Search

It's been told from the beginning that the liver is one of the most important organs of our body function. Once upon a time, it couldn't see that a large number of people are suffering from liver diseases. But in recent years, the number of patients with liver diseases is increasing day by day. For this reason, affected people must go to a medical center for checking. In this paper, a model for the liver-affected people is implemented by which they do not need to go outside for checking the possibility of liver disease problems. To implement this model, datasets were collected based on some basic attributes which are related to liver diseases so that the possibility of liver problem can be detected. The data carries both liver-affected people and non-affected people. Those data were used to train our model so that the model can identify the affected people and non-affected people easily. Several machine learning algorithms were applied to generate results. The evaluation was done by following two approaches. Firstly, the complete result was generated using all of the attributes from both of the datasets where KNN provides the highest accuracy for both of the datasets which are 73% and 75.19% respectively. But after gone through an important attribute selection process, SVM gave the highest and increased accuracy which is about 82.68% and 81.15% respectively.

As per the report of the World Health Organization (WHO), diabetes has become one of the rapidly expanding chronic diseases that has affected the life of 422 million people all over the world. The number of deaths in Bangladesh due to diabetes has reached 28,065, which is 3.61% of the total deaths of Bangladesh, according to the latest data published by the WHO in 2018. So we need to be concerned about the risks of diabetes disease. If we cannot take proper steps to diagnose diabetes at an early stage, eventually we have to face serious health issues. In this paper, we have shown the relation of different symptoms and diseases that cause diabetes so that we can help a person to diagnose diabetes at an early stage. Nowadays, machine learning classification approaches are well accepted by researchers for developing disease risk prediction models. Therefore eleven machine learning classification algorithms such as Logistic Regression (LR), Gaussian Process (GP), Adaptive Boosting (AdaBoost), Decision Tree (DT), K-Nearest Neighbors (KNN), Multilayer Perceptron (MLP), Support Vector Machine (SVM), Bernoulli Naive Bayes (BNB), Bagging Classifier (BC), Random Forest (RF), and Quadratic Discriminant Analysis (QDA) have been used in this study. Among all these machine learning classifiers, Random Forest (RF) classifier has showed the best accuracy of 98%. And its Area Under Curve(AUC) is also the highest.

A low loss and highly sensitive gold nano-film externally coated dual core refractive index (RI) sensor using photonic crystal fiber (PCF) platform is proposed in this work. The perfectly matched layer (PML) is circled around to the external side of analyte to avoid the radiation of reflected power. The numerical analysis of the recommended sensor has been done in the analyte range from 1.33 to 1.42 in the $x$ -polarization by means of the finite element method (FEM). The maximum amplitude sensitivity, maximum wavelength sensitivity, and sensor resolution are 1102.57 $RIU^{-1}$ , 19000 nm/RIU, and $5.26\times 10^{-6}$ RIU, respectively, with maximum confinement loss as low as 0.173 dB/cm. So, due to the low loss and high sensitive characteristics, the proposed sensor can be used in detecting unknown chemical and analyte.

The pH level of Bryophillum Pinnatum leaves (BPL) is 4.25, which is a suitable amount of acid for generating electricity. When the mixture of water and the acidic juice of Bryophillum Pinnatum comes to contact with Zinc and Copper plates, forming an organic cell which generates a small amount of voltage. In this paper, improved efficiency can be observed by changing the gap between the electrodes and the mixture procedure. The surprising fact is that, after using the cell for a week, the output voltage level continues to decline. It can be improved again by reusing the previous BPL fiber and juice. This well-developed cell shows more efficiency than previous works. A cheap battery can be developed with a combination of this type of cells. Using this technology it is possible to provide electricity to the isolated community like Bandarban, Sandip of Bangladesh where free space is available for cultivating this plant. So, in this paper, electricity generation is not only the main concern but to develop a more efficient cell with maximum utilization of natural leaves. This environment-friendly, cheap and easy way will be a good solution for the isolated community.

Every year, a large number of people become lame due to a road accident and unable to walk normally. Wheelchair is the best assistive device used by the older and differently abled people who cannot walk normally. The driving and controlling of traditional manual wheelchair are much harder task. The modern wheelchairs like joystick-controlled and voice-controlled wheelchair is a bit difficult to operate for certain people like the older and physically weak people. Moreover, this wheelchair is not cost effective. So the aim of this paper is to make such a cost-effective electronic gesture-based wheelchair which will be easy to operate rather than the joystick input to control a wheelchair using in-built gesture function of a smartphone and touch sensor. ATMega328 used as the processor long with the L298N motor driver, DC Gear Motor, Ultrasonic Sensor, TTP224 Capacitive Touch Sensor, Bluetooth Module and IP Camera. Special features of this wheelchair are that obstacles on the way of the wheelchair can be detected which can avoid the collision between the wheelchair and that detected obstacle. Another feature of this chair is to use an IP camera that gives visual and acoustic information to the guardian of the riding person.

Now-a-days implantable device for biomedical applications are the interesting research area through all over the world due to the health care industry is highly announced to develop better efficient system which are more suitable and user friendly for patient. The main concern of the implantable device is to monitor various ailments within the body efficiently and then sending the corresponding monitoring signal to the server or base station. In that case, an efficient and miniaturized implantable antenna design is a great concern to ensure proper communication between implanted device and base station. In this paper, an implantable microstrip patch antenna (IMPA) has been proposed and characterized for implantable biomedical applications in the frequency of industrial scientific and medical (ISM, 2.4 GHz to 2.48 GHz) band. The total size of the proposed antenna is 12×12×2 mm3. Here patch of the antenna is designed on a dielectric substrate named Teflon having dielectric constant of 2.1 and it also have tangent loss of 0.00028. This proposed antenna is designed and simulated by using CST (Computer Simulation Technology) microwave studio 3D electromagnetic solver. The proposed IMPA shows good results such as lower return loss (−31.628 dB), better gain (−13.6 dB), wide bandwidth (727 MHz), VSWR (1.05) very close to unity, good impedance matching as well as miniature in size (volume of 288mm3).

Although the demand and generation of electric power have kept on increasing at comet's pace post-independence of our country, maintaining systems reliability has been an issue we are struggling with even today. With large grid blackouts taking place, there exists a possibility of saving at least the cities of economic and strategic importance through power system islanding and hence avoid facing huge intermittent economic losses during collapse of power systems. In this paper the analysis of the existing islands in our country and the practices followed in India and abroad during such a scenario to come up with a design and analysis of the possibility of islanding system installation for Bhopal, the capital city of Madhya Pradesh which is also under proposed consideration at PGCIL. The power flows and disturbance dynamics were done with real time simulation and an accepted stability was observed and results with plots were documented for the different cases consideration.

Image inpainting has been an area of interest for researchers in this digital image processing era. All the methods of image inpainting start with a selection of the area which is needed to be inpainted. However, this selection of area or mask is to be done manually as there is no way for a machine to identify it. The process of creation of mask is cumbersome and time consuming and requires human expertise. Using image segmentation, the image can be divided into probable areas of inpainting and then the user can select area/s to generate the mask accurately and easily. In this paper we present a method to do selection of segments to generate mask which is further used to inpaint the image using exemplar based inpainting algorithm. The proposed scheme greatly reduces the overall inpainting time by drastically reducing the manual work of drawing the mask. Comparative results of masking object manually and by using proposed method are shown. Subjective analysis of inpainting using proposed method of masking and manual masking shows the effectiveness of the same.

This demo paper describes the architecture and communication protocols — both single hop and multi-hop — for DSRC devices. The paper also describes an Android application that enables visualization of real-time vehicle movements on Google map using DSRC and Bluetooth communication. The application receives information about position, speed and direction of mobility that multiple vehicles obtain through the GPS Receiver attached to their DSRC OBU. The android application communicates with one of the DSRC units through Bluetooth to gather real-time traces collected from all DSRC-equipped vehicles. The application displays live movement of these vehicles on Google map with their path history, speed and direction. The source code and installation files of this application will be released through the Open Source Application Development Portal (OSADP) hosted by the U.S. Department of Transportation.

Inpainting has been an attractive and useful art in this fast growing era of digital image processing. Its applications in numerous fields such as restoration, object removal, superimposed text removal, red-eye removal etc. has made an impression on researchers to evolve generation of inpainting methods. After introduction of exemplar-based algorithm, several variations have been made in order to improve quality of inpainting region containing structure and texture information. Since inpainting starts from the border of the structure and propagates inward, selection of appropriate points on the structure becomes a very important criterion. Data term is responsible for extracting the structures in an image. With overview of exemplar algorithm, few variations in gradient based data terms are presented. Moreover, a data term based on sparsity of the image structure can also be an area of research. Evaluation of all these techniques is done using common analytical measures such as Peak Signal to Noise Ratio (PSNR), Similarity Structure Index Measure (SSIM) and Feature Similarity Index Measure (FSIM) in this paper. Besides qualitative comparison, subjective evaluation using Mean Opinion Score (MOS) is also presented.

For many years network operators have struggled to maintain fragile, statically configured and extremely complex networks. The constant threat of viruses, malware, intruders and misconfigured devices has made the task even more difficult. The use of an Intrusion Detection System (IDS) has become a standard defense model in many networks, however they are expensive and difficult to maintain and further complicate a network. This paper introduces a novel approach that integrates a distributed Intrusion Detection System into a Software Defined Network (SDN) and in doing so provides a more scalable security and threat management solution. The core mechanisms that enable SDN to provide an IDS function have been implemented and their performance evaluated. The viability of this approach was evaluated and found to be an effective alternative to the current IDS deployment model.

One of the great attractions of Software Defined Networking (SDN) has been the promise of transforming the network by providing a degree of flexibility and automation that until now has never been available. Of interest however is that all of these exciting solutions are usually discussed in isolation and rarely do proponents allude to the fact that numerous applications can be deployed simultaneously. Unfortunately most SDN Controllers are incapable of managing multiple applications that require access to the same events notifications. In this paper, we introduce the concept of an Event Arbitration Manager (EAM), a modified event handler that provides mediation between applications on a single SDN controller. We show that the introduction of an event mediation layer between SDN applications would reduce conflicts between multiple applications on the same SDN controller and allows them to respond to controller events messages equally.