One aspect of quantitative image analysis (Q-IA) is shape analysis; in my
particular case, trying to gauge the complexity of neurons (not simply
counting dendritic branches but analysing the entire shape of the neuron).
In our lab, we have begun using Fourier and fractal analyses to help us
determine differences in neuronal shapes, assessing neuronal shape one
neuron at a time.
One problem that we have encountered, however, is how to gauge the
complexity associated with neuronal aggregation. In particular, in Case A
we have a collection of neurons forming a tight layer while in Case B we
have a collection of neurons with a distinctly dispersed pattern.
Case A Case B
xxxx x x x x
xxxx x x x x x
xxxx x x x x
xxxx x x x
xxxx x x xx
I would say that the neuronal pattern in Case B is more complex than Case
A--but what type of (mathematical) analysis should I use to quantify this
complexity?
Clint Young
Department of Psychiatry
University of British Columbia