Meeting Banner
Abstract #3966

Clustering Method for Estimating Principal Diffusion Directions

Mohammadreza Nazem-Zadeh1, Kourosh Jafari-Khouzani2, Abbas Babajani-Fermi2, Siamak Pourabdollah Nejad-Davarani1, Hamid Soltanian-Zadeh2,3, Quan Jiang1

1Neurology, Henry Ford Hospital, Detroit, MI, United States; 2Diagnostic Radiology, Henry Ford Hospital, Detroit, MI, United States; 3Control & Intelligent Processing Center of Excellence, School of Electrical & Computer Engineering, University of Tehran, Tehran, Iran

Using High angular resolution diffusion imaging (HARDI), the fiber orientation distribution function (ODF) on the unit sphere is calculated and used to extract the principal diffusion directions (PDDs). Fast and accurate estimation of PDDs is a prerequisite for tracking algorithms that deal with fiber crossings. In this paper, a clustering approach to estimate PDDs is proposed which is an extension of fuzzy c-means clustering developed for orientation coordinates of points on a sphere. Experimental results illustrate that the proposed clustering algorithm is more accurate, more resistant to noise, and faster than the techniques currently being utilized.