Araştırma Çıktıları | WoS | Scopus | TR-Dizin | PubMed
Permanent URI for this communityhttps://hdl.handle.net/20.500.14719/1741
Browse
6 results
Search Results
Publication Metadata only Kaleidomap visualizations of cardiovascular function in critical care medicine(2006) Bale, Kim; Chapman, Paul; Purdy, Jon H.; Aydın, Nizamettin; Dark, Paul Michael; Bale, Kim, Department of Computer Science, University of Hull, Hull, United Kingdom; Chapman, Paul, Department of Computer Science, University of Hull, Hull, United Kingdom; Purdy, Jon H., Department of Computer Science, University of Hull, Hull, United Kingdom; Aydın, Nizamettin, Department of Genetics and Bioinformatics, Bahçeşehir Üniversitesi, Istanbul, Turkey; Dark, Paul Michael, Intensive Care Research Group, The University of Manchester, Manchester, United KingdomIn this paper we consider how the use of Kaleidomaps can facilitate our understanding and interpretation of large complex multivariate medical datasets relating to cardiovascular function in critical care medicine. Kaleidomaps are a new technique for the visualization of multivariate time-series data. They build upon the classic cascade plot and use the curvature of a line to enhance the detection of periodic patterns within multivariate dual-periodicity datasets. Kaleidomaps keep user interaction to a minimum, facilitating the rapid identification of periodic patterns not only within their own variants but also across many different sets of the variants. By linking this technique with traditional line graphs and signal processing techniques, we are able to provide medical experts with a set of visualization tools that permit the combination of medical datasets in their raw form and also with the results of mathematical analysis. © 2006 IEEE. © 2008 Elsevier B.V., All rights reserved.Publication Metadata only Data mining usage in emboli detection(2007) Karahoca, Adem; Kucur, Turkalp; Aydın, Nizamettin; Karahoca, Adem, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Kucur, Turkalp, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Aydın, Nizamettin, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyAsymptomatic circulating cerebral emboli, which are particles bigger than blood cells, can be detected by transcranial Doppler ultrasound. In certain conditions asymptomatic embolic signals (ES) appear to be markers of increased stroke risk. ES, reflected by an embolus, have usually larger amplitude than the signals from normal blood flow and show a transient characteristic. A number of methods to detect cerebral emboli have been studied in the literature. In this study, data mining techniques have been used in order to increase sensitivity and specificity of an embolic signal detection system. The classification results of different methods have been compared by using a data set including 100 ES, 100 speckle and 100 artifact. The ROC analysis results show that adaptive neuro fuzzy inference (ANFIS) system method appears to give better results. © 2007 IEEE. © 2008 Elsevier B.V., All rights reserved.Publication Metadata only DWT based adaptive threshold determination in embolic signal detection(2007) Aydın, Nizamettin; Aydın, Nizamettin, Engineering Faculty, Bahçeşehir Üniversitesi, Istanbul, TurkeyEarly and accurate detection of emboli is important for monitoring of preventive therapy in stroke-prone patients. An embolic signal caused by emboli can be detected by Doppler ultrasound. Embolic signals are highly nonstationary and last only for a short time. Advanced processing methods are required to distinguish these signals from Doppler signals arising from red blood cells and artifacts. Any detection method involving change detection must rely on determination of an appropriate threshold. In this paper, an adaptive threshold determination method based on discrete wavelet transform (DWT) and statistical properties of the data under investigation is described. The results reveal that the proposed adaptive threshold results in a robust embolic signal detection system. © 2007 IEEE. © 2008 Elsevier B.V., All rights reserved.Publication Metadata only A novel information-theoretic clustering algorithm for robust, unsupervised classification(2007) Temel, Turgay; Aydın, Nizamettin; Temel, Turgay, Department of Electrical Engineering, Fatih Üniversitesi, Istanbul, Turkey; Aydın, Nizamettin, Engineering Faculty, Bahçeşehir Üniversitesi, Istanbul, TurkeyA new information-theoretic, unsupervised, subtractive clustering algorithm is proposed. The algorithm eliminates threshold constraint to detect possible cluster members. Cluster centers are formed with minimum entropy. Instead of using a fixed-threshold, a decision region is formed with the use of maximum mutual information. Cluster members are chosen with a relative-cost assigned in partitions of data set. The algorithm yields more reliably distributed cluster numbers in statistical sense, hence reducing further computation for validation, which is justified for a set of synthetic data. ©2007 IEEE. © 2008 Elsevier B.V., All rights reserved.Publication Metadata only Denoising embolic Doppler ultrasound signals using Dual Tree Complex Discrete Wavelet Transform(2010) Serbes, Görkem; Aydın, Nizamettin; Serbes, Görkem, Department of Electrical Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Aydın, Nizamettin, Department of Computer Engineering, Yıldız Teknik Üniversitesi, Istanbul, TurkeyEarly and accurate detection of asymptomatic emboli is important for monitoring of preventive therapy in stroke-prone patients. One of the problems in detection of emboli is the identification of an embolic signal caused by very small emboli. The amplitude of the embolic signal may be so small that advanced processing methods are required to distinguish these signals from Doppler signals arising from red blood cells. In this study instead of conventional discrete wavelet transform, the Dual Tree Complex Discrete Wavelet Transform was used for denoising embolic signals. Performances of both approaches were compared. Unlike the conventional discrete wavelet transform discrete complex wavelet transform is a shift invariant transform with limited redundancy. Results demonstrate that the Dual Tree Complex Discrete Wavelet Transform based denoising outperforms conventional discrete wavelet denoising. Approximately 8 dB improvement is obtained by using the Dual Tree Complex Discrete Wavelet Transform compared to the improvement provided by the conventional Discrete Wavelet Transform (less than 5 dB). © 2010 IEEE. © 2011 Elsevier B.V., All rights reserved.Publication Metadata only Pulmonary crackle detection using time-frequency analysis, Zaman-frekans anali̇zi̇ kullanarak pulmoner çitirti tespi̇ti̇(2012) Serbes, Görkem; Sakar, C. Okan; Kahya, Yasemin Palanduz; Aydın, Nizamettin; Serbes, Görkem, Bahçeşehir Üniversitesi, Istanbul, Turkey; Sakar, C. Okan, Bahçeşehir Üniversitesi, Istanbul, Turkey; Kahya, Yasemin Palanduz, Boğaziçi Üniversitesi, Bebek, Turkey; Aydın, Nizamettin, Bilgisayar Mühendisliǧi Bölümü, Yıldız Teknik Üniversitesi, Istanbul, TurkeyPulmonary crackles are used as indicators for the diagnosis of different pulmonary disorders. Crackles are very common adventitious sounds which have transient characteristics. From the characteristics of crackles such as timing and number of occurrences, the type and the severity of the pulmonary diseases can be obtained. In this study, a method is proposed for crackle detection. In this method, various feature sets are extracted using time-frequency analysis. In order to understand the effect of using different window types in time-frequency analysis in detecting crackles, various types of windows are used such as Gaussian, Blackman, Hanning, Hamming, Bartlett, Triangular and Rectangular. The extracted features both individually and as an ensemble of networks sets are fed into k-Nearest Neighbor classifier. Besides, in order to improve the success of the classifier, prior to the time frequency analysis, frequency bands containing no-crackle information are removed using dual tree complex wavelet transform, which is a shift invariant transform with limited redundancy compared to the conventional discrete wavelet transform. The comparative results of individual feature sets and ensemble of sets, which are extracted using different window types, for pre-processed and non pre-processed data with k-Nearest Neighbor are extensively evaluated and compared. © 2012 IEEE. © 2012 Elsevier B.V., All rights reserved.