Slide 34 of 42
Now here's a very different kind of dataset, this is a tomographically
reconstructed image of part of a dendrite from a rat neuron. This
dataset is from the Collaboratory for Microscopic Digital Anatomy, the
project I was associated with at Cornell. These scatterplots don't
have the ideal shape- they don't contain the nice curves we saw
before. So this is a dataset that is more challenging than the CT
datasets we saw before, since there is no one obvious iso-value to use
to find the surface. But, the position function calculated for this
dataset does have just one zero-crossing, so that's encouraging
