Support vector machines in medical image analysis

Description:  Support Vector Machines (SVM) are a classification tool with broad applications. The process is based on the representation of each element of the set in a multidimensional feature space.This machine learning method relies then on optimization to find a set of splitting hyperplanes that separate each class elements on a given labeled training set with maximum margin. This originates a classification on the feature space that is then used to classify other elements whose classification is unknown. The aim of this talk is to discuss the use of SVM in medical imaging, namely in image data acquired from optical coherence tomography (OCT) and magnetic resonance imaging (MRI). We are interested in image segmentation and image classification and will show some applications of SVM on current work developed at our institute. Date:  2011-05-13 Start Time:   14:30 Speaker:  Rui Bernardes, Pedro Serranho, João Duarte (IBILI, Coimbra) Institution:  IBILI Place:  Room 5.5 Organization:  LCM - Laboratory for Computational Mathematics See more:   <Main>   <Lab for Computational Mathematics>