Article
Engineering, Biomedical
Sareer Ul Amin, Sher Taj, Adnan Hussain, Sanghyun Seo
Summary: An ensemble-based machine learning approach has been developed to automate the classification of COVID-19, TB, and pneumonia using CXR images. Experimental results show that the proposed method achieved a classification accuracy of 98% using the CXR image dataset.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2024)
Article
Oncology
Ahmed Shaffie, Ahmed Soliman, Amr Eledkawy, Victor van Berkel, Ayman El-Baz
Summary: Lung cancer is the second most common cancer among men and the third most common cancer among women, but it is the leading cause of cancer-related deaths. This manuscript proposes a new computer-aided diagnosis system that uses chest CT scans to accurately diagnose the malignancy of pulmonary nodules. The system extracts appearance and shape features and applies deep learning algorithms for diagnosis. The validation results show high accuracy, sensitivity, and specificity, indicating the system's ability to accurately distinguish between malignant and benign nodules.
Article
Computer Science, Artificial Intelligence
Amenah Nazar Jabbar, Hakan Koyuncu
Summary: This study introduces a novel methodology utilizing deep learning and Grey Wolf Optimization for plant disease detection, achieving high classification accuracies through multiple feature extractors and optimal feature selection. The research is of significant importance for improving agricultural health management.
TRAITEMENT DU SIGNAL
(2023)
Article
Computer Science, Software Engineering
Wadhah Ayadi, Imen Charfi, Wajdi Elhamzi, Mohamed Atri
Summary: A new technique is proposed to improve the quality of MRI and classify brain tumors, achieving an accuracy of 90.27% and surpassing previous methods.
Article
Oncology
Mehran Ahmad, Muhammad Abeer Irfan, Umar Sadique, Ihtisham ul Haq, Atif Jan, Muhammad Irfan Khattak, Yazeed Yasin Ghadi, Hanan Aljuaid
Summary: This research successfully improved the accuracy of early diagnosis of Oral Squamous Cell Carcinoma (OSCC) through the development of hybrid methodologies. By employing a combination of transfer learning, CNN, and SVM, as well as the fusion of deep and texture-based features, the study achieved a significant overall accuracy.
Article
Clinical Neurology
Katharina Althaus, Martin Kasel, Albert C. Ludolph, Jan Kassubek, Rebecca Kassubek
Summary: This study retrospectively analyzed 23,948 cerebral MRIs and found that 84 images from 61 patients showed HARM-positive results. HARM was observed not only in stroke patients, but also in other neurological conditions, and it was not dependent on therapy.
INTERNATIONAL JOURNAL OF STROKE
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Michael Eliezer, Alexis Vaussy, Solenn Toupin, Remy Barbe, Stephan Kannengiesser, Alto Stemmer, Emmanuel Houdart
Summary: The purpose of this study was to evaluate the image quality of a three-dimensional fluid attenuated inversion recovery (3D-FLAIR) sequence acquired with a high acceleration factor and reconstructed with iterative denoising (ID) for brain MRI at 3-T. The results showed that using this method can reduce the scanning time by 37% while preserving image quality.
DIAGNOSTIC AND INTERVENTIONAL IMAGING
(2022)
Article
Clinical Neurology
Samantha Noteboom, D. R. van Nederpelt, A. Bajrami, B. Moraal, M. W. A. Caan, F. Barkhof, M. Calabrese, H. Vrenken, E. M. M. Strijbis, M. D. Steenwijk, M. M. Schoonheim
Summary: This study found that brain volumes derived from 3D-FLAIR images have similar associations with disability and cognitive impairment in patients with MS compared to those derived from 3D-T1 images. This highlights the potential of these techniques in clinical datasets.
JOURNAL OF NEUROLOGY
(2023)
Article
Engineering, Biomedical
Ela Kaplan, Mehmet Baygin, Prabal D. Barua, Sengul Dogan, Turker Tuncer, Erman Altunisik, Elizabeth Emma Palmer, U. Rajendra Acharya
Summary: The purpose of this study is to classify the neuroradiological features of patients with Alzheimer's disease (AD) using an automatic hand-modeled method with high accuracy. The proposed model, ExHiF, uses feature extraction, feature selection, and multiple classifiers to achieve the classification. The results show that the ExHiF model achieves 100% classification accuracy for AD patients using two datasets.
MEDICAL ENGINEERING & PHYSICS
(2023)
Article
Computer Science, Information Systems
Aasim Zafar, Arshad Iqbal
Summary: Research in the field of offline Arabic handwriting recognition has grown exponentially in the past decades. This paper presents a system based on word segmentation for recognizing Arabic Kufic script using Histogram of Oriented Gradient (HOG) and Local Binary Pattern (LBP) feature extraction techniques. Experimental results show that the proposed system performs better than previous recognition systems in recognizing Kufic script.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Radiology, Nuclear Medicine & Medical Imaging
Benjamin A. Hoff, Benjamin Lemasson, Thomas L. Chenevert, Gary D. Luker, Christina I. Tsien, Ghoncheh Amouzandeh, Timothy D. Johnson, Brian D. Ross
Summary: This study suggests that PRMrFLAIR may serve as an early biomarker for disease progression in glioblastoma. Changes in PRMrFLAIR+ could potentially predict patient outcomes and stratify them for progression-free survival. The study demonstrates the potential of utilizing voxel-wise comparison of FLAIR signal for improved treatment monitoring in glioblastoma.
ACADEMIC RADIOLOGY
(2021)
Article
Computer Science, Information Systems
Mona Ghahremani, Hamid Ghadiri, Mohammad Hamghalam
Summary: A new content-based image retrieval (CBIR) model is proposed in this study for reconstructing corrupted portions of images by combining color, texture, and shape features effectively. The process involves normalizing image scans, reducing noise with a median filter, modifying color channel shift with SLIC superpixel, introducing Histogram of Oriented Gradients (HOG) descriptor for feature extraction, and performing local thresholding based on Local Binary Patterns (LBP) to enhance image details. Experimental results show the method's efficiency in content retrieval, with a highest rate of 90.54% on a liver CT scan image compared to previous approaches.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Neuroimaging
Dejan Jakimovski, Robert Zivadinov, Niels Bergsland, Deepa P. Ramasamy, Jesper Hagemeier, Antonia Valentina Genovese, David Hojnacki, Bianca Weinstock-Guttman, Michael G. Dwyer
Summary: In this study, the researchers investigated the feasibility of atrophy measures as markers of disability progression (DP) in patients with multiple sclerosis (PwMS) scanned across different MRI field strengths. They found that changes in lateral ventricular volume (LVV) were significantly associated with DP, and that NeuroSTREAM and VIENA-based measure showed greater atrophy in PwMS scanned on different scanners. On the other hand, PBVC and SIENAX-based vCSF % changes were significantly affected by scanner changes.
NEUROIMAGE-CLINICAL
(2021)
Article
Medicine, Legal
Celine Berger, Christoph Birkl, Melanie Bauer, Eva Scheurer, Claudia Lenz
Summary: This study aims to determine the temperature effect on post mortem TInull in FLAIR sequence and provide a temperature correction method for achieving robust suppression of the CSF signal. Based on in situ MRI brain examination of nine deceased subjects, a significant positive linear relation was found between TInull and both brain and rectal temperatures, enabling the correction of TInull for varying temperatures of the deceased.
FORENSIC SCIENCE INTERNATIONAL
(2022)
Article
Computer Science, Information Systems
V Manimaran, K. G. Srinivasagan, S. Gokul, I. Jeena Jacob, S. Baburenagarajan
Summary: This paper introduces a unique framework called ILFP that integrates local and global features, as well as a new deep learning architecture m-XceptionNet that effectively extracts global features with lower complexity. The proposed framework outperforms existing works in Rank1 metric and mean Average Precision on different datasets, demonstrating its effectiveness.
KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Muhammad Zahid, Muhammad Attique Khan, Faisal Azam, Muhammad Sharif, Seifedine Kadry, Jnyana Ranjan Mohanty
Summary: In this article, a real-time architecture for pedestrian identification using motion-controlled deep neural networks (DNN) is presented. Motion vectors are calculated using optical flow and utilized for feature extraction. The features from both HOG and DNN are fused and a support vector machine classifier is used for final identification.
Article
Computer Science, Artificial Intelligence
Javeria Naz, Muhammad Sharif, Mudassar Raza, Jamal Hussain Shah, Mussarat Yasmin, Seifedine Kadry, S. Vimal
Summary: This paper introduces a hybrid approach based on texture and deep learning features for computer-aided diagnosis of stomach diseases. The method extracts texture features and deep convolutional neural network features, which are then serially fused to obtain a strong feature vector, thereby improving diagnostic accuracy.
NEURAL PROCESSING LETTERS
(2023)
Article
Computer Science, Artificial Intelligence
Muhammad Junaid Umer, Muhammad Sharif, Mudassar Raza, Seifedine Kadry
Summary: Optical coherence tomography (OCT) is a valuable imaging modality for detecting and classifying retinal eye diseases. This research proposes an automatic method for retinal eye disease detection and classification from OCT images, achieving high accuracy exceeding the current state-of-the-art performance. By utilizing modified-Alexnet and ResNet-50 for deep feature extraction and a feature selection framework, the proposed method demonstrates reliable and efficient automatic eye disease detection.
Article
Chemistry, Multidisciplinary
Rehna Batool, Nargis Bibi, Nazeer Muhammad, Samah Alhazmi
Summary: Cognitive Radio Network (CRN) is emerging technology to solve spectrum shortage problems. However, the spectrum sensing process in CRN is often disturbed by a security issue known as the Primary User Emulation Attack (PUEA). This study proposes a TDOA-based localization method using the DE algorithm to identify PUEA in CRNs.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Javaria Amin, Muhammad Almas Anjum, Muhammad Sharif, Seifedine Kadry, Ruben Gonzalez Crespo
Summary: Liver cancer is a major cause of death worldwide, and manually detecting infected tissues is challenging and time-consuming. Computerized methods can assist in making accurate decisions and therapies. Semantic segmentation plays a vital role in segmenting infected liver regions.
IEEE LATIN AMERICA TRANSACTIONS
(2023)
Article
Computer Science, Artificial Intelligence
Abdul Muiz Fayyaz, Mudassar Raza, Muhammad Sharif, Jamal Hussain Shah, Seifedine Kadry, Oscar Sanjuan Martinez
Summary: COVID-19 is a challenging pandemic that spreads rapidly worldwide, and there is a need to develop an automated technique for its identification. This study proposes a new framework for predicting COVID-19 based on X-ray images, which includes core phases such as preprocessing, feature extraction, selection, and categorization.
INTERNATIONAL JOURNAL OF UNCERTAINTY FUZZINESS AND KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Ahmed Iqbal, Muhammad Sharif
Summary: Breast cancer, the most commonly diagnosed cancer globally, poses challenges in accurate tumor segmentation due to random variation, irregular shapes, and blurred boundaries. To address this issue, we propose the BTS-ST network, which integrates Swin-Transformer and U-Net to enhance global modeling capability, achieving superior results in breast tumor segmentation and classification compared to other methods.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Muhammad Fayyaz, Mussarat Yasmin, Muhammad Sharif, Tasswar Iqbal, Mudassar Raza, Muhammad Imran Babar
Summary: In this manuscript, the imbalanced and small sample space dataset problems for pedestrian gender classification are addressed through the fusion of selected deep and traditional features. Various feature extraction and selection methods are employed, and multiple classifiers are used to perform gender classification. The proposed method achieves better results in terms of accuracy and area under curve on different datasets.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Agronomy
Javeria Amin, Muhammad Almas Anjum, Rida Zahra, Muhammad Imran Sharif, Seifedine Kadry, Lukas Sevcik
Summary: A method is developed for the localization and classification of pests in agriculture. YOLOv5 is used for localization and a quantum machine learning model is proposed for classification, achieving 0.93 F1 scores for localization and 99.9% classification accuracy. The developed model is compared to existing methods to demonstrate its novelty.
Article
Computer Science, Artificial Intelligence
Ahmed Iqbal, Muhammad Sharif
Summary: Rapid and precise segmentation of breast tumors is a challenge for diagnosing breast cancer in younger females. This research proposes a semi-supervised learning-based method, which incorporates DEN, PMG, and PDF-UNet network for accurate breast tumor segmentation. The results demonstrate that the proposed PDF-UNet achieves higher DSC on the Mendeley and SIIT datasets compared with classical UNet.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Javaria Amin, Muhammad Almas Anjum, Kainat Ibrar, Muhammad Sharif, Seifedine Kadry, Ruben Gonzalez Crespo
Summary: This research proposes two models, J. DCNN and J. QCNN, for identifying anomalous behavior in video surveillance. These models efficiently analyze and detect various types of abnormal behavior, such as violent robberies. The experimental results show that the J. QCNN model achieves an accuracy of 0.99, outperforming existing cutting-edge methods.
IMAGE AND VISION COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Javaria Amin, Muhammad Almas Anjum, Nadia Gul, Muhammad Sharif
Summary: The brain is a complex organ that can be affected by abnormal brain cells, such as tumors, abscesses, and cysts. The noisy nature of brain MRI images reduces detection accuracy, thus a 32-layers-denoise neural network is proposed to enhance image quality. A novel seven layers Javeria Quanvolutional Neural Network model named J. Qnet is introduced for classifying healthy/abnormal MRI slices, and an ONNX-YOLOv2tiny model is proposed to localize the classified images. A 34-layer U-net model is proposed for more accurate segmentation of the localized images. The proposed method achieves high accuracy in locally acquired images and the BRATS-2020 dataset, outperforming existing research works.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Sara Khalid, Jamal Hussain Shah, Muhammad Sharif, Muhammad Rafiq, Gyu Sang Choi
Summary: Traffic signs are crucial for facilitating traffic flow and ensuring the safety of drivers and pedestrians worldwide. Due to challenges such as low lighting conditions, occlusion, and similarity with other objects, detecting traffic signs automatically is a challenging task. Therefore, an innovative method utilizing color image processing techniques and features fusion is proposed for efficient and accurate traffic sign detection.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Information Systems
Faheem Shehzad, Muhammad Attique Khan, Muhammad Asfand E. Yar, Muhammad Sharif, Majed Alhaisoni, Usman Tariq, Arnab Majumdar, Orawit Thinnukool
Summary: Human action recognition based on Artificial intelligence reasoning is the most important research area in computer vision. This work proposes a deep learning and improved whale optimization algorithm based framework for HAR, which includes stages like pre-processing, transfer learning, feature fusion, and feature selection using modified serial approach and improved whale optimization algorithm. The proposed method achieves high testing accuracy on four datasets and outperforms state-of-the-art techniques.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Interdisciplinary Applications
Zainab Akhtar, Jong Weon Lee, Muhammad Attique Khan, Muhammad Sharif, Sajid Ali Khan, Naveed Riaz
Summary: This paper presents an automated OCR technique based on multi-properties features fusion and selection. The features are fused using serial formulation and selected using partial least square method with an entropy fitness function. The method achieves high accuracy in testing and outperforms existing techniques.
JOURNAL OF ENTERPRISE INFORMATION MANAGEMENT
(2023)
Article
Computer Science, Artificial Intelligence
Ursula Addison
Summary: Goal reasoning is an essential functionality for artificial systems to manage and execute goals in complex and changing environments. This survey investigates motivated agents that simulate human integrated-self, exploring the potential benefits of internal motivations in goal reasoning. By evaluating different systems, the study concludes the potential advantages of motives, mental simulation, and emotion in the goal reasoning paradigm.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Haoran Fu, Chundong Wang, Jiaqi Sun, Yumeng Zhao, Hao Lin, Junqing Sun, Baixue Zhang
Summary: Although natural language processing has shown strong performance, it is vulnerable to adversarial examples. Current methods for English are not suitable for Chinese due to the differences in language structure. This paper proposes a new algorithm called WordIllusion for generating Chinese adversarial texts and verifies its effectiveness.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Beatriz Garcia-Martinez, Patricia Fernandez-Sotos, Jorge J. Ricarte, Eva M. Sanchez-Morla, Roberto Sanchez-Reolid, Roberto Rodriguez-Jimenez, Antonio Fernandez-Caballero
Summary: This study aims to detect auditory hallucinations (AH) in schizophrenia patients using a wireless EEG device. The results show that AH is mainly activated in the right frontal locations, while the left hemisphere demonstrates stronger activation during hallucination-free periods. Additionally, a decrease in spectral power during hallucination episodes compared to non-hallucination periods is observed.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Shivant Kathusing, Natalie Samhan, Jan Treur
Summary: This paper introduces a fifth-order adaptive self-modelling network model to describe the role of epigenetics in the development of anxiety disorders and suggests a possible epigenetics-based therapeutic method.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Yanru Jiang, Rick Dale, Hongjing Lu
Summary: This study investigates the integration of large-language models and recurrent neural networks into a hybrid cognitive model for solving natural language tasks. The findings highlight the crucial role of global knowledge in adapting to new learning tasks, while having only local knowledge significantly reduces system transferability.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
configuration Viacheslav Wolfengagen, Larisa Ismailova, Sergey Kosikov, Igor Slieptsov, Sebastian Dohrn, Alexander Marenkov, Vladislav Zaytsev
Summary: This paper proposes a configuration-based approach to knowledge extraction, which enhances the expressive power of semantic networks. By representing functions as objects, the central issues of nodes and links in knowledge-based systems are addressed. The model is applicable to representing morphing and considers objects as processes, aligning with current ideas in computing. The concept of information channels for process transformations is introduced. The potential for generating displaced concepts and their morph families is demonstrated.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Leibovici Anat, Raizman Reut, Itzhaki Nofar, Tik Niv, Sapir Maayan, Tsarfaty Galia, Livny Abigail
Summary: Traditionally, neuroimaging studies have focused on brain activation in frontal-parietal regions for fluid intelligence. However, recent evidence suggests the involvement of the cerebellum in higher cognitive function. This study investigates the role of the cerebellum in processing fluid intelligence and provides evidence through task brain activation and network analysis.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Leticia Berto, Leonardo Rossi, Eric Rohmer, Paula Costa, Ricardo Gudwin, Alexandre Simoes, Esther Colombini
Summary: Integrating robots into daily life is becoming a reality, and bridging the gap between human developmental theories and robotics applications is crucial. This research focuses on the early stages of human development from 0 to 2 years old and aims to simulate motor and cognitive growth in robots through progressive experiments.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Alice Plebe, Henrik Svensson, Sara Mahmoud, Mauro Da Lio
Summary: This article reviews the research on autonomous driving influenced by cognitive science, neuroscience, and psychology, proposing the potential advantages of human driving ability for developing autonomous vehicles and discussing the methods of applying human thinking process to autonomous driving.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Md Ehtesham-Ul-Haque, Jacob D'Rozario, Rudaiba Adnin, Farhan Tanvir Utshaw, Fabiha Tasneem, Israt Jahan Shefa, A. B. M. Alim Al Islam
Summary: This paper aims to explore emotion generation, particularly for general-purpose conversations. Based on the Cognitive Appraisal Theory, a novel method to calculate informative variables for evaluating emotion-generating events and six primary emotions is proposed. The implementation of EmoBot, an emotional chatbot, demonstrates its ability to generate more accurate emotional and semantic responses compared to traditional chatbots.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Simon Jerome Han, Keith J. Ransom, Andrew Perfors, Charles Kemp
Summary: This study applies GPT-3.5 and GPT-4 models to explore the differences between human and machine intelligence in the context of property induction. The results show that GPT-4 performs qualitatively similar to humans in most cases, with the exception of premise non-monotonicity. This research provides interesting comparisons and two large datasets for future studies.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Vincent Frey, Julian Martinez
Summary: This study proposes a trust dynamics model based on a multi-agent reinforcement learning algorithm, aiming to quantitatively understand the characteristics and behavior of interpersonal trust, and explore the relationship between trust and agent performance.
COGNITIVE SYSTEMS RESEARCH
(2024)
Article
Computer Science, Artificial Intelligence
Hamit Basgol, Inci Ayhan, Emre Ugur
Summary: This study investigates the mechanism behind event segmentation, the process of dividing experiences into discrete units. The researchers propose a computational model inspired by event segmentation theory and predictive processing, which successfully produces human-like event boundaries and representations.
COGNITIVE SYSTEMS RESEARCH
(2024)