Search

In today’s era of increasing data complexity and pervasive noise, robust techniques for data processing, reconstruction, and denoising are crucial. Autoencoders, known for their adaptability in unsupervised learning, offer a strategic solution to these challenges. This research introduces two novel autoencoder architectures, ResoFocus and FragmentumZoom, complemented by a refined loss function. The ResoFocus Autoencoder is designed to optimize and denoise upsized images, while the FragmentumZoom Autoencoder targets small image segments to enhance denoising tasks. Empirical evaluations on diverse datasets, including CelebA and Autism Face data, demonstrate significant improvements in the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM). For the CelebA dataset, the ResoFocus 4X autoencoder achieved PSNR values of 31.578, with a peak increase of 1.9 dB compared to the baseline Simple autoencoder and SSIM values of 0.898 at 1000 epochs. In the Autism Face dataset, the FragmentumZoom 4X autoencoder recorded PSNR values of 30.951, with an increase of 1.289 dB over the baseline, and SSIM values of 0.906. These results underscore the efficacy of these novel architectures in improving image quality and advancing image processing techniques, particularly in scenarios requiring heightened resolution and noise reduction.

COVID-19 has affected many people across the globe. Though vaccines are available now, early detection of the disease plays a vital role in the better management of COVID-19 patients. An Artificial Neural Network (ANN) powered Computer Aided Diagnosis (CAD) system can automate the detection pipeline accounting for accurate diagnosis, overcoming the limitations of manual methods. This work proposes a CAD system for COVID-19 that detects and classifies abnormalities in lung CT images using Artificial Bee Colony (ABC) optimised ANN (ABCNN). The proposed ABCNN approach works by segmenting the suspicious regions from the CT images of non-COVID and COVID patients using an ABC optimised region growing process and extracting the texture and intensity features from those suspicious regions. Further, an optimised ANN model whose input features, initial weights and hidden nodes are optimised using ABC optimisation classifies those abnormal regions into COVID and non-COVID classes. The proposed ABCNN approach is evaluated using the lung CT images collected from the public datasets. In comparison to other available techniques, the proposed ABCNN approach achieved a high classification accuracy of 92.37% when evaluated using a set of 470 lung CT images.

Researchers gravitate towards Generative Adversarial Networks (GAN) to create artificial images. However, GANs suffer from convergence issues, mode collapse, and overall complexity in balancing the Nash Equilibrium. Images generated are often distorted, rendering them useless. We propose a combination of Variational Autoencoders (VAEs) and a statistical oversampling method called K-Nearest Neighbor OveRsampling (KNNOR) to create artificial images. This combination of VAE and KNNOR results in more life-like images with reduced distortion. We fine-tune several pre-trained networks on a separate set of real and fake face images to test images generated by our method against images generated by conventional Deep Convolutional GANs (DCGANs). We also compare the combination of VAEs and Synthetic Minority Oversampling Technique (SMOTE) to establish the efficacy of KNNOR against naive oversampling methods. Not only are our methods better able to convince the classifiers that the images generated are authentic, but the models are also half in size of DCGANs. The code is available at GitHub for public use.

Currently, numerous types of cybercrime are organized through the internet. Hence, this study mainly focuses on phishing attacks. Although phishing was first used in 1996, it has become the most severe and dangerous cybercrime on the internet. Phishing utilizes email distortion as its underlying mechanism for tricky correspondences, followed by mock sites, to obtain the required data from people in question. Different studies have presented their work on the precaution, identification, and knowledge of phishing attacks; however, there is currently no complete and proper solution for frustrating them. Therefore, machine learning plays a vital role in defending against cybercrimes involving phishing attacks. The proposed study is based on the phishing URL-based dataset extracted from the famous dataset repository, which consists of phishing and legitimate URL attributes collected from 11000+ website datasets in vector form. After preprocessing, many machine learning algorithms have been applied and designed to prevent phishing URLs and provide protection to the user. This study uses machine learning models such as decision tree (DT), linear regression (LR), random forest (RF), naive Bayes (NB), gradient boosting classifier (GBM), K-neighbors classifier (KNN), support vector classifier (SVC), and proposed hybrid LSD model, which is a combination of logistic regression, support vector machine, and decision tree (LR+SVC+DT) with soft and hard voting, to defend against phishing attacks with high accuracy and efficiency. The canopy feature selection technique with cross fold valoidation and Grid Search Hyperparameter Optimization techniques are used with proposed LSD model. Furthermore, to evaluate the proposed approach, different evaluation parameters were adopted, such as the precision, accuracy, recall, F1-score, and specificity, to illustrate the effects and efficiency of the models. The results of the comparative analyses demonstrate that the proposed approach outperforms the other models and achieves the best results.

Various vision-based tasks suffer from inaccurate navigation and poor performance due to inevitable problems, such as adverse weather conditions like haze, fog, rain, snow, and clouds affecting ground and aerial navigation, as well as underwater images being degraded with blue-green tones and mud affecting marine navigation. Existing techniques in the literature typically focus on restoring specific degradations using separate models, leading to computational inefficiency. To address this, an all-in-one Multidomain Attention-based Conditional Generative Adversarial Network (MACGAN) is proposed to improve scene visibility for optimal ground, aerial, and marine navigation, using the same set of parameters across all domains. The MACGAN is a lightweight network with four encoder and decoder blocks and multiple attention blocks in between, which enhances the image restoration process by focusing on the most important features. To evaluate the effectiveness of MACGAN, extensive qualitative and quantitative comparisons are conducted with state-of-the-art image-to-image translation models, all-in-one adverse weather removal models, and single-effect removal models. The results highlight the superior performance of MACGAN in terms of scene visibility improvement and restoration quality. Additionally, MACGAN is tested on real-world unseen image domains, including smog, dust, fog, rain, snow, and lightning, further validating its generalizability and robustness. Furthermore, an ablation study is conducted to analyze the contributions of the discriminator and attention blocks within the MACGAN architecture. The results confirm that both components play significant roles in the effectiveness of MACGAN, with the discriminator ensuring adversarial training and the attention blocks effectively capturing and enhancing important image features.

In the midst of the ongoing COVID-19 pandemic, there has been a surge in scientific literature aimed at understanding the virus and its impact. However, it has become challenging for a researcher to deal with thousands of articles published daily. This paper proposes a novel deep-learning architecture to organize a large dataset of COVID-19-related scientific literature and provides a clear overview of the current state of knowledge. The proposed model is developed based on two main bases to ensure robustness and efficiency. In particular, we trained a denoising autoencoder with clean and noisy data to make the model can balance, preserving the underline structure and generalizing the new unseen data. Furthermore, the cornerstone of the proposed architecture lies in training the autoencoder using a two-fold objective function that jointly incorporates the data’s reconstruction and clustering. The advantage behind this combination is to avoid the distortion of the latent space and to improve the model efficiency. Afterward, we use the Latent Dirichlet Allocation (LDA) to analyze the document’s topics. For the sake of computational efficiency, instead of feeding the LDA with the whole dataset of documents, we fed it with the clusters produced in the phase of dimensionality reduction and clustering to count the frequency of topics in each cluster. The model was trained on a large public corpus of COVID-19-related articles and evaluated using a set of evaluation metrics. Experimental results indicate the superiority of our proposed model compared to several recent studies.

EEG is a common and safe test that uses small electrodes to record electrical signals from the brain. It has a broad range of applications in medical diagnosis, including diagnosis of epileptic seizure, Alzheimer’s, brain tumors, head injury, sleep disorders, stroke, and other seizure and neurological disorders. EEG can also be used to help diagnose death in people who are in a persistent coma. The use of digital signal processing and machine learning to improve EEG analysis for medical diagnosis has gained traction in recent years. This is because EEG visual analysis can be complex and time-consuming, as it mostly involves high dimensions and consists of large datasets. The development of novel sensors for EEG recording, digital signal processing algorithms, feature engineering, and detection algorithms increases the need for efficient diagnostic systems. An extensive review of the recent approaches for EEG preprocessing, extraction of features, and diagnosis of brain disorders is provided. In this paper, the main focus is to identify reliable algorithms for preprocessing, feature engineering, and classification of EEG, applied to medical healthcare and diagnosis, providing practitioners with insights into the most effective strategies, as well as potential future directions for improving accuracy of the automatic diagnostic systems. The study of reliable feature extraction and classification algorithms is crucial for a more accurate analysis of EEG signals. This paper can provide valuable information to researchers and practitioners working in the fields of EEG analysis and machine learning, as it provides a summary of recent developments and highlights key areas for future research. This paper can help researchers and clinicians to stay up-to-date on the latest developments in this field.

Nature-inspired metaheuristics have been extensively investigated to solve challenging optimization problems. Particle Swarm Optimization (PSO) is one of the most famous nature-inspired algorithms owing to its simplicity and ability to be used in a wide range of applications. This paper presents an extended PSO variant, namely, Exponential Particle Swarm Optimization (ExPSO). To effectively explore the whole search space, the proposed algorithm divides the swarm population into three equal subpopulations and employs a new search strategy based on an exponential function (permitting particles to make leaps in the search space) and an adapted control of the velocity range of each particle (to balance the exploration and exploitation search phases). The leaping strategy is integrated into the velocity equation and a new linear decreasing cognitive parameter (including a dynamic inertia weight strategy) is integrated into the proposed method. The developed algorithm allows large jumps at the beginning of the search, and then small jumps for further improvements in specific regions of the solution search space. Our variant approach, ExPSO, has been intensively tested through a comparison with eight other well-known heuristic search algorithms, over 29 benchmark problems, and real optimization engineering problems. The Wilcoxon signed-rank test and Friedman rank have been applied to analyze the search performance of the algorithms. The comparisons and statistical results show that the exponential search strategy significantly contributes to the search process and proves the superiority of ExPSO in terms of the convergence velocity and optimization accuracy.

Understanding protein subcellular localization is vital and indispensable in proteomics research. Molecular biology and computer science developments have enabled the use of computational approaches to identify proteins in cells. An excellent method for locating proteins is confocal microscopy, used by the Human Protein Atlas (HPA). By categorizing human proteins, it can assist researchers in better comprehending human pathophysiology and assist doctors in automating medical image interpretation. Human protein Atlas comprises millions of images annotated with single or multiple labels. However, only a few methods for automated prediction of protein localization have been developed, and they mostly concentrate on single-label classification. Therefore, a recognition system for multi-label classification of HPA with acceptable performance should be developed. Hence, this study aims to develop a deep learning-based system for the multi-label classification of HPA. Specifically, two architectures have been proposed in this work for automatically extracting features from the images and predicting the localization of the proteins in 28 subcellular compartments. First, a convolutional neural network has been proposed, which has been trained from scratch and second an ensemble-based model using transfer learning architectures has been proposed. The results demonstrate that both models are effective in classifying proteins according to their location in the major cellular organelles. Yet, in this study, the proposed convolutional network outperforms the ensemble model in classification of images with multiple simultaneous protein localizations. Three performance metrics standards—recall, accuracy, and f1-score—were used to assess the models. The proposed convolutional neural network beats the ensemble model by achieving recall of 0.75, precision of 0.75 and f1-score of 0.74.

Dementia is a term used to describe a variety of symptoms related to cognitive impairment in which Alzheimer disease represents 60% – 70% of the cases. As of today, there is no cure for this disease and the only way to prevent any associated medical, economic, and financial impacts or losses is to detect the disease early and work closely with suspected patients to prevent any further progress. In this research, a methodology consisting of 4 modules is proposed: (1) preprocessing, exemplar pyramid along with bi-linear interpolation followed by (2) feature extraction using Gray Level Co-Occurrence Matrix and Local Binary Pattern then (3) concatenation of all extracted features and finally (4) classification of Alzheimer disease stage using deep learning, Multi-Layer Perceptron, in particular. Our proposed method was tested using the MPRAGE structural MRI dataset from Alzheimer Disease Neuro Imaging Initiative (ADNI), and it outperformed other techniques used in the literature review. An accuracy result of 89.80 was reported for multi-class classification of 4 stages of Alzheimer disease (Cognitive Normal, Early Mild Cognitive Impairment, Late Mild Cognitive Impairment and Alzheimer Disease) for both Gray Matter (GM) and White Matter (WM). In term of binary-class classification, we were able to achieve very good results using both GM and WM. By using GM, we were able to distinguish between CN vs EMCI, EMCI vs AD and LMCI vs AD with accuracy results of 96.43%, 90.91% and 95.24% respectively. And using WM, we were able to distinguish between CN vs LMCI with 100% accuracy and EMCI vs LMCI with 95.65% accuracy. While we achieved the same accuracy result of 96.15 using both WM and GM.

Advances in wearable technologies have the ability to revolutionize and improve people’s lives. The gains go beyond the personal sphere, encompassing business and, by extension, the global economy. The technologies are incorporated in electronic devices that collect data from consumers’ bodies and their immediate environment. Human activities recognition, which involves the use of various body sensors and modalities either separately or simultaneously, is one of the most important areas of wearable technology development. In real-life scenarios, the number of sensors deployed is dictated by practical and financial considerations. In the research for this article, we reviewed our earlier efforts and have accordingly reduced the number of required sensors, limiting ourselves to first-person vision data for activities recognition. Nonetheless, our results beat state of the art by more than 4% of F1 score.

This paper presents a multi-unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) system, where multiple UAVs (variable number of UAVs) are deployed to serve Internet of Things devices (IoTDs). We aim to minimize the sum of hovering and flying energies of UAVs by optimizing the trajectories of UAVs. The problem is very complicated as we have to consider the deployment of stop points (SPs), the association between UAVs and SPs, and the order of SPs for UAVs. To solve the problem, this paper proposed a novel genetic trajectory planning algorithm with variable population size (GTPA-VP), which consists of two phases. In the first phase, a genetic algorithm (GA) with a variable population size is used to update the deployment of SPs. Accordingly, a multi-chrome GA is adopted to find the association between UAVs and SPs, an optimal number of UAVs, and the optimal order of SPs for UAVs. The effectiveness of the proposed GTPA-VP is demonstrated through several experiments on a set of ten instances with up to 200 IoTDs. It is evident from the experimental results that the proposed GTPA-VP outperforms the benchmark algorithms in terms of the energy consumption of the system.

Cardiovascular disease (CVD) is reported to be the leading cause of mortality in the middle eastern countries, including Qatar. But no comprehensive study has been conducted on the Qatar specific CVD risk factors identification. The objective of this case-control study was to develop machine learning (ML) model distinguishing healthy individuals from people having CVD, which could ultimately reveal the list of potential risk factors associated to CVD in Qatar. To the best of our knowledge, this study considered the largest collection of biomedical measurements representing the anthropometric measurements, clinical biomarkers, bioimpedance, spirometry, VICORDER readings, and behavioral factors of the CVD group from Qatar Biobank (QBB). CatBoost model achieved 93% accuracy, thereby outperforming the existing model for the same purpose. Interestingly, combining multimodal datasets into the proposed ML model outperformed the ML model built upon currently known risk factors for CVD, emphasizing the importance of incorporating other clinical biomarkers into consideration for CVD diagnosis plan. The ablation study on the multimodal dataset from QBB revealed that physio-clinical and bioimpedance measurements have the most distinguishing power to classify these two groups irrespective of gender and age of the participants. Multiple feature subset selection techniques confirmed known CVD risk factors (blood pressure, lipid profile, smoking, sedentary life, and diabetes), and identified potential novel risk factors linked to CVD-related comorbidities such as renal disorder (e.g., creatinine, uric acid, homocysteine, albumin), atherosclerosis (intima media thickness), hypercoagulable state (fibrinogen), and liver function (e.g., alkaline phosphate, gamma-glutamyl transferase). Moreover, the inclusion of the proposed novel factors into the ML model provides better performance than the model with traditional known risk factors for CVD. The association of the proposed risk factors and comorbidities are required to be investigated in clinical setup to understand their role in CVD better.

Diabetes is a costly and burdensome metabolic disorder that occurs due to the elevation of glucose levels in the bloodstream. If it goes unchecked for an extended period, it can lead to the damage of different body organs and develop life-threatening health complications. Studies show that the progression of diabetes can be stopped or delayed, provided a person follows a healthy lifestyle and takes proper medication. Prevention of diabetes or the delayed onset of diabetes is crucial, and it can be achieved if there exists a screening process that identifies individuals who are at risk of developing diabetes in the future. Although machine learning techniques have been applied for disease diagnosis, there is little work done on long term prediction of disease, type 2 diabetes in particular. Moreover, finding discriminative features or risk-factors responsible for the future development of diabetes plays a significant role. In this study, we propose two novel feature extraction approaches for finding the best risk-factors, followed by applying a machine learning pipeline for the long term prediction of type 2 diabetes. The proposed methods have been evaluated using data from a longitudinal clinical study, known as the San Antonio Heart Study. Our proposed model managed to achieve 95.94% accuracy in predicting whether a person will develop type 2 diabetes within the next 7-8 years or not.

Particle Swarm Optimization (PSO) is a metaheuristic evolutionary computation technique inspired by the social behavior of birds and fish flock. The classical PSO has limitations of slow convergence rate and trapping in local minima, as the dimensions of the data increase. Moreover, majority of improvements made by the researchers in optimization techniques have focused on the accuracy of solution and have overlooked the convergence speed of an algorithm. Keeping in view the need of an optimization algorithm with fast convergence speed, suitable for high dimensional data space, this article proposes a novel concept of Multi-Cluster Jumping PSO. In the proposed method, the particles in the swarm are divided in different clusters to search for the global optimum solution. Each cluster in the swarm has its own cluster best position which is the best position within a cluster and the global best position is located by clusters communication. In order to avoid trapping in the local optima, a jumping strategy is incorporated for stuck particles through relocation of particles to a random new position. Instead of random initialization of the particles, a semi-random initialization is opted by dividing the entire search space and the distribution of particles over a search space is done in independent slots. The proposed approach has the ability to overcome the limitations of classical evolutionary computation methods and is suitable for high dimensional dynamic data. An extensive experimentation is carried out to optimize twelve benchmark functions using the proposed Multi-Cluster Jumping PSO and a significant difference is observed in the convergence speed of the proposed method over the existing state-of-the-art approaches.

Smart health is one of the most popular and important components of smart cities. It is a relatively new context-aware healthcare paradigm influenced by several fields of expertise, such as medical informatics, communications and electronics, bioengineering, ethics, to name a few. Smart health is used to improve healthcare by providing many services such as patient monitoring, early diagnosis of disease and so on. The artificial neural network (ANN), support vector machine (SVM) and deep learning models, especially the convolutional neural network (CNN), are the most commonly used machine learning approaches where they proved to be performance in most cases. Voice disorders are rapidly spreading especially with the development of medical diagnostic systems, although they are often underestimated. Smart health systems can be an easy and fast support to voice pathology detection. The identification of an algorithm that discriminates between pathological and healthy voices with more accuracy is needed to obtain a smart and precise mobile health system. The main contribution of this paper consists of proposing a multiclass-pathologic voice classification using a novel multileveled textural feature extraction with iterative feature selector. Our approach is a simple and efficient voice-based algorithm in which a multi-center and multi threshold based ternary pattern is used (MCMTTP). A more compact multileveled features are then obtained by sample-based discretization techniques and Neighborhood Component Analysis (NCA) is applied to select features iteratively. These features are finally integrated with MCMTTP to achieve an accurate voice-based features detection. Experimental results of six classifiers with three diagnostic diseases (frontal resection, cordectomy and spastic dysphonia) show that the fused features are more suitable for describing voice-based disease detection.

Sleep Apnea is a breathing disorder that occurs while the patient is sleeping. Traditionally, Polysomnography is used to diagnose it. However, it is quite inconvenient and expensive. Because of the troublesome diagnosis, this ailment often remained undiagnosed. This paper aims at the development of such a method that provides an easy diagnostic solution to the doctors. Electrocardiogram (ECG) is one of the most common tests done at the hospitals. In this paper, we aim to develop a method which deploys ECG data to diagnose the sleep ailment, Apnea. A technique deploying wavelet packet transform on RR interval of ECG has been presented. Probability density functions of data, both when Apnea is present and when it is not, are obtained by constructing histograms of decision variable for each signal segment. From the overlapping PDFs of the normal and abnormal cases, a threshold is then derived. This helped in segregating the Apnea cases from normal cases. The stated method provided a 100% accuracy in diagnosing Sleep Apnea.