Zernike coefficients - data reduction?

Hello all! I am currently looking into nuclear shape changes under various conditions, and have been using the MeasureSizeShape module’s Zernike features. Measurements show condition-dependent difference over a wide range of Zernike coefficients.

Coming from a background with only basic mathematics skills, my question is: does anybody know of an approach to to reduce this multi-parameter data into a single parameter (e.g. indicating a more general shape change)?



Our group has had some success using both Principal Component Analysis and Linear Discriminant Analysis (using JMP) to reduce the dimensionality of our morphological data. Hope this was helpful.


1 Like