Download Full Text (2.1 MB)
Studies and explorations of human visual perception have been the main source of inspiration for computer vision algorithms. Understanding how the human brain represents basic attributes of objects helps in developing computer vision algorithms for automatic object interpretation and understanding. Human visual perception is based on the neural coding of fundamental features, such as object boundaries, color, orientation, shape, etc. Thus, finding the contours and boundaries of objects provides the first step for object recognition and interpretation. Form here, the idea of this research inspired to introduce an automatic boundary detection technique based on active contours that is designed to detect the contours of abnormalities in X-ray and MRI imagery. Our research is aimed to aid healthcare professionals to sort and analyze large amount of imagery more effectively. Our segmentation algorithm incorporates prior information within segmentation framework to enhance the performance of object region and boundary extraction of defected tissue regions in medical imagery. We exploit Self Organizing Map (SOM) unsupervised neural network to train our prior information. One reason to prefer SOMs to other neural network models is the specific ability of SOMs to learn the intensity information via their topology preservation property. In addition, SOMs have several characteristics that make them pretty much similar to the way the human brain works. A dual self-organizing map approach is being used to learn the object of interest and the background independently in order to guide the active contour to extract the target region. The segmentation process is achieved by the construction of a level set cost function, in which, the dynamic variables are the Best Matching Units (BMU)s coming from the SOM maps. We evaluate our algorithm by comparing our detection results to the results of the manually segmented by health professionals.
Vijayan K Asari
Primary Advisor's Department
Electrical and Computer Engineering
Stander Symposium poster
Arts and Humanities | Business | Education | Engineering | Life Sciences | Medicine and Health Sciences | Physical Sciences and Mathematics | Social and Behavioral Sciences
Albalooshi, Fatema A.; Diskin, Yakov; and Paheding, Sidike, "A Self Organizing Maps Approach to Segmenting Tumors in Computed Tomography (CAT) and Magnetic Resonance Imaging (MRI) Scans" (2015). Stander Symposium Posters. 643.
Arts and Humanities Commons, Business Commons, Education Commons, Engineering Commons, Life Sciences Commons, Medicine and Health Sciences Commons, Physical Sciences and Mathematics Commons, Social and Behavioral Sciences Commons