A structural method for pediatric MRI segmentation

• Grant number: 17/09465-7
• Support Opportunities: Scholarships in Brazil - Master
• Start date: December 01, 2017
• End date: December 31, 2018
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Roberto Marcondes Cesar Junior
• Grantee: Mateus Riva
• Host Institution: Instituto de Matemática e Estatística (IME). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

Segmentation and recognition in medical imaging of fetuses and children, as well as non-standard individuals (e.g. patients with missing organs after major surgery), presents important challenges. One of the main reasons is that most of the state-of-art methods have been developed and applied to standard models and data (typically adults). Similar challenges arise on the continuous study of an individual over time, which allows a better understanding of the individual's evolution and development. Tracking the changes of recognized organs and other structures along time may show information not available on a single image. This project aims at developing a structural (graph-based) method for the analysis in pediatric imaging. More specifically, a new structural method for model learning and application to image segmentation will be developed and applied to real MRI data. (AU)

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

RIVA, Mateus. A new calibration approach to graph-based semantic segmentation. 2018. Master's Dissertation - Universidade de São Paulo (USP). Instituto de Matemática e Estatística (IME/SBI) São Paulo.