Clinical keratoconus (KCN) detection is a challenging and time-consuming task. In the diagnosis process, ophthalmologists must revise demographic and clinical ophthalmic examinations. The latter include slit-lamb, corneal topographic maps, and Pentacam indices (PI). We propose an Ensemble of Deep Transfer Learning (EDTL) based on corneal topographic maps. We consider four pretrained networks, SqueezeNet (SqN), AlexNet (AN), ShuffleNet (SfN), and MobileNet-v2 (MN), and fine-tune them on a dataset of KCN and normal cases, each including four topographic maps. We also consider a PI classifier. Then, our EDTL method combines the output probabilities of each of the five classifiers to obtain a decision based on the fusion of probabilities. Individually, the classifier based on PI achieved 93.1% accuracy, whereas the deep classifiers reached classification accuracies over 90% only in isolated cases. Overall, the average accuracy of the deep networks over the four corneal maps ranged from 86% (SfN) to 89.9% (AN). The classifier ensemble increased the accuracy of the deep classifiers based on corneal maps to values ranging (92.2% to 93.1%) for SqN and (93.1% to 94.8%) for AN. Including in the ensemble-specific combinations of corneal maps’ classifiers and PI increased the accuracy to 98.3%. Moreover, visualization of first learner filters in the networks and Grad-CAMs confirmed that the networks had learned relevant clinical features. This study shows the potential of creating ensembles of deep classifiers fine-tuned with a transfer learning strategy as it resulted in an improved accuracy while showing learnable filters and Grad-CAMs that agree with clinical knowledge. This is a step further towards the potential clinical deployment of an improved computer-assisted diagnosis system for KCN detection to help ophthalmologists to confirm the clinical decision and to perform fast and accurate KCN treatment.
The major of DDoS attacks use TCP protocol and the TCP SYN flooding attack is the most common one among them. The SYN Cookie mechanism is used to defend against the TCP SYN flooding attack. It is an effective defense, but it has a disadvantage of high calculations and it doesn’t differentiate spoofed packets from legitimate packets. Therefore, filtering the spoofed packet can effectively enhance the SYN Cookie activity. Hop Count Filtering (HCF) is another mechanism used at the server side to filter spoofed packets. This mechanism has a drawback of being not a perfect and final solution in defending against the TCP SYN flooding attack. An enhanced mechanism of Integrating and combining the SYN Cookie with Hop Count Filtering (HCF) mech
... Show MoreThe diagnosis of acute appendicitis (AA) sometimes is illusive and the accompanying clinical and laboratory manifestations cannot be used for definitive diagnosis. Objective: This study aimed to evaluate the diagnostic value of neutrophil/lymphocyte ratio (NLR) in detection of AA. Materials and Methods: This is a cross-sectional study that included a total of 80 adult patients with AA and 62 age- and gender-matched patients with abdominal pain due to causes other than AA. Three milliliter of peripheral blood were collected from each participant. The NLR was calculated by dividing the absolute neutrophil count by the absolute lymphocyte count. Receiver operating characteristic curve was used to assess the diagnostic value of NLR in detection
... Show MoreCorona virus sickness has become a big public health issue in 2019. Because of its contact-transparent characteristics, it is rapidly spreading. The use of a face mask is among the most efficient methods for preventing the transmission of the Covid-19 virus. Wearing the face mask alone can cut the chance of catching the virus by over 70\%. Consequently, World Health Organization (WHO) advised wearing masks in crowded places as precautionary measures. Because of the incorrect use of facial masks, illnesses have spread rapidly in some locations. To solve this challenge, we needed a reliable mask monitoring system. Numerous government entities are attempting to make wearing a face mask mandatory; this process can be facilitated by using face m
... Show MorePlagiarism is becoming more of a problem in academics. It’s made worse by the ease with which a wide range of resources can be found on the internet, as well as the ease with which they can be copied and pasted. It is academic theft since the perpetrator has ”taken” and presented the work of others as his or her own. Manual detection of plagiarism by a human being is difficult, imprecise, and time-consuming because it is difficult for anyone to compare their work to current data. Plagiarism is a big problem in higher education, and it can happen on any topic. Plagiarism detection has been studied in many scientific articles, and methods for recognition have been created utilizing the Plagiarism analysis, Authorship identification, and
... Show MoreFor several applications, it is very important to have an edge detection technique matching human visual contour perception and less sensitive to noise. The edge detection algorithm describes in this paper based on the results obtained by Maximum a posteriori (MAP) and Maximum Entropy (ME) deblurring algorithms. The technique makes a trade-off between sharpening and smoothing the noisy image. One of the advantages of the described algorithm is less sensitive to noise than that given by Marr and Geuen techniques that considered to be the best edge detection algorithms in terms of matching human visual contour perception.
Goal of research is to investigate the impact of the use of effective learning model in the collection of the fourth grade students/Department of physics in the material educational methods and the development of critical thinking .to teach this goal has been formulated hypothesis cefereeten zero subsidiary of the second hypothesis .To investigate the research hypothesis were selected sample of fourth-grade students of the department of physics at the univers
... Show MoreAbstract
The current study presents numerical investigation of the fluid (air) flow characteristics and convection heat transfer around different corrugated surfaces geometry in the low Reynolds number region (Re<1000). The geometries are included wavy, triangle, and rectangular. The effect of different geometry parameters such as aspect ratio and number of cycles per unit length on flow field characteristics and heat transfer was estimated and compared with each other. The computerized fluid dynamics package (ANSYS 14) is used to simulate the flow field and heat transfer, solve the governing equations, and extract the results. It is found that the turbulence intensity for rectangular extended surface was larg
... Show MoreTheoretical and experimental investigations have been carried out on developing laminar
combined free and forced convection heat transfer in a vertical concentric annulus with uniformly
heated outer cylinder (constant heat flux) and adiabatic inner cylinder for both aiding and opposing
flows. The theoretical investigation involved a mathematical modeling and numerical solution for
two dimensional, symmetric, simultaneously developing laminar air flows was achieved. The
governing equations of motion (continuity, momentum and energy) are solved by using implicit
finite difference method and the Gauss elimination technique. The theoretical work covers heat flux
range from (200 to 1500) W/m2, Re range from 400 to 2000 an