Supertoroid-based representation of the diffusion tensor: assessment of brain white matter in bipolar disorder

• Grant number: 11/00893-0
• Support Opportunities: Regular Research Grants
• Duration: May 01, 2011 - April 30, 2013
• Field of knowledge: Engineering - Biomedical Engineering - Medical Engineering
• Principal Investigator: Marcel Parolin Jackowski
• Grantee: Marcel Parolin Jackowski
• Host Institution: Instituto de Matemática e Estatística (IME). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated researchers: Choukri Mekkaoui ; Sheila Cavalcante Caetano

Abstract

White matter (WM) is of fundamental importance in understanding brain function because of the communication network it establishes. It is known that neuropsychiatric disorders such as bipolar disorder (BD) may involve subtle WM abnormalities that require robust and efficient analysis methods. Diffusion tensor imaging (DTI) has emerged as a promising noninvasive imaging modality to characterize WM structural alterations. However, the multidimensional and multivariate nature of DTI makes the neuropathological study in BD a challenging task. Hence, robust and appropriate DTI analysis methods are required to improve the understanding of the role of such abnormalities in psychiatric disorders. This project aims the development and evaluation of a novel DTI analysis methodology based on the supertoroidal representation of the diffusion tensor. This representation attempts to improve the characterization of WM both visually and quantitatively by offering two new indices of diffusivity and anisotropy. Experiments using tissue-mimicking phantoms have revealed that these indices are sensitive to subtle structural alterations and thus may be able to describe changes in the brain. Preliminary results using DTI dataset of a normal human brain showed that major WM tracts could be identified unambiguously. In order to evaluate the effectiveness of the supertoroidal model, DTI datasets of BD patients and healthy individuals will be analyzed and compared statistically. We expect that our approach may improve the understanding of the underlying neurophysiology of BD and provide meaningful information for WM characterization in other psychiatric disorders as well. (AU)