Thi Qar University

In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

The sol‐gel process is a more chemical method (wet chemical method) for the synthesis of various nanostructures, especially metal oxide nanoparticles. In this method, the molecular precursor (usually metal alkoxide) is dissolved in water or alcohol and converted to gel by heating and stirring by hydrolysis/alcoholysis. Since the gel obtained from the hydrolysis/alcoholysis process is wet or damp, it should be dried using appropriate methods depending on the desired properties and application of the gel. For example, if it is an alcoholic solution, the drying process is done by burning alcohol. After the drying stage, the produced gels are powdered and then calcined. The sol‐gel method is a cost‐effective method and due to the low reaction temperature there is good control over the chemical composition of the products. The sol‐gel method can be used in the process of making ceramics as a molding material and can be used as an intermediate between thin films of metal oxides in various applications. The materials obtained from the sol‐gel method are used in various optical, electronic, energy, surface engineering, biosensors, and pharmaceutical and separation technologies (such as chromatography). The sol‐gel method is a conventional and industrial method for the synthesis of nanoparticles with different chemical composition. The basis of the sol‐gel method is the production of a homogeneous sol from the precursors and its conversion into a gel. The solvent in the gel is then removed from the gel structure and the remaining gel is dried. The properties of the dried gel depend significantly on the drying method. In other words, the “removing solvent method” is selected according to the application in which the gel will be used. Dried gels in various ways are used in industries such as surface coating, building insulation, and the production of special clothing. It is worth mentioning that, by grinding the gel by special mills, it is possible to achieve nanoparticles.

Abstract Data scarcity is a major challenge when training deep learning (DL) models. DL demands a large amount of data to achieve exceptional performance. Unfortunately, many applications have small or inadequate data to train DL frameworks. Usually, manual labeling is needed to provide labeled data, which typically involves human annotators with a vast background of knowledge. This annotation process is costly, time-consuming, and error-prone. Usually, every DL framework is fed by a significant amount of labeled data to automatically learn representations. Ultimately, a larger amount of data would generate a better DL model and its performance is also application dependent. This issue is the main barrier for many applications dismissing the use of DL. Having sufficient data is the first step toward any successful and trustworthy DL application. This paper presents a holistic survey on state-of-the-art techniques to deal with training DL models to overcome three challenges including small, imbalanced datasets, and lack of generalization. This survey starts by listing the learning techniques. Next, the types of DL architectures are introduced. After that, state-of-the-art solutions to address the issue of lack of training data are listed, such as Transfer Learning (TL), Self-Supervised Learning (SSL), Generative Adversarial Networks (GANs), Model Architecture (MA), Physics-Informed Neural Network (PINN), and Deep Synthetic Minority Oversampling Technique (DeepSMOTE). Then, these solutions were followed by some related tips about data acquisition needed prior to training purposes, as well as recommendations for ensuring the trustworthiness of the training dataset. The survey ends with a list of applications that suffer from data scarcity, several alternatives are proposed in order to generate more data in each application including Electromagnetic Imaging (EMI), Civil Structural Health Monitoring, Medical imaging, Meteorology, Wireless Communications, Fluid Mechanics, Microelectromechanical system, and Cybersecurity. To the best of the authors’ knowledge, this is the first review that offers a comprehensive overview on strategies to tackle data scarcity in DL.

Current evidence links oxidative stress (OS) to male infertility, reduced sperm motility, sperm DNA damage and increased risk of recurrent abortions and genetic diseases. A review of PubMed, Medline, Google Scholar, and Cochrane review databases of published articles from years 2000-2018 was performed focusing on physiological and pathological consequences of reactive oxygen species (ROS), sperm DNA damage, OS tests, and the association between OS and male infertility, pregnancy and assisted reproductive techniques outcomes. Generation of ROS is essential for reproductive function, but OS is detrimental to fertility, pregnancy, and genetic status of the newborns. Further, there is a lack of consensus on selecting OS test, type, and duration of antioxidants treatment as well as on the target patients group. Developing advanced diagnostic and therapeutic options for OS is essential to improve fertility potential and limit genetic diseases transmitted to offspring.

Developing accurate soft computing methods for groundwater level (GWL) forecasting is essential for enhancing the planning and management of water resources. Over the past two decades, significant progress has been made in GWL prediction using machine learning (ML) models. Several review articles have been published, reporting the advances in this field up to 2018. However, the existing review articles do not cover several aspects of GWL simulations using ML, which are significant for scientists and practitioners working in hydrology and water resource management. The current review article aims to provide a clear understanding of the state-of-the-art ML models implemented for GWL modeling and the milestones achieved in this domain. The review includes all of the types of ML models employed for GWL modeling from 2008 to 2020 (138 articles) and summarizes the details of the reviewed papers, including the types of models, data span, time scale, input and output parameters, performance criteria used, and the best models identified. Furthermore, recommendations for possible future research directions to improve the accuracy of GWL prediction models and enhance the related knowledge are outlined.

BACKGROUND: For more than three decades, the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) has provided a framework to quantify health loss due to diseases, injuries, and associated risk factors. This paper presents GBD 2023 findings on disease and injury burden and risk-attributable health loss, offering a global audit of the state of world health to inform public health priorities. This work captures the evolving landscape of health metrics across age groups, sexes, and locations, while reflecting on the remaining post-COVID-19 challenges to achieving our collective global health ambitions. METHODS: The GBD 2023 combined analysis estimated years lived with disability (YLDs), years of life lost (YLLs), and disability-adjusted life-years (DALYs) for 375 diseases and injuries, and risk-attributable burden associated with 88 modifiable risk factors. Of the more than 310 000 total data sources used for all GBD 2023 (about 30% of which were new to this estimation round), more than 120 000 sources were used for estimation of disease and injury burden and 59 000 for risk factor estimation, and included vital registration systems, surveys, disease registries, and published scientific literature. Data were analysed using previously established modelling approaches, such as disease modelling meta-regression version 2.1 (DisMod-MR 2.1) and comparative risk assessment methods. Diseases and injuries were categorised into four levels on the basis of the established GBD cause hierarchy, as were risk factors using the GBD risk hierarchy. Estimates stratified by age, sex, location, and year from 1990 to 2023 were focused on disease-specific time trends over the 2010-23 period and presented as counts (to three significant figures) and age-standardised rates per 100 000 person-years (to one decimal place). For each measure, 95% uncertainty intervals [UIs] were calculated with the 2·5th and 97·5th percentile ordered values from a 250-draw distribution. FINDINGS: Total numbers of global DALYs grew 6·1% (95% UI 4·0-8·1), from 2·64 billion (2·46-2·86) in 2010 to 2·80 billion (2·57-3·08) in 2023, but age-standardised DALY rates, which account for population growth and ageing, decreased by 12·6% (11·0-14·1), revealing large long-term health improvements. Non-communicable diseases (NCDs) contributed 1·45 billion (1·31-1·61) global DALYs in 2010, increasing to 1·80 billion (1·63-2·03) in 2023, alongside a concurrent 4·1% (1·9-6·3) reduction in age-standardised rates. Based on DALY counts, the leading level 3 NCDs in 2023 were ischaemic heart disease (193 million [176-209] DALYs), stroke (157 million [141-172]), and diabetes (90·2 million [75·2-107]), with the largest increases in age-standardised rates since 2010 occurring for anxiety disorders (62·8% [34·0-107·5]), depressive disorders (26·3% [11·6-42·9]), and diabetes (14·9% [7·5-25·6]). Remarkable health gains were made for communicable, maternal, neonatal, and nutritional (CMNN) diseases, with DALYs falling from 874 million (837-917) in 2010 to 681 million (642-736) in 2023, and a 25·8% (22·6-28·7) reduction in age-standardised DALY rates. During the COVID-19 pandemic, DALYs due to CMNN diseases rose but returned to pre-pandemic levels by 2023. From 2010 to 2023, decreases in age-standardised rates for CMNN diseases were led by rate decreases of 49·1% (32·7-61·0) for diarrhoeal diseases, 42·9% (38·0-48·0) for HIV/AIDS, and 42·2% (23·6-56·6) for tuberculosis. Neonatal disorders and lower respiratory infections remained the leading level 3 CMNN causes globally in 2023, although both showed notable rate decreases from 2010, declining by 16·5% (10·6-22·0) and 24·8% (7·4-36·7), respectively. Injury-related age-standardised DALY rates decreased by 15·6% (10·7-19·8) over the same period. Differences in burden due to NCDs, CMNN diseases, and injuries persisted across age, sex, time, and location. Based on our risk analysis, nearly 50% (1·27 billion [1·18-1·38]) of the roughly 2·80 billion total global DALYs in 2023 were attributable to the 88 risk factors analysed in GBD. Globally, the five level 3 risk factors contributing the highest proportion of risk-attributable DALYs were high systolic blood pressure (SBP), particulate matter pollution, high fasting plasma glucose (FPG), smoking, and low birthweight and short gestation-with high SBP accounting for 8·4% (6·9-10·0) of total DALYs. Of the three overarching level 1 GBD risk factor categories-behavioural, metabolic, and environmental and occupational-risk-attributable DALYs rose between 2010 and 2023 only for metabolic risks, increasing by 30·7% (24·8-37·3); however, age-standardised DALY rates attributable to metabolic risks decreased by 6·7% (2·0-11·0) over the same period. For all but three of the 25 leading level 3 risk factors, age-standardised rates dropped between 2010 and 2023-eg, declining by 54·4% (38·7-65·3) for unsafe sanitation, 50·5% (33·3-63·1) for unsafe water source, and 45·2% (25·6-72·0) for no access to handwashing facility, and by 44·9% (37·3-53·5) for child growth failure. The three leading level 3 risk factors for which age-standardised attributable DALY rates rose were high BMI (10·5% [0·1 to 20·9]), drug use (8·4% [2·6 to 15·3]), and high FPG (6·2% [-2·7 to 15·6]; non-significant). INTERPRETATION: Our findings underscore the complex and dynamic nature of global health challenges. Since 2010, there have been large decreases in burden due to CMNN diseases and many environmental and behavioural risk factors, juxtaposed with sizeable increases in DALYs attributable to metabolic risk factors and NCDs in growing and ageing populations. This long-observed consequence of the global epidemiological transition was only temporarily interrupted by the COVID-19 pandemic. The substantially decreasing CMNN disease burden, despite the 2008 global financial crisis and pandemic-related disruptions, is one of the greatest collective public health successes known. However, these achievements are at risk of being reversed due to major cuts to development assistance for health globally, the effects of which will hit low-income countries with high burden the hardest. Without sustained investment in evidence-based interventions and policies, progress could stall or reverse, leading to widespread human costs and geopolitical instability. Moreover, the rising NCD burden necessitates intensified efforts to mitigate exposure to leading risk factors-eg, air pollution, smoking, and metabolic risks, such as high SBP, BMI, and FPG-including policies that promote food security, healthier diets, physical activity, and equitable and expanded access to potential treatments, such as GLP-1 receptor agonists. Decisive, coordinated action is needed to address long-standing yet growing health challenges, including depressive and anxiety disorders. Yet this can be only part of the solution. Our response to the NCD syndemic-the complex interaction of multiple health risks, social determinants, and systemic challenges-will define the future landscape of global health. To ensure human wellbeing, economic stability, and social equity, global action to sustain and advance health gains must prioritise reducing disparities by addressing socioeconomic and demographic determinants, ensuring equitable health-care access, tackling malnutrition, strengthening health systems, and improving vaccination coverage. We live in times of great opportunity. FUNDING: Gates Foundation and Bloomberg Philanthropies.

Deep learning requires a large amount of data to perform well. However, the field of medical image analysis suffers from a lack of sufficient data for training deep learning models. Moreover, medical images require manual labeling, usually provided by human annotators coming from various backgrounds. More importantly, the annotation process is time-consuming, expensive, and prone to errors. Transfer learning was introduced to reduce the need for the annotation process by transferring the deep learning models with knowledge from a previous task and then by fine-tuning them on a relatively small dataset of the current task. Most of the methods of medical image classification employ transfer learning from pretrained models, e.g., ImageNet, which has been proven to be ineffective. This is due to the mismatch in learned features between the natural image, e.g., ImageNet, and medical images. Additionally, it results in the utilization of deeply elaborated models. In this paper, we propose a novel transfer learning approach to overcome the previous drawbacks by means of training the deep learning model on large unlabeled medical image datasets and by next transferring the knowledge to train the deep learning model on the small amount of labeled medical images. Additionally, we propose a new deep convolutional neural network (DCNN) model that combines recent advancements in the field. We conducted several experiments on two challenging medical imaging scenarios dealing with skin and breast cancer classification tasks. According to the reported results, it has been empirically proven that the proposed approach can significantly improve the performance of both classification scenarios. In terms of skin cancer, the proposed model achieved an F1-score value of 89.09% when trained from scratch and 98.53% with the proposed approach. Secondly, it achieved an accuracy value of 85.29% and 97.51%, respectively, when trained from scratch and using the proposed approach in the case of the breast cancer scenario. Finally, we concluded that our method can possibly be applied to many medical imaging problems in which a substantial amount of unlabeled image data is available and the labeled image data is limited. Moreover, it can be utilized to improve the performance of medical imaging tasks in the same domain. To do so, we used the pretrained skin cancer model to train on feet skin to classify them into two classes-either normal or abnormal (diabetic foot ulcer (DFU)). It achieved an F1-score value of 86.0% when trained from scratch, 96.25% using transfer learning, and 99.25% using double-transfer learning.

Abstract Streamflow ( Q flow ) prediction is one of the essential steps for the reliable and robust water resources planning and management. It is highly vital for hydropower operation, agricultural planning, and flood control. In this study, the convolution neural network (CNN) and Long-Short-term Memory network (LSTM) are combined to make a new integrated model called CNN-LSTM to predict the hourly Q flow (short-term) at Brisbane River and Teewah Creek, Australia. The CNN layers were used to extract the features of Q flow time-series, while the LSTM networks use these features from CNN for Q flow time series prediction. The proposed CNN-LSTM model is benchmarked against the standalone model CNN, LSTM, and Deep Neural Network models and several conventional artificial intelligence (AI) models. Q flow prediction is conducted for different time intervals with the length of 1-Week, 2-Weeks, 4-Weeks, and 9-Months, respectively. With the help of different performance metrics and graphical analysis visualization, the experimental results reveal that with small residual error between the actual and predicted Q flow , the CNN-LSTM model outperforms all the benchmarked conventional AI models as well as ensemble models for all the time intervals. With 84% of Q flow prediction error below the range of 0.05 m 3 s −1 , CNN-LSTM demonstrates a better performance compared to 80% and 66% for LSTM and DNN, respectively. In summary, the results reveal that the proposed CNN-LSTM model based on the novel framework yields more accurate predictions. Thus, CNN-LSTM has significant practical value in Q flow prediction.

BACKGROUND: Timely and comprehensive analyses of causes of death stratified by age, sex, and location are essential for shaping effective health policies aimed at reducing global mortality. The Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2023 provides cause-specific mortality estimates measured in counts, rates, and years of life lost (YLLs). GBD 2023 aimed to enhance our understanding of the relationship between age and cause of death by quantifying the probability of dying before age 70 years (70q0) and the mean age at death by cause and sex. This study enables comparisons of the impact of causes of death over time, offering a deeper understanding of how these causes affect global populations. METHODS: GBD 2023 produced estimates for 292 causes of death disaggregated by age-sex-location-year in 204 countries and territories and 660 subnational locations for each year from 1990 until 2023. We used a modelling tool developed for GBD, the Cause of Death Ensemble model (CODEm), to estimate cause-specific death rates for most causes. We computed YLLs as the product of the number of deaths for each cause-age-sex-location-year and the standard life expectancy at each age. Probability of death was calculated as the chance of dying from a given cause in a specific age period, for a specific population. Mean age at death was calculated by first assigning the midpoint age of each age group for every death, followed by computing the mean of all midpoint ages across all deaths attributed to a given cause. We used GBD death estimates to calculate the observed mean age at death and to model the expected mean age across causes, sexes, years, and locations. The expected mean age reflects the expected mean age at death for individuals within a population, based on global mortality rates and the population's age structure. Comparatively, the observed mean age represents the actual mean age at death, influenced by all factors unique to a location-specific population, including its age structure. As part of the modelling process, uncertainty intervals (UIs) were generated using the 2·5th and 97·5th percentiles from a 250-draw distribution for each metric. Findings are reported as counts and age-standardised rates. Methodological improvements for cause-of-death estimates in GBD 2023 include a correction for the misclassification of deaths due to COVID-19, updates to the method used to estimate COVID-19, and updates to the CODEm modelling framework. This analysis used 55 761 data sources, including vital registration and verbal autopsy data as well as data from surveys, censuses, surveillance systems, and cancer registries, among others. For GBD 2023, there were 312 new country-years of vital registration cause-of-death data, 3 country-years of surveillance data, 51 country-years of verbal autopsy data, and 144 country-years of other data types that were added to those used in previous GBD rounds. FINDINGS: The initial years of the COVID-19 pandemic caused shifts in long-standing rankings of the leading causes of global deaths: it ranked as the number one age-standardised cause of death at Level 3 of the GBD cause classification hierarchy in 2021. By 2023, COVID-19 dropped to the 20th place among the leading global causes, returning the rankings of the leading two causes to those typical across the time series (ie, ischaemic heart disease and stroke). While ischaemic heart disease and stroke persist as leading causes of death, there has been progress in reducing their age-standardised mortality rates globally. Four other leading causes have also shown large declines in global age-standardised mortality rates across the study period: diarrhoeal diseases, tuberculosis, stomach cancer, and measles. Other causes of death showed disparate patterns between sexes, notably for deaths from conflict and terrorism in some locations. A large reduction in age-standardised rates of YLLs occurred for neonatal disorders. Despite this, neonatal disorders remained the leading cause of global YLLs over the period studied, except in 2021, when COVID-19 was temporarily the leading cause. Compared to 1990, there has been a considerable reduction in total YLLs in many vaccine-preventable diseases, most notably diphtheria, pertussis, tetanus, and measles. In addition, this study quantified the mean age at death for all-cause mortality and cause-specific mortality and found noticeable variation by sex and location. The global all-cause mean age at death increased from 46·8 years (95% UI 46·6-47·0) in 1990 to 63·4 years (63·1-63·7) in 2023. For males, mean age increased from 45·4 years (45·1-45·7) to 61·2 years (60·7-61·6), and for females it increased from 48·5 years (48·1-48·8) to 65·9 years (65·5-66·3), from 1990 to 2023. The highest all-cause mean age at death in 2023 was found in the high-income super-region, where the mean age for females reached 80·9 years (80·9-81·0) and for males 74·8 years (74·8-74·9). By comparison, the lowest all-cause mean age at death occurred in sub-Saharan Africa, where it was 38·0 years (37·5-38·4) for females and 35·6 years (35·2-35·9) for males in 2023. Lastly, our study found that all-cause 70q0 decreased across each GBD super-region and region from 2000 to 2023, although with large variability between them. For females, we found that 70q0 notably increased from drug use disorders and conflict and terrorism. Leading causes that increased 70q0 for males also included drug use disorders, as well as diabetes. In sub-Saharan Africa, there was an increase in 70q0 for many non-communicable diseases (NCDs). Additionally, the mean age at death from NCDs was lower than the expected mean age at death for this super-region. By comparison, there was an increase in 70q0 for drug use disorders in the high-income super-region, which also had an observed mean age at death lower than the expected value. INTERPRETATION: We examined global mortality patterns over the past three decades, highlighting-with enhanced estimation methods-the impacts of major events such as the COVID-19 pandemic, in addition to broader trends such as increasing NCDs in low-income regions that reflect ongoing shifts in the global epidemiological transition. This study also delves into premature mortality patterns, exploring the interplay between age and causes of death and deepening our understanding of where targeted resources could be applied to further reduce preventable sources of mortality. We provide essential insights into global and regional health disparities, identifying locations in need of targeted interventions to address both communicable and non-communicable diseases. There is an ever-present need for strengthened health-care systems that are resilient to future pandemics and the shifting burden of disease, particularly among ageing populations in regions with high mortality rates. Robust estimates of causes of death are increasingly essential to inform health priorities and guide efforts toward achieving global health equity. The need for global collaboration to reduce preventable mortality is more important than ever, as shifting burdens of disease are affecting all nations, albeit at different paces and scales. FUNDING: Gates Foundation.

One of the main challenges of employing deep learning models in the field of medicine is a lack of training data due to difficulty in collecting and labeling data, which needs to be performed by experts. To overcome this drawback, transfer learning (TL) has been utilized to solve several medical imaging tasks using pre-trained state-of-the-art models from the ImageNet dataset. However, there are primary divergences in data features, sizes, and task characteristics between the natural image classification and the targeted medical imaging tasks. Therefore, TL can slightly improve performance if the source domain is completely different from the target domain. In this paper, we explore the benefit of TL from the same and different domains of the target tasks. To do so, we designed a deep convolutional neural network (DCNN) model that integrates three ideas including traditional and parallel convolutional layers and residual connections along with global average pooling. We trained the proposed model against several scenarios. We utilized the same and different domain TL with the diabetic foot ulcer (DFU) classification task and with the animal classification task. We have empirically shown that the source of TL from the same domain can significantly improve the performance considering a reduced number of images in the same domain of the target dataset. The proposed model with the DFU dataset achieved F1-score value of 86.6% when trained from scratch, 89.4% with TL from a different domain of the targeted dataset, and 97.6% with TL from the same domain of the targeted dataset.

Coronaviruses (CoVs) infect many species causing a variety of diseases with a range of severities. Their members include zoonotic viruses with pandemic potential where therapeutic options are currently limited. Despite this diversity CoVs share some common features including the production, in infected cells, of elaborate membrane structures. Membranes represent both an obstacle and aid to CoV replication - and in consequence - virus-encoded structural and nonstructural proteins have membrane-binding properties. The structural proteins encounter cellular membranes at both entry and exit of the virus while the nonstructural proteins reorganize cellular membranes to benefit virus replication. Here, the role of each protein in membrane binding is described to provide a comprehensive picture of their role in the CoV replication cycle.

Abstract Feature extraction is the main core in diagnosis, classification, clustering, recognition, and detection. Many researchers may by interesting in choosing suitable features that used in the applications. In this paper, the most important features methods are collected, and explained each one. The features in this paper are divided into four groups; Geometric features, Statistical features, Texture features, and Color features. It explains the methodology of each method, its equations, and application. In this paper, we made acomparison among them by using two types of image, one type for face images (163 images divided into 113 for training and 50 for testing) and the other for plant images(130 images divided into 100 for training and 30 for testing) to test the features in geometric and textures. Each type of image group shows that each type of images may be used suitable features may differ from other types.

The turbulent forced convection heat transfer of water/functionalized multi-walled carbon nanotube (FMWCNT) nanofluids over a forward-facing step was studied in this work. Turbulence was modeled using the shear stress transport K-ω model. Simulations were performed for Reynolds numbers ranging from 10,000 to 40,000, heat fluxes from 1,000 to 10,000 W/m2, and nanoparticle volume fractions of 0.00% to 0.25%. The two-dimensional governing equations were discretized with the finite volume method. The effects of nanoparticle concentration, shear force, heat flux, contraction, and turbulence on the hydraulics and thermal behavior of nanofluid flow were studied. The model predictions were found to be in good agreement with previous experimental and numerical studies. The results indicate that the Reynolds number and FMWCNT volume fraction considerably affect the heat transfer coefficient; a rise in local heat transfer coefficient was noted when both Reynolds number and FMWCNT volume fraction were increased for all cases. Moreover, the contraction of the channel passage leads to the formation of two recirculation regions with augmented local heat transfer coefficient value.

Opportunity has investigated in detail rocks on the rim of the Noachian age Endeavour crater, where orbital spectral reflectance signatures indicate the presence of Fe(+3)-rich smectites. The signatures are associated with fine-grained, layered rocks containing spherules of diagenetic or impact origin. The layered rocks are overlain by breccias, and both units are cut by calcium sulfate veins precipitated from fluids that circulated after the Endeavour impact. Compositional data for fractures in the layered rocks suggest formation of Al-rich smectites by aqueous leaching. Evidence is thus preserved for water-rock interactions before and after the impact, with aqueous environments of slightly acidic to circum-neutral pH that would have been more favorable for prebiotic chemistry and microorganisms than those recorded by younger sulfate-rich rocks at Meridiani Planum.

Smart cities result from integrating advanced technologies and intelligent sensors into modern urban infrastructure. The Internet of Things (IoT) and data integration are pivotal in creating interconnected and intelligent urban spaces. In this literature review, we explore the different methods of information fusion used in smart cities, along with their advantages and challenges. However, there are notable challenges in managing diverse data sources, handling large data volumes, and meeting the near-real-time demands of various smart city applications. The review aims to examine smart city applications in detail, incorporating quality evaluation and information fusion techniques and identifying critical issues while outlining promising research directions. In order to accomplish our goal, we conducted a comprehensive search of literature and applied selective criteria. We identified 59 recent studies addressing machine learning (ML) and deep learning (DL) techniques in smart city applications. These studies were obtained from various databases such as ScienceDirect (SD), Scopus, Web of Science (WoS), and IEEE Xplore. The main objective of this study is to provide more detailed insights into smart cities by supplementing existing research. The word cloud visualisation of machine learning/deep learning and information fusion in smart cities papers shows a diverse landscape, covering both technical aspects of artificial intelligence and practical applications in urban settings. Apart from technical exploration, the study also delves into the ethical and privacy implications arising in smart cities. Moreover, it thoroughly examines the challenges that must be addressed to realise this urban revolution's potential fully.

Sickle cell anemia, which is also called sickle cell disease (SCD), is a hematological disorder that causes occlusion in blood vessels, leading to hurtful episodes and even death. The key function of red blood cells (erythrocytes) is to supply all the parts of the human body with oxygen. Red blood cells (RBCs) form a crescent or sickle shape when sickle cell anemia affects them. This abnormal shape makes it difficult for sickle cells to move through the bloodstream, hence decreasing the oxygen flow. The precise classification of RBCs is the first step toward accurate diagnosis, which aids in evaluating the danger level of sickle cell anemia. The manual classification methods of erythrocytes require immense time, and it is possible that errors may be made throughout the classification stage. Traditional computer-aided techniques, which have been employed for erythrocyte classification, are based on handcrafted features techniques, and their performance relies on the selected features. They also are very sensitive to different sizes, colors, and complex shapes. However, microscopy images of erythrocytes are very complex in shape with different sizes. To this end, this research proposes lightweight deep learning models that classify the erythrocytes into three classes: circular (normal), elongated (sickle cells), and other blood content. These models are different in the number of layers and learnable filters. The available datasets of red blood cells with sickle cell disease are very small for training deep learning models. Therefore, addressing the lack of training data is the main aim of this paper. To tackle this issue and optimize the performance, the transfer learning technique is utilized. Transfer learning does not significantly affect performance on medical image tasks when the source domain is completely different from the target domain. In some cases, it can degrade the performance. Hence, we have applied the same domain transfer learning, unlike other methods that used the ImageNet dataset for transfer learning. To minimize the overfitting effect, we have utilized several data augmentation techniques. Our model obtained state-of-the-art performance and outperformed the latest methods by achieving an accuracy of 99.54% with our model and 99.98% with our model plus a multiclass SVM classifier on the erythrocytesIDB dataset and 98.87% on the collected dataset.

Nanoparticles have revolutioned materials secience and manufacturing industry thanks to their ultrafine scale. Nano-scale titanium oxide (nano-TiO2) are key elements of a wide variety of sectors of advanced materials, devices, and systems in view of their versatility and extraordinary characetristics, mainly optical, electrical and catalytic properties. Thnaks to aforementioned features, they found a key position in different industries such as colors and pigments, photocatalysts, photovoltaic solar cells, adsorbents, cosmetics, electric and electronic devices. Nano-SiO2 application gives rise to a wide range of properties and performance, which severely depends on particle size, particle size distribution, and agglomeration in the bulk of host materials or systems. When they are used as surface coating additive or even functionalized, nano- TiO2 takes more roles, which are state-of-the-art depending on synthesis method. Therefore, review of synthesis, properties, surface- and bulk-functionalization/modification (respectively with function groups and doping) methods in achieving high-performance nano- TiO2 seems essential. In this review, background information and developments from pure Titanium Oxide to chemically modified Titanium Oxide -based materials as photocatalysts were discussed in detail, Photocatalytic activity of Titanium oxide-based materials in nitrogen group pollutant decomposition. It also Titanium Oxide Nanoparticles as a Vehicle for Chemotherapeutics and Other Applications of Titanium Oxide Nanoparticles in Medicine and also an overview of the synthesis methods and factors affecting their properties in different synthesis methods have been reviewed.

In this paper, some recent piezoelectric wafer active sensors (PWAS) progress achieved in our laboratory for active materials and smart structures (LAMSS) at the University of South Carolina: http: //www.me.sc.edu/research/lamss/ group is presented. First, the characterization of the PWAS materials shows that no significant change in the microstructure after exposure to high temperature and nuclear radiation, and the PWAS transducer can be used in harsh environments for structural health monitoring (SHM) applications. Next, PWAS active sensing of various damage types in aluminum and composite structures are explored. PWAS transducers can successfully detect the simulated crack and corrosion damage in aluminum plates through the wavefield analysis, and the simulated delamination damage in composite plates through the damage imaging method. Finally, the novel use of PWAS transducers as acoustic emission (AE) sensors for in situ AE detection during fatigue crack growth is presented. The time of arrival of AE signals at multiple PWAS transducers confirms that the AE signals are originating from the crack, and that the amplitude decay due to geometric spreading is observed.

The utilization of biomass waste as a raw material for renewable energy is a global concern. Pyrolysis is one of the thermal treatments for biomass wastes that results in the production of liquid, solid and gaseous products. Unfortunately, the complex structure of the biomass materials matrix needs elevated heating to convert these materials into useful products. Microwave heating is a promising alternative to conventional heating approaches. Recently, it has been widely used in pyrolysis due to easy operation and its high heating rate. This review tries to identify the microwave-assisted pyrolysis treatment process fundamentals and discusses various key operating parameters which have an effect on product yield. It was found that several operating parameters govern this process such as microwave power and the degree of temperature, microwave absorber addition and its concentration, initial moisture content, initial sweep gas flow rate/residence time. Moreover, this study highlighted the most attractive products of the microwave pyrolysis process. These products include synthesis gas, bio-char, and bio-oil. The benefits and challenges of microwave heating are discussed.

at 0.9 V, highlighting that β is not an intrinsic molecular property. Further theoretical analysis using density functional theory underlines the role of intersite coupling and indicates that this large increase in conductance with length at increasing voltages can be generalized to other molecular oligomers.