Browsing by Subject "Subgraphs"

Now showing 1 - 4 of 4

Open Access
Altçizge modellemesi kullanarak kolon bez tespiti
(IEEE, 2011-04) Özgül, Etkin Barış; Sökmensüer, C.; Gündüz-Demir, Çiğdem
Kolon adenokarsinomu, kolon bez yapılarında değişimlere yol açar. Patologlar bezlerdeki bu değişimleri değerlendirerek kolon adenokarsinom tanı ve derecelendirmesi yaparlar. Ancak değişimlerin değerlendirme süreci kaydadeğer öznellik taşıyabilir. Bezlerin matematiksel özniteliklerle karakterize edilmesiyle bu öznelliği azaltabilmek olasıdır. Bunun içinse ilk aşama, bezlerin yerlerinin otomatik olarak tespit edilmesidir. Literatürdeki bez tespit etme yöntemleri çoğunlukla piksel tabanlıdır. Ancak doku görüntüleri, doğaları gereği ve biyopsi hazırlama ve görüntü alma işlemlerindeki değişkenlik nedeni ile piksel bazında değişkenlik gösterebilir. Öte yandan, bu değişkenliğe rağmen, bezleri oluşturan doku bileşenlerinin uzaysal dağılımı benzer özellik gösterir. Bu dağılımı gözönüne alarak tasarlanan yöntemler, bölütleme başarısını artırma potansiyeline sahiptir. Bu çalışmada önerdiğimiz yöntem, ilk olarak, doku bileşenlerinin dağılımını, bu bileşenler üzerinde oluşturduğu bir çizge ile modeller. Daha sonra, oluşturduğu bu çizgeyi altçizgelere ayırır ve bu altçizgelerin öznitelikleri yardımıyla bezleri tespit eder. Kolon doku görüntüleri üzerinde yaptığımız çalışmalar, önerilen bu yöntemin bezlerin yüksek doğrulukta tespit edilmesinde umut verici sonuçlar verdiğini göstermiştir. The colon adenocarcinoma causes changes in glandular structures of colon tissues. Pathologists assess these changes to diagnose and grade the colon adenocarcinoma. However, this assessment may consist of a considerable amount of subjectivity. It is possible to reduce this subjectivity by characterizing the glands with mathematical features. For that, the first step is to detect gland locations. In literature, most of the gland detection methods are pixel-based. However, tissue images may show pixel-level variances due to their nature and differences in biopsy preparation and image acquisition procedures. On the other hand, in spite of these variances, the distribution of tissue components forming glands show similar properties. The methods that consider this distribution has the potential of improving the performance. The method proposed in this study first models the distribution of the components by constructing a graph on them. Then, it breaks the constructed graph down into subgraphs and detects the glands using the features of these subgraphs. The experiments conducted on colon tissue images show that the proposed method leads to promising results for detecting the glands. © 2011 IEEE.
Open Access
Finding compound structures in images using image segmentation and graph-based knowledge discovery
(IEEE, 2009-07) Zamalieva, Daniya; Aksoy, Selim; Tilton J. C.
We present an unsupervised method for discovering compound image structures that are comprised of simpler primitive objects. An initial segmentation step produces image regions with homogeneous spectral content. Then, the segmentation is translated into a relational graph structure whose nodes correspond to the regions and the edges represent the relationships between these regions. We assume that the region objects that appear together frequently can be considered as strongly related. This relation is modeled using the transition frequencies between neighboring regions, and the significant relations are found as the modes of a probability distribution estimated using the features of these transitions. Experiments using an Ikonos image show that subgraphs found within the graph representing the whole image correspond to parts of different high-level compound structures. ©2009 IEEE.
Open Access
Graph walks for classification of histopathological images
(IEEE, 2013) Olgun, Gülden; Sokmensuer, C.; Gündüz-Demir, Çiğdem
This paper reports a new structural approach for automated classification of histopathological tissue images. It has two main contributions: First, unlike previous structural approaches that use a single graph for representing a tissue image, it proposes to obtain a set of subgraphs through graph walking and use these subgraphs in representing the image. Second, it proposes to characterize subgraphs by directly using distribution of their edges, instead of employing conventional global graph features, and use these characterizations in classification. Our experiments on colon tissue images reveal that the proposed structural approach is effective to obtain high accuracies in tissue image classification. © 2013 IEEE.
Open Access
Image classification using subgraph histogram representation
(IEEE, 2010) Özdemir, Bahadır; Aksoy, Selim
We describe an image representation that combines the representational power of graphs with the efficiency of the bag-of-words model. For each image in a data set, first, a graph is constructed from local patches of interest regions and their spatial arrangements. Then, each graph is represented with a histogram of subgraphs selected using a frequent subgraph mining algorithm in the whole data. Using the subgraphs as the visual words of the bag-of-words model and transforming of the graphs into a vector space using this model enables statistical classification of images using support vector machines. Experiments using images cut from a large satellite scene show the effectiveness of the proposed representation in classification of complex types of scenes into eight high-level semantic classes. © 2010 IEEE.