Transfer Learning for Chest X-rays Diagnosis Using Dipper ThroatedAlgorithm

Most children and elderly people worldwide die from pneumonia, which is a contagious illness that causes lung ulcers. For diagnosing pneumonia from chest X-ray images, many deep learning models have been put forth. The goal of this research is to develop an effective and strong approach for detecting and categorizing pneumonia cases. By varying the deep learning approach, three pre-trained models, GoogLeNet, ResNet18, and DenseNet121, are employed in this research to extract the main features of pneumonia and normal cases. In addition, the binary dipper throated optimization (DTO) algorithm is utilized to select the most significant features, which are then fed to the K-nearest neighbor (KNN) classifier for getting the final classification decision. To guarantee the best performance of KNN, its main parameter (K) is optimized using the continuous DTO algorithm. To test the proposed approach, six evaluation metrics were employed namely, positive and negative predictive values, accuracy, specificity, sensitivity, and F1-score. Moreover, the proposed approach is compared with other traditional approaches, and the findings confirmed the superiority of the proposed approach in terms of all the evaluation metrics. The minimum accuracy achieved by the proposed approach is (98.5%), and the maximum accuracy is (99.8%) when different test cases are included in the evaluation experiments.     

Multi-source shared nearest neighbours for multi-modal image clustering

Shared Nearest Neighbours (SNN) techniques are well known to overcome several shortcomings of traditional clustering approaches, notably high dimensionality and metric limitations. However, previous methods were limited to a single information source whereas such methods appear to be very well suited for heterogeneous data, typically in multi-modal contexts. In this paper, we propose a new technique to accelerate the calculation of shared neighbours and we introduce a new multi-source shared neighbours scheme applied to multi-modal image clustering. We first extend existing SNN-based similarity measures to the case of multiple sources and we introduce an original automatic source selection step when building candidate clusters. The key point is that each resulting cluster is built with its own optimal subset of modalities which improves the robustness to noisy or outlier information sources. We experiment our method in the scope of multi-modal search result clustering, visual search mining and subspace clustering. Experimental results on both synthetic and real data involving different information sources and several datasets show the effectiveness of our method.     

Deterioration of a Tunisian coastal aquifer due to agricultural activities and possible approaches for better water management

The purpose of this study is to provide a better understanding of the impact of agricultural activities on groundwater quality in a semi‐arid coastal environment (Korba, Tunisia). Impacts have been assessed through groundwater samples that were analysed for a range of chemical and biological parameters. Results indicate that groundwater overexploitation for irrigation and fertiliser use have affected the quality of the groundwater not only by direct contamination but also seawater intrusion. The implementation of appropriate regulatory and economic instruments to foster sustainable practices in the agricultural sector have been analysed within the framework of the EC (European Community) FP6 (funding program 6) INECO (Institutional and Economic Instruments for Sustainable Water Management in the Mediterranean Region) Project. Through stakeholder involvement processes, the project analysed issues relating to the regulation and control of groundwater abstractions, the promotion of wastewater reuse, efficiency improvements in irrigation water use, as well as instruments to strengthen the socioeconomic and institutional environment towards better managing and conserving available groundwater resources.     

Synergistic Effect of Atmospheric Pressure Plasma Pre-Treatment on Alkaline Etching of Polyethylene Terephthalate Fabrics and Films

Dyeing of PET materials by traditional methods presents several problems.Plasma technology has received enormous attention as a solution for the environmental problems related with textile surface modifications,and there has been a rapid development and commercialization of plasma technology over the past decade.In this work,the synergistic effect of atmospheric pressure plasma on alkaline etching and deep coloring of dyeing properties on polyethylene terephthalate(PET)fabrics and films was investigated.The topographical changes of the PET surface were investigated by atomic force microscopy(AFM)images,which revealed a smooth surface morphology of the untreated sample whereas a high surface roughness for the plasma and/or alkaline treated samples.The effects of atmospheric pressure plasma on alkaline etching of the structure and properties of PET were investigated by means of differential scanning calorimetry(DSC),the main objective of performing DSC was to investigate the effect of the plasma pre-treatment on the T_g and T_m.Using a tensile strength tester YG065 H and following a standard procedure the maximum force and elongation at maximum force of PET materials was investigated.Oxygen and argon plasma pre-treatment was found to increase the PET fabric weight loss rate.The color strength of PET fabrics was increased by various plasma pre-treatment times.The penetration of plasma and alkaline reactive species deep into the PET structure results in better dyeability and leaves a significant effect on the K/S values of the plasma pre-treated PET.It indicated that plasma pre-treatment has a great synergistic effect with the alkaline treatment of PET.     

Meta-heuristics for Feature Selection and Classification in Diagnostic BreastCancer

One of the most common kinds of cancer is breast cancer. The early detection of it may help lower its overall rates of mortality. In this paper, we robustly propose a novel approach for detecting and classifying breast cancer regions in thermal images. The proposed approach starts with data preprocessing the input images and segmenting the significant regions of interest. In addition, to properly train the machine learning models, data augmentation is applied to increase the number of segmented regions using various scaling ratios. On the other hand, to extract the relevant features from the breast cancer cases, a set of deep neural networks (VGGNet, ResNet-50, AlexNet, and GoogLeNet) are employed. The resulting set of features is processed using the binary dipper throated algorithm to select the most effective features that can realize high classification accuracy. The selected features are used to train a neural network to finally classify the thermal images of breast cancer. To achieve accurate classification, the parameters of the employed neural network are optimized using the continuous dipper throated optimization algorithm. Experimental results show the effectiveness of the proposed approach in classifying the breast cancer cases when compared to other recent approaches in the literature. Moreover, several experiments were conducted to compare the performance of the proposed approach with the other approaches. The results of these experiments emphasized the superiority of the proposed approach.     

Study of selected trace elements in cancerous and non-cancerous human breast tissues from Sudanese subjects using instrumental neutron activation analysis

This study was performed to investigate the influence of cancer on selected trace elements among Sudanese patients with confirmed breast cancer. Instrumental neutron activation analysis was used to estimate contents of Se, Zn, Fe, Cr, Rb, Cs, Co and Sc in 40 subjects. Wilcoxon signed ranks test was used to examine if there was any difference in the concentrations of elements from normal and malignant tissues. It was found that Se, Zn and Cr elements from the malignant tissues are significantly elevated (P < 0.05) compared to the normal tissue. The results obtained have shown consistency with results obtained by some previous studies.     

Investigation of the corrosion behavior of carbon steel coated with fluoropolymer thin films

Hydrophobic thin fluoropolymer films were deposited on carbon steel specimens using inductively coupled radio frequency plasma. Octafluorocyclobutane (C₄F₈) was used as gas source. The anticorrosion resistance of the deposited films in 3% sodium chloride aqueous solution was estimated using X-ray photoelectron spectroscopy (XPS), Fourier-transform infrared spectroscopy (FTIR), atomic force microscopy (AFM), scanning electron microscopy (SEM) as well as by electrochemical techniques. Results have shown that a 90 nm fluoropolymer film shows the optimal film thickness in terms of adhesion stability and corrosive resistance. Upon aging in 3% sodium chloride aqueous solutions for a week a decrease of the fluorine concentration together with a layer-by-layer delamination of the polymer coating was observed together with some coating oxidation. A thin polymer film of about 9 nm remained on the steel interface which proved to stay unaltered over a period of a month. This makes the fluoropolymer film interesting as protective layer of metal and metal oxides.     

Investigation of the corrosion protection of SiOx-like oxide films deposited by plasma-enhanced chemical vapor deposition onto carbon steel

The paper reports on the corrosion behavior of carbon steel coated with thin SiO_x-like oxide films. The SiO_x-like coatings were deposited by plasma-enhanced chemical vapor deposition (PECVD) and their thickness was varied between 20 and 200 nm. The coated carbon steel interfaces were investigated for their corrosion protection efficiency when immersed in an aqueous saline solution of 3% NaCl. FTIR measurements and electrochemical impedance spectroscopy (EIS) experiments revealed that thin SiO_x-like coating layers (20 nm thick) do not prevent the carbon steel from corrosion, while thicker silica layers (d ≥ 100 nm) protect efficiently carbon steel interfaces in highly saline media with a protection efficiency of about 96% for a 200 nm thick coating.     

Theoretical Study of the Ozone Production and NOx Molecules Processing in N2/O2 Pulsed Discharge

Environmental decontamination is a major challenge due to the rapid growth of industrial and technological development, requiring an important consumption of fossil energies. Nowadays, a new way to treat polluting molecules based on the use of nonequilibrium reactive plasmas, is under development. These plasmas are generated by electrical discharges at atmospheric pressure, for neutralizing or transforming the toxic oxides. In this respect, the goal of the present work is to analyze the possibilities of treatment of NO_x molecules and ozone generation by pulsed discharge. The study was carried out by using a one-dimensional model based on the parallel resistor network concept, consisting in dividing the discharge volume into plasma elements which are connected in no way but through their contribution to the total resistance of the plasma. The model calculations indicate the effect of gas heating and inhomogeneous preionization on the ozone production and NO_x destruction.