Search

Sepsis is a life-threatening disease that is associated with organ dysfunction. It occurs due to the body’s dysregulated response to infection. It is difficult to identify sepsis in its early stages, this delay in identification has a dramatic effect on mortality rate. Developing prognostic tools for sepsis prediction has been the focus of various studies over previous decades. However, most of these studies relied on tracking a limited number of features, as such, these approaches may not predict sepsis sufficiently accurately in many cases. Therefore, in this study, we concentrate on building a more accurate and medically relevant predictive model for identifying sepsis. First, both NSGA-II (a multi-objective genetic algorithm optimization approach) and artificial neural networks are used concurrently to extract the optimal feature subset from patient data. In the next stage, a deep learning model is built based on the selected optimal feature set. The proposed model has two layers. The first is a deep learning classification model used to predict sepsis. This is a stacking ensemble of neural network models that predicts which patients will develop sepsis. For patients who were predicted to have sepsis, data from their first six hours after admission to the ICU are retrieved, this data is then used for further model optimization. Optimization based on this small, recent timeframe leads to an increase in the effectiveness of our classification model compared to other models from previous works. In the second layer of our model, a multitask regression deep learning model is used to identify the onset time of sepsis and the blood pressure at that time in patients that were predicted to have sepsis by the first layer. Our study was performed using the medical information from the intensive care MIMIC III real-world dataset. The proposed classification model achieved 0.913, 0.921, 0.832, 0.906 for accuracy, specificity, sensitivity, and AUC, respectively. In addition, the multitask regression model obtained an RMSE of 10.26 and 9.22 for predicting the onset time of sepsis and the blood pressure at that time, respectively. There are no other studies in the literature that can accurately predict the status of sepsis in terms of its onset time and predict medically verifiable quantities like blood pressure to build confidence in the onset time prediction. The proposed model is medically intuitive and achieves superior performance when compared to all other current state-of-the-art approaches.

Objectives: To measure the performance of docker swarm technology in virtual labs. Methods : The virtual laboratory is developed as a group of four system machines (VMs) on the same host computer as a cluster. The simulation depends on Linux OS, VirtualBox, Docker Swarm, Nginx, and Redis tools. Visualizing the tracing process by using portainer. Findings: The performance analysis of building virtual labs and running six main educational services using docker swarm virtualization technology are explained in detail. The experimental results have shown that the maximum utilization of the central processing unit (CPU) has reached 13% only for the nodes, 11% for the services, and 1% for the container, which considered very efficient in terms of processing. Moreover, the results have proved the effectiveness of the docker swarm in terms of memory usage since the maximum memory usage of nodes reached 101 MB, 103 MB for Container, and only 2% for each service. Additionally, the maximum network transition has reached (941 Bps) for service. Novelty/Applications: Building Cloud Virtual Labs enable students to connect remotely to the virtual machine at anytime and anywhere. Also, these labs enable instructors to trace the students’ progress and manage the evaluation process. Keywords: Container; cloud; docker; hypervisor; orchestration; swarm

Background/Objectives: In recent years, visualization systems were not only entertainment but also essential in training in our various fields. They do not depend on fixed devices only. They use visual systems, such as the headmounted display that Microsoft developed as the Mixed Reality (MR) HoloLens. Its features are equipped with engines that allow the user to interact with the headset via oral orders to communicate remotely with specialists in the surgical field in real-life situations. The main objective of the study is to use the 3D anatomical information models, the Digital Imaging and Communications in Medicine (DICOM) file and all the patient\'s data, its registration on Azure cloud computing system, to obtain the necessary training and support in case of encountering any emergency before and during the surgical planning. Method: This study presents the application as divided into two stages of anatomical simulation: training for local and international trainees through MR. The first stage classified (DICOM) files to the 3D model using the machine learning and HoloLens emulator of anatomy operational structure. The second stage involves Microsoft Azure and stores on cloud network by Data Lake, Azure Cosmo DB, and utilization of the Azure Spatial Anchors service to access the trainee to locate it at any time through the ID that is displayed by the IoT Sensor. Finding: This study examines Mixed Reality technology, HoloLens, and Head Mount Display to show the expected potential results to improve the surgeon\'s actions in surgery. This examination finished by 3D displaying anatomical models because they allow the surgeon to access the best solutions in real-life situations during the process to assess the three-dimensional holograms related to patient imaging or surgical techniques. Keywords: Mixed reality; 3D anatomical; HoloLens emulator; medical images; Azure spatial anchors; Azure cosmo DB

In the recent times, social media has become important in the field of health care as a major resource of valuable health information. Social media can provide massive amounts of data in real-time through user interaction, and this data can be analysed to reflect the harms and benefits of treatment by using the personal health experiences of users to improve health outcomes. In this study, we propose an enhanced data mining framework for analysing user opinions on Twitter and on a health-care forum. The proposed framework measures the degree of satisfaction of consumers regarding the drug Xeljanz, which is used to treat rheumatoid arthritis. The proposed framework is based on seven steps distributed in two phases. The first phase involves aggregating data related to the drug Xeljanz. This data is pre-processed to produce a list of words with a term frequency-inverse document frequency score. The word list is then classified into the following three categories: positive, negative and neutral. The second phase involves modelling social media posts using network analysis, identifying sub-graphs, calculating average opinions and detecting influential users. The results showed 77.3% user satisfaction with Xeljanz. Positive opinions were especially pronounced among users who switched to Xeljanz based on advice from a physician. Negative opinions of Xeljanz typically pertained to the high cost of the drug.

This paper presents the possibility of converting (2D) medical image data (Digital Imaging and Communications in Medicine (DICOM) files) to 3D model. Medical data and image processing software’s, namely Seg3D2 and ImageVis3D, were used to analyze images, create 3D models of the liver and export them in OBJ images for performing a range of surgical procedures, and measure the accuracy of the size and weight of the liver, kidneys and arteries with their conformity to DICOM file. It is compared to the image processing before and after the conversion stage of medical image using the Python language program to ensure the integrity of the images after the conversion process is identical to the original pictures of DICOM without causing any distortions or changes to it. We reduce file size while maintaining the model’s highest quality, while employing mixed reality techniques, applied on Liver Surgical Operation [living donor liver Transplantation (LDLT)].

The intensive care unit (ICU) admits the most seriously ill patients requiring extensive monitoring. Early ICU mortality prediction is crucial for identifying patients who are at great risk of dying and for providing suitable interventions to save their lives. Accordingly, early prediction of patients at high mortality risk will enable their provision of appropriate and timely medical services. Although various severity scores and machine-learning models have recently been developed for early mortality prediction, such prediction remains challenging. This paper proposes a novel stacking ensemble approach to predict the mortality of ICU patients. Our approach is more accurate and medically intuitive compared to the literature work. Data were prepared and feature selection was processed under the supervision of the ICU domain expert. The data were split into six modalities based on the expert's decisions. For the prediction process, a separate classifier was selected for each modality based on the performance of the classifiers. We utilized the most popular and diverse classifiers in the literature, including linear discriminant analysis, decision tree (DT), multilayer perceptron, k-nearest neighbor, and logistic regression (LR). Then, a stacking ensemble classifier was constructed and optimized based on the fusion of these five classifier decisions. The framework was evaluated using 10,664 patients from the medical information mart for intensive care (MIMIC III) benchmark dataset. To predict patient mortality, extensive experiments were conducted using the patients' time series data of different lengths. For each patient, the first 6, 12, and 24 hours of the first stay were tested. The results indicate that our model outperformed the state-of-the-art approaches in terms of accuracy (94.4%), F1 score (93.7%), precision (96.4%), recall (91.1%), and area under the receiver operator characteristic (ROC) curve (93.3%). These results demonstrate the ability and efficiency of our approach to predict ICU mortality.

Nowadays, due to easiness and expansion in property of smart mobile devices, it is becoming inevitable for mobile applications to have an important role in higher education systems. The Egyptian public universities are facing the problem of students’ large number enrolled in each year. Thus, we lack proper communication between educators and learners. Mobile learning can solve that problem, and it enables adjustment of the curriculum to meet students' learning time and life situations. It provides different solutions better than traditional educational methods. Students and professors could exchange educational material or information even if they are not in the same class. Furthermore, the cost of universities’ materials reduced, as all course materials can be found online through mobile applications. This paper proposes a mobile learning system named “Easy-Edu.” The proposed system intended to make the learning process easier, focus on students’ needs, and encourage communication and collaboration between students and professors and supports collaborative scenario-based learning for university students. Unlike other traditional systems, the proposed “Easy-Edu” was built using an Agile-based approach that delivers sustainable and high-quality mobile learning system. In addition, it eliminates the chances of absolute system failure and detects and fixes issues faster. Summarily, everything related to the design and implementation of “Easy-Edu” is discussed.

Liver cirrhosis, the end stage of chronic liver disease, is one of the major risk factors for the development of liver cancer, and may result in premature death. This research proposes a fuzzy fibrosis decision support (F2DS) system. It is a fuzzy knowledge-based expert system for liver fibrosis stage prediction. F2DS is carefully based on a set of knowledge acquisition and machine learning techniques. In addition, the system depends on domain expert knowledge for designing the membership functions and validating the fuzzy knowledge base. It depends on a suitable list of 17 symptoms, and laboratory test features that can accurately and significantly describe fibrosis patients. The experimental results of the expert system were obtained using a real dataset from the Liver Institute, Mansoura University, Egypt, of 119 patients infected by chronic viral hepatitis C. The performance of the system was evaluated with many metrics, achieving a testing accuracy of 95.7%. The evaluation of proposed fuzzy expert system shows its capability of diagnosing the stages of liver fibrosis with a high degree of accuracy, and it can be embedded as a component in a healthcare system to assist physicians in their daily practice. In addition, students training in medicine can benefit from this system.

Despite the extensive research of using web services for security purposes, there is a big challenge towards finding a no radical solution for NoSQL injection attack. This paper presents an independent RESTful web service in a layered approach to detect NoSQL injection attacks in web applications. The proposed method is named DNIARS. DNIARS depends on comparing the generated patterns from NoSQL statement structure in static code state and dynamic state. Accordingly, the DNIARS can respond to the web application with the possibility of NoSQL injection attack. The proposed DNIARS was implemented in PHP plain code and can be considered as an independent framework that has the ability for responding to different requests formats like JSON, XML. To evaluate its performance, DNIARS was tested using the most common testing tools for RESTful web service. According to the results, DNIARS can work in real environments where the error rate did not exceed 1%.

Voice calls is a popular way of communication between persons. Making voice call is cheap, fast and simple. Because of mobile networks attack or smartphones hackers, telephone conversations are vulnerable. This paper presents an encryption scheme for Voice Calls. It helps the user to encrypt the voice call before transmitting it on the mobile network. The idea of the proposed system is to encrypt voice calls without using any secure servers or any intermediate systems between mobile phone and the GSM network. The encryption process occurs before reaching to mobile phone. To maintain intensive security, an encryption algorithm based on RSA encryption is used. Moreover, it uses a public key known by all users and a private key which is a secret and restricted key to each user.

Early detection of cancer increases the probability of recovery. This paper presents an intelligent decision support system (IDSS) for the early diagnosis of cancer based on gene expression profiles collected using DNA microarrays. Such datasets pose a challenge because of the small number of samples (no more than a few hundred) relative to the large number of genes (in the order of thousands). Therefore, a method of reducing the number of features (genes) that are not relevant to the disease of interest is necessary to avoid overfitting. The proposed methodology uses the information gain (IG) to select the most important features from the input patterns. Then, the selected features (genes) are reduced by applying the grey wolf optimization (GWO) algorithm. Finally, the methodology employs a support vector machine (SVM) classifier for cancer type classification. The proposed methodology was applied to two datasets (Breast and Colon) and was evaluated based on its classification accuracy, which is the most important performance measure in disease diagnosis. The experimental results indicate that the proposed methodology is able to enhance the stability of the classification accuracy as well as the feature selection.

Nowadays, Healthcare is one of the most critical issues that need efficient and effective analysis. Data mining provides many techniques and tools that help in getting a good analysis for healthcare data. Data classification is a form of data analysis for deducting models. Mining on a reduced version of data or a lower number of attributes increases the efficiency of system providing almost the same results. In this paper, a comparative study between different data reduction techniques is introduced. Such comparison is tested against classification algorithms accuracy. The results showed that fuzzy rough feature selection outperforms rough set attribute selection, gain ratio, correlation feature selection and principal components analysis. General Terms Data mining, bioinformatics

messages are one of the popular ways of communication. Sending an SMS/MMS is cheap, fast and simple. Because of mobile networks attack or smartphones hackers, the GSM networks are not secure, so that all information or SMS/MMS messages are vulnerable. This paper describe an android application that helps the user to encrypt the message (SMS/Multimedia files) before it is transmitted over the mobile network. The new idea of the program is to transmit encrypted messages and multimedia files via mobile networks or the internet as an alternative mean. To maintain intensive security, the program uses a Hybrid encryption algorithm based on Blowfish and S-Boxes of DES encryption. Moreover, it uses a private key encrypts the files and another private key encrypts file name. The transferring media is maintained online in the absence of mobile network coverage.

Supervisory Control and Data Acquisition(SCADA) systems is one of important software systems which are used for monitoring and controlling industrial systems that are geographically spread over thousands of kilometers. These systems need to monitor and control so many field sites through thousands of devices that are varying in type, technology and usage. There are different types of people need to access SCADA systems for different purposes. Because of the sensitivity and spreading of these systems, they are vulnerable by hackers and crackers and there are many risks may causes partially or fully breakdown. To managing the SCADA systems, there are number of solutions that had been placed. These solutions varied from detecting one to more of SCADA system risk and assessed them on real system once it occurs. This way causes some damages could happen till the risk is eliminate or could need adaption that difficult or impossible to process. We propose in this paper a new framework for assessing and managing risks of the SCADA systems before they actually implemented by using one of risk management methodologies through scanning and testing proposed SCADA system architecture and its components.

A NoSQL injection attack targets interactive Web applications that employ NoSQL database services. These applications accept user inputs and use them to form query statements at runtime. During NoSQL injection attack, an attacker might provide malicious query segments as user input which could result in a different database request. In this paper, a testing tool is presented to detect NoSQL injection attacks in web application which is called “NoSQL Racket”. The basic idea of this tool depends on checking the intended structure of the NoSQL query by comparing NoSQL statement structure in code query statement (static code analysis) and runtime query statement (dynamic analysis). But we faced a big challenge, there is no a common query language to drive NoSQL databases like the same way in relational database using SQL as a standardized query language. The proposed tool is tested on four different vulnerable web applications and its effectiveness is compared against three different well known testers, none of them is able to detect any NoSQL Injection attacks. However, the implemented testing tool has the ability to detect the NoSQL injection attacks.

Software architects cannot avoid the consideration of quality attributes when designing software architecture. Architectural styles such as Layers and Client-Server are often used by architects to describe the overall structure and behavior of software. Although an architectural style affects the achievement of quality attributes, these quality attributes are directly performed by design decisions called architectural tactics. While the implementation of an architectural tactic supports a specific quality attribute, it often enhances or hurts other quality attributes in the software. In this paper, a framework for selecting the most appropriate architectural tactics according to their best achievement of the required levels of quality attributes when developing transaction processing systems is proposed. The proposed framework is based on fuzzy measures using Choquet Integral approach and takes into account the impact of architectural tactics on quality attributes, the preferences of quality attributes and the interactions between them. It can also be used to compare different potential architectures in terms of their supporting of quality attributes. The abilities and the advantages of the proposed framework are clarified via practical experiments using a case study.

As population is increasing rapidly all over the world, the need for delivering products is being more difficult especially for conditional products (products with life time). Many Customers require conditional products to be delivered to their locations. Distribution center may have multi depots (multi store branches) instead of one depot. Every depot has limited number of vehicles to minimize cost. Capacities of these vehicles are based on two dimensions (weight and volume). Geographic information system (GIS) is used for localizing customers’ destinations. Then OD Cost Matrix is used to assign every customer destination to the least cost depot to be served from it. Finally Network analyst is used to solve the vehicle routing problem generating final route directions for every vehicle and calculating the best time for lunch break of drivers automatically. This case study is applied on Mansoura city in Egypt.

This paper presents a proposed Adaptive Online Lecture Model that presents different pedagogical aspects fo r re commending th e most suit able lea rning materials for s tudents based o n th eir learning profiles and preferences, involving students in the learni ng process from the very early beginning of the lecture, a nd p reparing for th e next /upcoming lect ure, so stud ents feel the personalization an d customization of the lec ture, and h opefully this model en hances the learning pr ocess and students’ online learning experience.

Concatenative speech synthesis systems generate speech by concatenating small prerecorded speech units which are stored in the speech unit inventory. The most commonly used type of these units is the diphone which is a unit that starts at the middle of one phone and extends to the middle of the following one. Diphones have the advantage of modeling coarticulation by including the transition to the next phone inside the diphone itself. In this paper, a diphone speech synthesis system for the Arabic language using MARY TTS has been developed and evaluated by two types of tests which are the Diagnostic Rhyme Test (DRT) that measures the intelligibility of the synthesized speech and the Categorical Estimation (CE) test that measures the overall quality of the synthesized speech. The results of these tests are illustrated in the experiments and results section.

Fast packet detection is very important to overcome intrusion attack. In this paper, a new approach for fast packet detection in serial data sequence is presented. Such algorithm uses fast time delay neural networks (FTDNNs). The operation of these networks relies on performing cross correlation in the frequency domain between the input data and the input weights of neural networks. It is proved mathematically and practically that the number of computation steps required for the presented FTDNNs is less than that needed by conventional time delay neural networks (CTDNNs). Simulation results using MATLAB confirm the theoretical computations.

Achieving the desired levels of quality attributes is inevitable when developing software. In reality, software engineers take into account stakeholders' preferences of quality attributes when developing software. Furthermore, it is not practical to perform the needed level of each quality attribute individually without considering its interaction with other quality attributes in the potential system. While the conventional aggregations methods such as weighted arithmetic mean consider the preferences among criteria, they do not take into consideration the interactions between them. To accomplish this, a method of fuzzy measures using Choquet Integral can be utilized. Choquet Integral method also has an advantage of helping decision makers in providing insights about interactions among quality attributes. It can define if two quality attributes interplay in complementary way or in redundancy way. In this paper we utilized Choquet Integral approach to investigate the preferences and the interactions of quality attributes when developing transaction processing information systems. The investigation results are drawn based on analyzing the impact of architectural tactics on quality attributes when building software architecture.

Evaluating software quality is an important and essential issue in the development process because it helps to deliver a competitive software product. A decision of selecting the best software based on quality attributes is a type of multi-criteria decision-making (MCDM) processes where interactions among criteria should be considered. This paper presents and develops quantitative evaluations by considering interactions among criteria in the MCDM problems. The aggregator methods such as Arithmetic Mean (AM) and Weighted Arithmetic Mean (WAM) are introduced, described and compared to Choquet Integral (CI) approach which is a type of fuzzy measure used as a new method for MCDM. The comparisons are shown by evaluating and ranking software alternatives based on six main quality attributes as identified by the ISO 9126-1 standard. The evaluation experiments depend on real data collected from case studies.

In this paper, a new model for adaptive e-learning based on learner's styles is presented. In the previous work, the dimensions of learner's styles given by Felder-Silverman did not consider some important issues of the learner himself. Here, new learner's parameters such as his social environment, health conditions, psychological and economical states are taken into account. Such parameters greatly affect the ability of student to learn and understand. Therefore, in order to perfectly recognize the ability of the student to be interactive in the leaning environment and accept information, new learner's styles are added to the dimensions of Felder-Silverman learning style model and our previous work [24]. The new proposed model is applied for logic gates and functions used in data encoding and computer networks. Such model presents suitable courses for each student in a dynamic and adaptive manner using existing database and workflow technologies. Furthermore, it is powerful, user friendly and easy to interpret. Moreover, it suggests a learning strategy and appropriate electronic media that match the learner’s preference.

In this paper, a new fast neural model for testing massive volume of medical data is presented. The idea is to accelerate the process of detecting and classifying pediatric respiratory diseases by using neural networks. This is done by applying cross correlation between the input patterns and the input weights of neural networks in the frequency domain rather than time domain. Furthermore, such model is very useful for understanding the internal relation between the medical patterns. In addition, the input patterns are collected in one vector and manipulated as a one pattern. Moreover, before training neural networks, rough sets are used to reduce the length of the feature input vector. The most important feature elements are used to train the neural networks. The reduced input medical patterns are classified to one of eight diseases. Simulation results confirm the theoretical considerations as 98% of all tested cases are classified correctly. The presented model can be applied successfully for any other classification application.

Distributed database security has become an important issue. In this paper, a new computational method for protecting the distributed databases is presented. Such approach is applied for protecting both textual and numerical data. This is done by adding only one hidden record with a secret function. For each attribute, the value of this function depends on the data stored in all other records. Therefore, this technique is more powerful against any attacks or modifications such as deleting or updating cell values. Furthermore, the problems associated with the work in literature are solved. For example, there is no need for additional storage area as required when adding additional columns especially with large databases. In addition, in case of protecting data by adding columns, we need to add a number of columns equal to the number of data types to be protected. Here, only one record is sufficient to protect all types of data. Another advantage is that, there is a possibility to use a different function for each field results in more robustness. Moreover, a real-time intrusion detection algorithm is introduced for fast attack detection. Finally, the proposed technique does not have any other requirements or restrictions on either database design or database administration.

In this paper, a new approach for protecting the ownership of relational database is presented. Such approach is applied for protecting both textual and numerical data. This is done by adding only one hidden record with a secret function. For each attribute, the value of this function depends on the data stored in all other records. Therefore, this technique is more powerful against any attacks or modifications such as deleting or updating cell values. Furthermore, the problems associated with the work in literature are solved. For example, there is no need for additional storage area as required when adding additional columns especially with large databases. In addition, in case of protecting data by adding columns, we need to add a number of columns equal to the number of data types to be protected. Here, only one record is sufficient to protect all types of data. Moreover, there is a possibility to use a different function for each field results in more robustness. Finally, the proposed technique does not have any other requirements or restrictions on either database design or database administration.

This paper presents a new approach to speed up the operation of time delay neural networks for detecting a record in databases. The entire data are collected together in a long vector and then tested as a one input pattern. The proposed fast time delay neural networks (FTDNNs) use cross correlation in the frequency domain between the tested data and the input weights of neural networks. It is proved mathematically and practically that the number of computation steps required for the presented time delay neural networks is less than that needed by conventional time delay neural networks (CTDNNs). Simulation results using MATLAB confirm the theoretical computations.

In recent years, time delay neural networks are successfully used in many applications. Here, a new idea to speed the operation of time delay neural networks is presented. The whole data are collected together in a long vector and then tested as a one input pattern. The proposed fast time delay neural networks uses cross correlation in the frequency domain between the tested data and the input weights of neural networks. It is proved mathematically that the number of computation steps required for the presented fast time delay neural networks is less than that needed by classical time delay neural networks. Simulation results after these corrections using MATLAB confirms the theoretical computations.

This paper presents a new approach to speed up the operation of time delay neural networks. The entire data are collected together in a long vector and then tested as a one input pattern. The proposed fast time delay neural networks (FTDNNs) use cross correlation in the frequency domain between the tested data and the input weights of neural networks. It is proved mathematically and practically that the number of computation steps required for the presented time delay neural networks is less than that needed by conventional time delay neural networks (CTDNNs). Simulation results using MATLAB confirm the theoretical computations.

Single-layer training for high speed English capital or small letters recognition is presented. A new approach to the hardware implementation of the artificial processing element (PE) and control circuits with learning is introduced. The programmable synaptic weights are computed during the training period by a software program. The proposed learning algorithm is very fast and significant in many ways. The results are computed in real time and appear to be perfect. This system is very suitable for analog-digital VLSI implementation.