Classification of brain tumours by CT image Walsh spectra

Wolf, Martin; Ziegengeist, S.; Michalik, Michał; F, Bornholdt; Michalik, S; Meffert, Beate

doi:10.1007/bf02426456

Cited by 1 publication

(1 citation statement)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Albrecht et al presented in [11] a pattern classification method that combines the classical perceptron algorithm with simulated annealing in order to recognize focal liver tumors in CT images. Wolf et al [12] studied the hierarchical classification of brain tumors by CT-image Walsh spectra. Our work will provide texture extraction and classification of normal tissue across a wide range of organs and tissues in Computed Tomography (CT) scans.…”

Section: Image Classification: Related Workmentioning

confidence: 99%

Mining Knowledge in Computer Tomography Image Databases

Raicu¹

Multimedia Data Mining and Knowledge Discovery

View full text Add to dashboard Cite

Summary: This chapter presents our research results obtained for texture extraction, classification, segmentation, and retrieval of normal soft tissues in Computed Tomography studies of the chest and abdomen. The texture extraction step consists of various texture methods applied to the collection of tissue data in order to derive a set of features characterizing the best the visual perception of texture. The classification step involves different data mining learning models used to automatically map similar texture features to the same type of tissues, and produce a set of rules that can be used for automatic classification and annotation of unlabelled image data.When the classification approach is applied at the local (pixel) level, it can be used for automatic segmentation. Each pixel will receive a label through the classification rules and connected pixels having the same labels will form a region or segment in the corresponding image. This type of segmentation will have a significant impact on the current research efforts for providing automatic context (i.e. that the cursor is hovering over "liver" in CT images). The image retrieval step consists of the selection of the best similarity metric and the best texture feature representation in order to retrieve the most similar images with the organ query image. Since there is no similarity measure known to perform the best for the CT modality, we compare eight metrics and three different feature representations, and show how the selection of a similarity metric affects the texture-based retrieval. Furthermore, since our work deals with normal tissues, the system proposed here can be considered as a first step for supporting the clinical decision-making process.

show abstract

Section: Image Classification: Related Workmentioning

confidence: 99%

Mining Knowledge in Computer Tomography Image Databases

Raicu¹

Multimedia Data Mining and Knowledge Discovery

View full text Add to dashboard Cite

show abstract

Classification of brain tumours by CT image Walsh spectra

Cited by 1 publication

References 2 publications

Mining Knowledge in Computer Tomography Image Databases

Mining Knowledge in Computer Tomography Image Databases

Contact Info

Product

Resources

About