Araştırma Çıktıları | WoS | Scopus | TR-Dizin | PubMed
Permanent URI for this communityhttps://hdl.handle.net/20.500.14719/1741
Browse
11 results
Search Results
Publication Metadata only Dendritic Spine Shape Classification from Two-Photon Microscopy Images(IEEE, 2015) Ghani, Muhammad Usman; Kanik, Sumeyra Demir; Argunsah, Ali Ozgur; Tasdizen, Tolga; Unay, Devrim; Cetin, Mujdat; Sabanci University; Fundacao Champalimaud; Utah System of Higher Education; University of Utah; Bahcesehir UniversityFunctional properties of a neuron are coupled with its morphology, particularly the morphology of dendritic spines. Spine volume has been used as the primary morphological parameter in order the characterize the structure and function coupling. However, this reductionist approach neglects the rich shape repertoire of dendritic spines. First step to incorporate spine shape information into functional coupling is classifying main spine shapes that were proposed in the literature. Due to the lack of reliable and fully automatic tools to analyze the morphology of the spines, such analysis is often performed manually, which is a laborious and time intensive task and prone to subjectivity. In this paper we present an automated approach to extract features using basic image processing techniques, and classify spines into mushroom or stubby by applying machine learning algorithms. Out of 50 manually segmented mushroom and stubby spines, Support Vector Machine was able to classify 98% of the spines correctly.Publication Metadata only Coupled Shape Priors for Dynamic Segmentation of Dendritic Spines(IEEE, 2017) Atabakilachini, Naeimeh; Erdil, Ertunc; Argunsah, A. Ozgur; Rada, Lavdie; Unay, Devrim; Cetin, Mujdat; Sabanci University; University of Zurich; Bahcesehir University; Izmir Ekonomi UniversitesiSegmentation of biomedical images is a challenging task, especially when there is low quality or missing data. The use of prior information can provide significant assistance for obtaining more accurate results. In this paper we propose a new approach for dendritic spine segmentation from microscopic images over time, which is motivated by incorporating shape information from previous time points to segment a spine in the current time point. In particular, using a training set consisting of spines in two consecutive time points to construct coupled shape priors, and given the segmentation in the previous time point, we can improve the segmentation process of the spine in the current time point. Our approach has been evaluated on 2-photon microscopy images of dendritic spines and its effectiveness has been demonstrated by both visual and quantitative results.Publication Metadata only Biomedical image time series registration with particle filtering(IEEE, 2013) Yagci, A. Murat; Erdil, Ertunc; Argunsah, A. Ozgur; Unay, Devrim; Cetin, Mujdat; Akarun, Lale; Gurgen, Fikret; Bogazici University; Sabanci University; Fundacao Champalimaud; Bahcesehir UniversityWe propose a family of methods for biomedical image time series registration based on Particle filtering. The first method applies an intensity-based information-theoretic approach to calculate importance weights. An effective second group of methods use landmark-based approaches for the same purpose by automatically detecting intensity maxima or SIFT interest points from image time series. A brute-force search for the best alignment usually produces good results with proper cost functions, but becomes computationally expensive if the whole search space is explored. Hill climbing optimizations seek local optima. Particle filtering avoids local solutions by introducing randomness and sequentially updating the posterior distribution representing probable solutions. Thus, it can be more robust for the registration of image time series. We show promising preliminary results on dendrite image time series.Publication Metadata only Concordance Between Computer-based Neuroimaging Findings and Expert Assessments in Dementia Grading(IEEE, 2013) Iheme, Leonardo O.; Unay, Devrim; Baskaya, Osman; Sennaz, Ahmet; Kandemir, Melek; Yalciner, Z. Betul; Tepe, M. Savas; Kahraman, Turgay; Unal, Gozde; Bahcesehir UniversityWe present computer-based techniques for the measurement, detection and quantification of dementia related neuroimaging findings. The three focal points of these are: brain atrophy, lateral ventricle enlargement, and white matter hyperintensity (WMH) lesion detection and quantification. Research has shown that pathological changes in these three areas have strong correlations with dementia. Our proposed methods seek to provide consistent and accurate measures of these pathologies by applying advanced image processing techniques to brain MRIs. The results are evaluated by investigating correlations with experts' grades. We report correlations of up to 0.79, 0.70 and -0.24 for hemispheric cortical atrophy, central sulcus width, and central sulcus depth, respectively. For lateral ventricle (LV) enlargement, we obtain an LV-to-brain ratio correlation of up to 0.81 among other measurements. We automatically segment and quantify periventricular WMH lesions by adaptive thresholding and 3D connectivity analysis on fluid attenuation inversion recovery MRIs. A mean Dice score of 0.83 (standard deviation is 0.06), and a correlation score of up to 0.84 is obtained.Publication Metadata only Automated Dendritic Spine Tracking on 2-Photon Microscopic Images(IEEE, 2015) Kilic, Bike; Rada, Lavdie; Erdil, Ertunc; Argunsah, A. Ozgur; Cetin, Mujdat; Unay, Devrim; Bahcesehir University; Sabanci University; Fundacao ChampalimaudThe rapid and spontaneous morphological changes of dendritic spines have been an important observation to understand how information is stored in brain. Manual assessment of spine structure has been a useful tool to understand the differences between wild type (normal) and diseased cases. In order to perform a more through analysis, automatic tools need to be developed due to the immense amount of image data collected throughout the experiments. Additionally, dendritic spines are very dynamic structures and florescence microscopy contains high level of noise, blur and shift due to the optical properties. In this study, we track locations of dendritic spines in a full series of a time-lapse two photon microscopic images. To achieve this we propose a combined detection and tracking framework. For the detection we use a SIFT based algorithm, while the tracking requires a combination of registration and distance based spine matching. Experimental results show that this technique helps to track detected spines in time series even though the noise or blur deformed the image and complicated the detection.Publication Metadata only A Watershed and Active Contours Based Method for Dendritic Spine Segmentation in 2-Photon Microscopy Images(IEEE, 2013) Erdil, Ertunc; Argunsah, A. Ozgur; Unay, Devrim; Cetin, Mujdat; Sabanci University; Fundacao Champalimaud; Bahcesehir UniversityAnalysing morphological and volumetric properties of dendritic spines from 2-photon microscopy images has been of interest to neuroscientists in recent years. Developing robust and reliable tools for automatic analysis depends on the segmentation quality. In this paper, we propose a new segmentation algorithm for dendritic spine segmentation based on watershed and active contour methods. First, our proposed method coarsely segments the dendritic spine area using the watershed algorithm. Then, these results are further refined using a region-based active contour approach. We compare our results and the results of existing methods in the literature to manual delineations of a domain expert. Experimental results demonstrate that our proposed method produces more accurate results than the existing algorithms proposed for dendritic spine segmentation.Publication Metadata only Automated Aortic Supravalvular Sinus Detection in Conventional Computed Tomography Image(IEEE, 2013) Unay, Devrim; Harmankaya, Ibrahim; Oksuz, Ilkay; Kadipasaoglu, Kamuran; Cubuk, Rahmi; Celik, Levent; Bahcesehir University; Bahcesehir University; Maltepe UniversityValvular diseases are those where one or more of the cardiac valves are affected. Treatment of valvular diseases often involves replacement or restoration of the affected valve(s). In such a surgical procedure, the medical expert performing the procedure can largely benefit from a patient-specific and dynamic valvular model containing information complementary to the 2D/3D static images. To this end, in this study a novel automated supravalvular sinus detection method (to be used as a first step in aortic valve segmentation) on conventional contrast-enhanced ECG-gated multislice CT data and its evaluation on expert annotated 31 real cases are presented. Results demonstrate a highly accurate detection performance with average error rate inferior to 1.12 mm.Publication Metadata only COMPARISON OF CEREBRAL ATROPHY GRADE WITH SIZES OF LATERAL VENTRICLE AND CENTRAL SULCI(IEEE, 2014) Gokay, Gokhan; Kandemir, Melek; Tepe, M. Savas; Yalciner, Betul; Unay, Devrim; Bahcesehir University; Bahcesehir University; Bahcesehir UniversityCerebral atrophy, which can be detected as ventricular enlargement or sulcal widening on magnetic resonance images, is one of the risk factors of dementia characterized as progressive impairment of memory and intellectual function beyond normal aging. To this end, the agreement between cerebral atrophy grade and the size (length and width) of lateral ventricles and central sulci is explored in this study. Atrophy grading is realized visually by a group of experts on magnetic resonance imaging data of 20 subjects with complaints of memory or cognitive ability. Whereas the latter is measured automatically and manually on the same data using image processing techniques. Results show that agreement with atrophy is more pronounced for lateral ventricle than central strict's, and for the width than the length of both structures. Finally, both automated and manual measurements of lateral ventricle width show high agreement (around 0.9) with atrophy grades.Publication Metadata only REGION GROWING ON FRANGI VESSELNESS VALUES IN 3-D CTA DATA(IEEE, 2013) Oksuz, Ilkay; Unay, Devrim; Kadipasaoglu, Kamuran; Bahcesehir University; Bahcesehir UniversityIn cardiac related diagnostic methods, the shape and curvature of coronary arteries is essential. Consequently, one of the most important requirements for Computer Aided Diagnosis (CAD) Systems is automated segmentation of vasculature. In this paper, we propose a new hybrid algorithm, which segment the coronary arterial tree in CTA images by merging methodologies-, namely, Region Growing and Frangi Approach. The algorithm first runs a region growing on Frangi vesselness values and subsequently optimizes the results with several threshold values. Comparison of the present results with optimal results of existing segmentation algorithms reveals that the proposed approach outperforms its predecessors. The diagnostic accuracy of the algorithm will next be validated on the segmentation of coronary arteries from real CT data.Publication Metadata only CONTENT AND CONTEXT BASED SIMILARITY FOR DISEASE DIAGNOSIS(IEEE, 2014) Ozdemir, Fehmi Volkan; Fettahoglu, Anil; Unay, Devrim; Bahcesehir University; Bahcesehir UniversityWhen there is a new patient, medical experts usually compare the results with those of previously seen cases for diagnosis and treatment. Besides, the number of digital medical images acquired, stored and managed at the healthcare centres is exponentially increasing each day with the advances in the medical imaging technology.Therefore, automated search and retrieval of similar casesfrom large medical databases may especially improve diagnosis and treatment of diseases whose causes and effects have not yet been fully known. Database management systems used at clinical sites store textual information like name, age, and gender along with medical images. Accordingly, combined usage of image features (content) with demographic and clinical information (context) for similar case search may improve accuracy andreliability of diagnosis, and thus result in increased patient care and lower healthcare costs. In this study we explored the effect of content features extracted from magnetic resonance images and context features such as patient demographics and clinical data on similarity-based disease diagnosis. Our validation on data composed of Alzheimer Parkinson, Dementia and healty cases showed that combined usage of content and context provides more accurate diagnosis with less features.