COVID-19 Infected Lung Computed Tomography Segmentation and Supervised Classification Approach

The purpose of this research is the segmentation of lungs computed tomography (CT) scan for the diagnosis of COVID-19 by using machine learning methods. Our dataset contains data from patients who are prone to the epidemic. It contains three types of lungs CT images (Normal, Pneumonia, and COVID-19) collected from two different sources; the first one is the Radiology Department of Nishtar Hospital Multan and Civil Hospital Bahawalpur, Pakistan, and the second one is a publicly free available medical imaging database known as Radiopaedia. For the preprocessing, a novel fuzzy c-mean automated region-growing segmentation approach is deployed to take an automated region of interest (ROIs) and acquire 52 hybrid statistical features for each ROIs. Also, 12 optimized statistical features are selected via the chi-square feature reduction technique. For the classification, five machine learning classifiers named as deep learning J4, multilayer perceptron, support vector machine, random forest, and naive Bayes are deployed to optimize the hybrid statistical features dataset. It is observed that the deep learning J4 has promising results (sensitivity and specificity: 0.987; accuracy: 98.67%) among all the deployed classifiers. As a complementary study, a statistical work is devoted to the use of a new statistical model to fit the main datasets of COVID-19 collected in Pakistan.     

Race Classification Using Deep Learning

Race classification is a long-standing challenge in the field of face image analysis. The investigation of salient facial features is an important task to avoid processing all face parts. Face segmentation strongly benefits several face analysis tasks, including ethnicity and race classification. We propose a race-classification algorithm using a prior face segmentation framework. A deep convolutional neural network (DCNN) was used to construct a face segmentation model. For training the DCNN, we label face images according to seven different classes, that is, nose, skin, hair, eyes, brows, back, and mouth. The DCNN model developed in the first phase was used to create segmentation results. The probabilistic classification method is used, and probability maps (PMs) are created for each semantic class. We investigated five salient facial features from among seven that help in race classification. Features are extracted from the PMs of five classes, and a new model is trained based on the DCNN. We assessed the performance of the proposed race classification method on four standard face datasets, reporting superior results compared with previous studies.     

Robot trajectory optimisation with dynamic constraints

Objects located within a robot arm workspace fragment this space, since a path obstructed by an object is non-viable. Point-to-point motion of a robot arm in such a fragmented workspace has to be carefully planned in order to avoid collision. The problem of planning the trajectory of the robot arm can be represented as a constrained optimisation problem from which an optimal time can be deduced. The solution of the optimisation problem for an optimal continuous trajectory is complex and costly. However, a motion strategy, whereby the robot arm path is formed of single linear segments, renders the optimisation problem solvable and produces a sub-optimal but practical trajectory. This paper describes a motion strategy that is based on the concept of defining via intermediate and control points which the robot arm has to pass by along its path, and takes into consideration the torque limitation of the different axis drives. Joint motion planning is based on a function of time approximated by a sequence of polynomials with boundary conditions that would result in continuous motion at the via points. Motion between each two points is divided into two transition zones and a constant speed zone in between. A result is obtained by selecting polynomial coefficients that would minimise, subject to the constraints of the robot dynamics, the time required to accelerate and decelerate the joint to the appropriate speeds at the boundary of these zones. The joint is allowed to move at the maximum speed for the remainder of the time.     

Cluster-Based RF Fingerprint Positioning Using LTE and WLAN Signal Strengths

Wireless Local Area Network (WLAN) positioning has become a popular localization system due to its low-cost installation and widespread availability of WLAN access points. Traditional grid-based radio frequency (RF) fingerprinting (GRFF) suffers from two drawbacks. First it requires costly and non-efficient data collection and updating procedure; secondly the method goes through time-consuming data pre-processing before it outputs user position. This paper proposes Cluster-based RF Fingerprinting (CRFF) to overcome these limitations by using modified Minimization of Drive Tests data which can be autonomously collected by cellular operators from their subscribers. The effect of environmental changes and device variation on positioning accuracy has been carried out. Experimental results show that even under these variations CRFF can improve positioning accuracy by 15.46 and 22.30% in 95 percentile of positioning error as compared to that of GRFF and K-nearest neighbour methods respectively.     

MODELING THE HYDROCRACKING PROCESS USING ARTIFICIAL NEURAL NETWORKS

Feed-forward neural networks that models the hydrocracking process of Arabian light vacuum gas oil are presented. The input-output data to the neural networks was obtained from actual local refineries. Several network architectures were tried and the networks that best simulate the hydrocracking process were retained. The networks are able to predict yields and properties of products of the hydrocracking unit (e.g. iC₄, nC₄, light and heavy naphtha, light and heavy ATK, Diesel, etc.). The predictions of yields and properties of various desired and undesired products at different conditions are required by refineries for process optimization, control, design, catalyst selection, and planning. The predictions of the prepared neural networks have been cross validated against data not originally used in the training process. The networks compared well against this new set of data with an average percent error always less than 8.71 for the different products of the hydrocracking unit.     

Performance analysis of hybrid/single nanofluids on augmentation of heat transport in lid-driven undulated cavity

This numerical study reveals the heat transfer performance of hybrid/single nanofluids inside a lid-driven sinusoidal trapezoidal-shaped enclosure. The right and left inclined surfaces of the trapezium have been considered as insulated, whereas the bottom sinusoidal wavy and the flat top surfaces of the enclosure as hot and cold, respectively. The governing partial differential equations of fluid's velocity and temperature have been resolved by applying the finite element method. The implications of Prandtl number (4.2-6.2), Richardson number (0.1-10.0), undulation number (0-3), nanoparticles volume fraction (0%-3%), and nanofluid/base fluid (water, water–copper (Cu), water–Cu–carbon nanotube, water–Cu–copper oxide (CuO), water–Cu–TiO₂, and water–Cu–Al₂O₃) on the velocity and temperature profiles have been studied. Simulated findings have been represented by means of streamlines, isothermal lines, and average Nusselt number of above-mentioned hybrid nanofluids for varying the governing parameters. The comparison of heat transfer rates using hybrid nanofluids and pure water has been also shown. The heat transfer rate is increased about 15% for varying Richardson number from 0.1 to 10.0. Blending of two nanoparticles suspension in base fluid has a higher heat transfer rate—approximately 5% than a mononanoparticle. Moreover, a higher average Nusselt number is obtained by 14.7% using the wavy surface than the flat surface of the enclosure. Thus, this study showed that applying hybrid nanofluid may be beneficial to obtain expected thermal performance.     

Occurrence of infected amoebae in cooling towers compared with natural aquatic environments: implications for emerging pathogens

Many species of bacteria pathogenic to humans, such as Legionella, are thought to have evolved in association with amoebal hosts. Several novel unculturable bacteria related to Legionella have also been found in amoebae, a few of which have been thought to be causes of nosocomial infections in humans. Because amoebae can be found in cooling towers, we wanted to know whether cooling tower environments might enhance the association between amoebae and bacterial pathogens of amoebae in order to identify potential "hot spots" for emerging human pathogens. To compare occurrence of infected amoebae in natural environments with those in cooling towers, 40 natural aquatic environments and 40 cooling tower samples were examined. Logistic regression analysis determined variables that were significant predictors of the occurrence of infected amoebae, which were found in 22 of 40 cooling tower samples but in only 3 of the 40 natural samples. An odds ratio showed that it is over 16 times more likely to encounter infected amoebae in cooling towers than in natural environments. Environmental data from cooling towers and natural habitats combined revealed dissolved organic carbon (DOC) and pH were predictors of the occurrence of the pathogens, however, when cooling tower data alone were analyzed, no variables accounted for the occurrence. Several bacteria have novel rRNA sequences, and most strains were not culturable outside of amoebae. Such pathogens of amoebae may spread to the environment via aerosols from cooling towers. Studies of emerging infectious diseases should strongly consider cooling towers as a source of amoeba-associated pathogens.     

Effect of Al Doping on Structural, Electrical, Optical and Photoluminescence Properties of Nano-Structural ZnO Thin Films

The nano-structural Al-doped ZnO thin films of different morphologies deposited on glass substrate were successfully fabricated at substrate temperature of 350 C by an inexpensive spray pyrolysis method. The structural, electrical, optical and photoluminescence properties were investigated. X-ray diffraction study revealed the crystalline wurtzite (hexagonal) structure of the films with nano-grains. Scanning electron mi- croscopy (SEM) micrographs indicated the formation of a large variety of nano-structures during film growth. The spectral absorption of the films occurred at the absorption edge of ～410 nm. In the present study, the optical band gap energy 3.28 eV of ZnO decreased gradually to 3.05 eV for 4 mol% of Al doping. The deep level activation energy decreased and carrier concentrations increased substantially with increasing doping. Exciting with the energy 3.543 eV (λ=350 nm), a narrow and a broad characteristic photoluminescence peaks that correspond to the near band edge (NBE) and deep level emissions (DLE), respectively emerged.