Mark E. Bastin1, Susana Muoz Maniega1, Jonathan D. Clayden2, Amos J. Storkey1, Laura J. E. Brown1, Alasdair M J MacLullich1
1University of Edinburgh, Edinburgh, Midlothian, UK; 2Institute of Child Health, University College London, London, UK
Tractography provides a promising tool for assessing white matter connectivity in old age. However, tractography output is usually strongly dependent on user-specified seed points. We have shown, however, that it is possible to segment the same fasciculus in groups of subjects using a method we term neighborhood tractography (NT). In addition to the original heuristic NT approach, we have recently developed two new NT methods which create probabilistic tract-matching models using supervised and unsupervised learning techniques. Here we investigate which of these three NT methods performs best in segmenting tracts in the brains of a cohort of elderly subjects.