How the brain changes its shape
My research program is focused on how the brain changes its shape. For this we use both mouse models as well as human subjects combined with imaging and advanced image processing techniques to understand subtle changes in neuroanatomy. Some particular examples include:
MRI of learning and memory
Research using human imaging, including the famous study of the hippocampus
of London Taxi drivers, has indicated that experience and expertise
is reflected on local structures of the brain. In order to understand
these effects better we are training mice on different mazes, scanning
their brains using high resolution Magnetic Resonance Images, and have
found that we can grow differing regions of the brain depending on
the training paradigm used. These studies, augmented by complimentary
imaging techniques to understand corresponding cellular events, are
continuing in order to allow us to understand how the brain changes
with learning and experience.
The effect of specialized training on the human brain
There is increasing evidence that subtle variations in our brains reflect
past experience. A particularly potent modifier of brain shape appears
to be learning a specialized skill, such as a musical instrument, ballet,
etc. This part of my research program will use imaging to attempt to
delineate precisely how such learning changes our brains, and relate
these findings back to work in the mouse (described above) in order
to ascertain more precise cellular and genetic contributions.
Methods of analyzing neuroanatomy
The research projects outlined above depend on precise automated measurements
of brain shape from MRI. We use a combination of image registration
techniques, tissue classification and deformable models to obtain these
measurements. Ongoing research will thus also focus on improving these
methods as well as understanding precisely what they can and cannot
tell us. Image processing will further be augmented by traditional
as well as novel statistics so as to identify reliable signals in our
data.
