Search

In this study, a computer-aided detection system was developed for detection and investigation of lung nodules in computed tomography images using image processing techniques. The computer aided detection system consists of four stages. First and second stages are preprocessing stages. First preprocessing stage is two-dimensional preprocessing stage and second preprocessing stage is three-dimensional preprocessing stage. Third stage of the developed system is the feature extraction stage. Five different groups of features are extracted from volume of interests in this stage. Last stage is the the nodule detection stage. The support vector machine algorithm is optimized using evolutionary algorithms to classify volume of interests using the features. The computer aided detection system achieves 93.76% sensitivity, 81.69% selectivity, 84.52% accuracy and 3.63 false positive per scan using only morphologic, statistical and histogram based features. After the inclusion of groups of outer surface statistical and outer surface GLCM and Gabor filter based textural features, performance rates of the computer aided detection system reaches 98.35% sensitivity, 90.42% selectivity, 92.28% accuracy and 2.33 false positive per scan. Results of experiments reveal that outer surface statistical and textural features are useful to increase sensitivity of the system. These features also decrease the number of false positives of the developed systemi

Intelligent systems change familiar objects as well as rapidly change common technologies. Internet of Things (IoT) applications are made with Cloud Computing can not meet the expectation of service quality due to bandwidth and network delays. It focuses on solving these needs in real-time intelligent applications that use Edge Computing, which is introduced as a new platform between the cloud and the objects. Security and privacy concerns are a major obstacle to the widespread use of intelligent objects. Edge Computing has both its proximity to objects and sufficient resources are available to provide an environment for the operation of security applications. In this study, Edge Computing is introduced. Edge Computing Security Aplication Kilic is recommend, which works on the Edge Layer. To illustrate the effective use of the Kilic, the real-time industry application has been shown in the context of the Smart Factory.

The purpose of this paper was to develop a novel actual time cyber security approach to smart grids. The developed way securely backed the smart meters among smart grids against external attacks in a distribution substation. Security is provided using Message Passing Interface (MPI). Newly configured Floyd-Warshall Algorithm was applied to MPI so as to represent a distribution of key pairs which was generated by Homomorphic Encryption. Simulation case studies were conducted on Raspberry Pi 2 Model B in order to illustrate how the mentioned approach can be built to execute the cyber security essentials via parallel programming techniques and Valgrind tools on a virtual distributed mechanism. Simulation results which use real world attack data demonstrated that the developed way is capable of accomplishing cyber security against outer attacks accurately, quickly and real time based sorting of the key pairs into each smart meter node in the demanded area.

WOS: 000284213700008, The goal of this study was to develop an efficient clustering framework for processing high-dimensional datasets with reasonable memory and computing power requirements. Strehl and Ghosh proposed a novel clustering approach and developed a framework which is called "relationship-based clustering framework" [1]. In this study, a preprocessing system has been implemented on top of their approach and it has been integrated into the relationship-based clustering framework. Three different benchmark datasets were used to evaluate its efficiency. The results are presented in various tables and charts, and in addition CLUSION graphs are plotted to enable visual evaluation of cluster quality. It is demonstrated that CPU and memory usage has been substantially decreased compared with Strehl and Ghosh's framework 1, without any noticeable decrease in clustering quality. This fact enables the use of the relationship-based clustering framework for much larger datasets than was heretofore possible, and also increases its scalability with respect to number of dimensions., Marmara University [FEN-DKR-270306-0055], The research for this article was supported by grant number FEN-DKR-270306-0055 from Marmara University Scientific Research Projects Committee (BAPKO).

Developing the system which will help doctors with the result to be obtained from the medical data sets by realizing the design of the medical decision support system in which data mining methods are used is the primary objective of this study. The survivability condition of metastatic colorectal cancer disease has been predicted using data mining methods in the developed system. In the process of data mining, after the phase of data preprocessing, Support Vector Machines, Naive Bayes, Decision Trees, Artificial Neural Networks, Multilayer Perceptron, Logistic Regression algorithms have been used. In the study, two different medical decision support system models, classifying prediction and hybrid models, have been developed, and the results obtained from the two models have been compared after being examined. While the most successful algorithm is Support Vector Machines in Classifying Prediction Model in which only classification algorithms are applied, it is Decision Trees and Artificial Neural Networks which are the most successful algorithms with an accuracy rate of 100 per cent in Hybrid Prediction Model. In consequence of the classifying processes, when the accuracy rates of the models are examined, it is seen that while the accuracy rate of Classifying Prediction Model is 65-70%, this rate reaches 95-100% in Hybrid Prediction Model. It has been seen that accuracy rates have elicited very high and very close values for all of the algorithms with the realized hybrid structure, that is, with clustering, in the applications of the classifying algorithms combined.

In this study, a computer-aided detection (CAD) system was developed for the detection of lung nodules in computed tomography images. The CAD system consists of four phases, including two-dimensional and three-dimensional preprocessing phases. In the feature extraction phase, four different groups of features are extracted from volume of interests: morphological features, statistical and histogram features, statistical and histogram features of outer surface, and texture features of outer surface. The support vector machine algorithm is optimized using particle swarm optimization for classification. The CAD system provides 97.37% sensitivity, 86.38% selectivity, 88.97% accuracy and 2.7 false positive per scan using three groups of classification features. After the inclusion of outer surface texture features, classification results of the CAD system reaches 98.03% sensitivity, 87.71% selectivity, 90.12% accuracy and 2.45 false positive per scan. Experimental results demonstrate that outer surface texture features of nodule candidates are useful to increase sensitivity and decrease the number of false positives in the detection of lung nodules in computed tomography images.

Computer-aided detection (CAD) systems help to radiology experts in mass detection using image processing techniques. This study aims to realize mass detection process on brain MRI (Magnetic Resonance Imaging) images. This paper describes the CAD system that is based on histogram equalization and morphological image processing techniques. The processes are carried out through 125 MR images which are taken from 11 people that are 8 people with tumors, 3 people without tumors. In classification stage which is the last stage of the computer-aided detection systems, 6 classification algorithms are tested in the RapidMiner program, and these algorithms are compared with each other to show CAD system accuracy.

In this study, we developed a computer aided detection system (CAD) to detect lung nodules on computed tomography images. Template matching and similarity measurement methodsused for examine whether region of interest which extracted using image pre-processing techniques are nodule candidate. Intensity thresholding, distance thresholding, neighbourhood analysis are preprocessing techniques of the developed CAD system. Pearson's correlation coefficient, simple matching coefficient, Jaccard's coefficient, Euclidean Distance and Sokal & Sneath similarity coeeficient calculated to measure similarity between nodule candidate and the template. Sensitivity of the CAD and number of false positives per slice are given in conclusion.