44,99 €
inkl. MwSt.
Versandkostenfrei*
Versandfertig in über 4 Wochen
  • Broschiertes Buch

The identification of normal breast tissues in mammograms is an important step in identifying abnormal tissues, or masses. The major focus of this study is to use image-processing techniques to characterize the normal breast tissues. Clustering techniques have been developed to classify the breast tissue into one of 5 different tissue types including fat, glandular, connective, dense tissues and pectoral muscle. The classification was achieved by using a novel technique (dSGLD) involving a subset of the elements (the values on the diagonals) in the Spatial Grey Level Dependence (SGLD) matrix…mehr

Produktbeschreibung
The identification of normal breast tissues in mammograms is an important step in identifying abnormal tissues, or masses. The major focus of this study is to use image-processing techniques to characterize the normal breast tissues. Clustering techniques have been developed to classify the breast tissue into one of 5 different tissue types including fat, glandular, connective, dense tissues and pectoral muscle. The classification was achieved by using a novel technique (dSGLD) involving a subset of the elements (the values on the diagonals) in the Spatial Grey Level Dependence (SGLD) matrix which was calculated using texture features. The ultimate goal of this study is therefore to design and implement a computerized characterization method that mimics the radiologist s characterization of tissue composition in digitized mammograms. The classification results using dSGLD and SGLD features achieved an acceptable accuracy.
Autorenporträt
The Author born in 1965-Sudan, joined the University studies at Sudan University of Science and Technology, College of Medical Radiologic Sience and graduated in 1998 with B. Sc. degree in Radiation Therapy, and awarded M.Sc. in Radiation Therapy in 2000 SUST and Awarded Ph.D. in Medical Physics 2007-Natal University-South Africa.