3D salient point detector based on the dual-tree complex wavelet transform with application to the positioning of deformable meshes in MR brain images

• Grant number: 17/24391-0
• Support Opportunities: Scholarships in Brazil - Master
• Start date: March 01, 2018
• End date: July 31, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Agreement: Coordination of Improvement of Higher Education Personnel (CAPES)
• Principal Investigator: Ricardo José Ferrari
• Grantee: Breno da Silveira Souza
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

Deformable geometric models have been used with great success in various medical imaging applications. The main appealing of this type of approach in the segmentation of medical images is the fact that such models bring in themselves the topology of the anatomical organ they represent. Thus, this method becomes less susceptible to image noise and can be applied even in situations where the contrast level of the anatomy under study in the image is too low. One of the main difficulties of the use of deformable geometric models in the segmentation of medical images is the adequate initial positioning of the model. For the success of the segmentation, the model must be positioned in a region very close to the structure that is to be segmented in the image. Therefore, in this research, we intend to develop a 3D salient point detector based on the dual-tree complex wavelet transform for application in the positioning of deformable meshes in magnetic resonance (MR) images of the brain. The Hidden Markov Tree (HMT) model will be studied using three different scenarios to analyze the persistence of the magnitude of the complex coefficients and thus to define stable points for further analysis. For each detected salient point, a descriptor will be calculated from the neighborhood coefficients of that position in the different decomposition subbands of the image. The detector will be evaluated using both synthetic images and clinical MR images. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

SOUZA, BRENO DA SILVEIRA; POLONI, KATIA M.; FERRARI, RICARDO J.. Detector of 3-D salient points based on the dual-tree complex wavelet transform for the positioning of hippocampi meshes in magnetic resonance images. JOURNAL OF NEUROSCIENCE METHODS, v. 341, JUL 15 2020. (18/06049-5, 18/08826-9, 17/24391-0)