated how well fine-tuning a few well-known convolutional neural network (CNN)-based pre-trained models—including VGG16, ResNet-50, ResNet152, and VGG-Face—worked for the multi-class facial syndrome classification task. We obtained the most encouraging results by adjusting the VGG-Face model. The proposed fine-tuned VGG-Face model not only demonstrated the best performance in this study, but it also performed better than other state-of-the-art pre-trained CNN models for the multi-class facial syndrome classification task. The fine-tuned model achieved both accuracy and an F1-Score of 90%, indicating significant progress in accurately detecting the specified genetic disorders.
- Open Access
- 1 January 2024
- Published by Tech Science Press in Computers, Materials & Continua
- Vol. 81 (1) , 271-307
Enhancing Early Detection of Lung Cancer through Advanced Image Processing Techniques and Deep Learning Architectures for CT Scans
Lung cancer remains a major concern in modern oncology due to its high mortality rates and multifaceted origins, including hereditary factors and various clinical changes. It stands as the deadliest type of cancer and a significant cause of cancer-related deaths globally. Early diagnosis enables healthcare providers to administer appropriate treatment measures promptly and accurately, leading to improved prognosis and higher survival rates. The significant increase in both the incidence and mortality rates of lung cancer, particularly its ranking as the second most prevalent cancer among women worldwide, underscores the need for comprehensive research into efficient screening methods. Advances in diagnostic techniques, particularly the use of computed tomography (CT) scans, have revolutionized the identification of lung cancer. CT scans are renowned for their ability to provide high-resolution images and are particularly effective in detecting small, calcified areas, crucial for identifying early-stage lung cancer. Consequently, there is growing interest in enhancing computer-aided detection (CAD) systems. These algorithms assist radiologists by reducing false-positive interpretations and improving the accuracy of early cancer diagnosis. This study aims to enhance the effectiveness of CAD systems through various methods. Initially, the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm is employed to preprocess CT scan images, thereby improving their visual quality. Further refinement is achieved by integrating different optimization strategies with the CLAHE method. The CutMix data augmentation technique is applied to boost the performance of the proposed model. A comparative analysis is conducted using deep learning architectures such as Inception-V3, ResNet101, Xception, and EfficientNet. The study evaluates the performance of these architectures in image classification tasks, both with and without the implementation of the CLAHE algorithm. The empirical findings of the study demonstrate a significant reduction in the false positive rate (FPR) and an overall enhancement in diagnostic accuracy. This research not only contributes to the field of medical imaging but also holds significant implications for the early detection and treatment of lung cancer, ultimately aiming to reduce its mortality rates.
- 1 January 2024
- Published by Tech Science Press in Computers, Materials & Continua
- Vol. 81 (1) , 271-307
Open AccessEnhancing Early Detection of Lung Cancer through Advanced Image Processing Techniques and Deep Learning Architectures for CT Scans
Lung cancer remains a major concern in modern oncology due to its high mortality rates and multifaceted origins, including hereditary factors and various clinical changes. It stands as the deadliest type of cancer and a significant cause of cancer-related deaths globally. Early diagnosis enables healthcare providers to administer appropriate treatment measures promptly and accurately, leading to improved prognosis and higher survival rates. The significant increase in both the incidence and mortality rates of lung cancer, particularly its ranking as the second most prevalent cancer among women worldwide, underscores the need for comprehensive research into efficient screening methods. Advances in diagnostic techniques, particularly the use of computed tomography (CT) scans, have revolutionized the identification of lung cancer. CT scans are renowned for their ability to provide high-resolution images and are particularly effective in detecting small, calcified areas, crucial for identifying early-stage lung cancer. Consequently, there is growing interest in enhancing computer-aided detection (CAD) systems. These algorithms assist radiologists by reducing false-positive interpretations and improving the accuracy of early cancer diagnosis. This study aims to enhance the effectiveness of CAD systems through various methods. Initially, the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm is employed to preprocess CT scan images, thereby improving their visual quality. Further refinement is achieved by integrating different optimization strategies with the CLAHE method. The CutMix data augmentation technique is applied to boost the performance of the proposed model. A comparative analysis is conducted using deep learning architectures such as Inception-V3, ResNet101, Xception, and EfficientNet. The study evaluates the performance of these architectures in image classification tasks, both with and without the implementation of the CLAHE algorithm. The empirical findings of the study demonstrate a significant reduction in the false positive rate (FPR) and an overall enhancement in diagnostic accuracy. This research not only contributes to the field of medical imaging but also holds significant implications for the early detection and treatment of lung cancer, ultimately aiming to reduce its mortality rates.
- 29 June 2022
- Published by Springer Nature in Journal of Digital Imaging
- Vol. 35 (5) , 1308-1325
Multimodal Medical Image Fusion Using Stacked Auto-encoder in NSCT Domain
Medical image fusion is a process that aims to merge the important information from images with different modalities of the same organ of the human body to create a more informative fused image. In recent years, deep learning (DL) methods have achieved significant breakthroughs in the field of image fusion because of their great efficiency. The DL methods in image fusion have become an active topic due to their high feature extraction and data representation ability. In this work, stacked sparse auto-encoder (SSAE), a general category of deep neural networks, is exploited in medical image fusion. The SSAE is an efficient technique for unsupervised feature extraction. It has high capability of complex data representation. The proposed fusion method is carried as follows. Firstly, the source images are decomposed into low- and high-frequency coefficient sub-bands with the non-subsampled contourlet transform (NSCT). The NSCT is a flexible multi-scale decomposition technique, and it is superior to traditional decomposition techniques in several aspects. After that, the SSAE is implemented for feature extraction to obtain a sparse and deep representation from high-frequency coefficients. Then, the spatial frequencies are computed for the obtained features to be used for high-frequency coefficient fusion. After that, a maximum-based fusion rule is applied to fuse the low-frequency sub-band coefficients. The final integrated image is acquired by applying the inverse NSCT. The proposed method has been applied and assessed on various groups of medical image modalities. Experimental results prove that the proposed method could effectively merge the multimodal medical images, while preserving the detail information, perfectly.
- 29 June 2022
- Published by Springer Nature in Journal of Digital Imaging
- Vol. 35 (5) , 1308-1325
Multimodal Medical Image Fusion Using Stacked Auto-encoder in NSCT Domain
Medical image fusion is a process that aims to merge the important information from images with different modalities of the same organ of the human body to create a more informative fused image. In recent years, deep learning (DL) methods have achieved significant breakthroughs in the field of image fusion because of their great efficiency. The DL methods in image fusion have become an active topic due to their high feature extraction and data representation ability. In this work, stacked sparse auto-encoder (SSAE), a general category of deep neural networks, is exploited in medical image fusion. The SSAE is an efficient technique for unsupervised feature extraction. It has high capability of complex data representation. The proposed fusion method is carried as follows. Firstly, the source images are decomposed into low- and high-frequency coefficient sub-bands with the non-subsampled contourlet transform (NSCT). The NSCT is a flexible multi-scale decomposition technique, and it is superior to traditional decomposition techniques in several aspects. After that, the SSAE is implemented for feature extraction to obtain a sparse and deep representation from high-frequency coefficients. Then, the spatial frequencies are computed for the obtained features to be used for high-frequency coefficient fusion. After that, a maximum-based fusion rule is applied to fuse the low-frequency sub-band coefficients. The final integrated image is acquired by applying the inverse NSCT. The proposed method has been applied and assessed on various groups of medical image modalities. Experimental results prove that the proposed method could effectively merge the multimodal medical images, while preserving the detail information, perfectly.
- 27 April 2021
- Published by Springer Nature in Journal of Ambient Intelligence and Humanized Computing
- Vol. 12 (6) , 6001-6018
Hybrid pixel-feature fusion system for multimodal medical images
- 27 April 2021
- Published by Springer Nature in Journal of Ambient Intelligence and Humanized Computing
- Vol. 12 (6) , 6001-6018
Hybrid pixel-feature fusion system for multimodal medical images
- 15 October 2020
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 80 (4) , 6369-6396
Survey study of multimodality medical image fusion methods
- 15 October 2020
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 80 (4) , 6369-6396
Survey study of multimodality medical image fusion methods