2D STICS (space,time)

For the theory/introduction to spatiotemporal image correlation spectroscopy (STICS), I would recommend these papers (one, two).

As an input, this command takes an 8-,16-,32- bit image stack (only with time or only with Z, not both). It is assumed that the z-axis corresponds to time.

As an output it generates average x,y 2D cross correlation maps for different values of time delay.

For example, we have a stack with N frames, labeled as 1,2,3..N. If time delay value is equal to 3, the command will calculate 2D cross-correlation maps between all image pairs (1,4), (2,5), (3,6)...(N-3,N) and average their values at each pixel (at each displacement).

So as output this command will provide a 32-bit stack with those average normalized cross-correlation values for time delays in the range from 0 to specified maximum value (in frames):

The values of corresponding time delays are shown as image names in the stack. The origin of coordinates in the output images is calibrated similarly to 2D cross-correlation.

In general, this analysis can be improved by some pre-filtering/cleaning of the stack, like background subtraction or immobile fraction removal. This is not provided in the current plugin, you have to do it yourself in ImageJ/FIJI with macros.

There are alternative plugins for this purpose with more functionality, (for example MICS).

The current plugin was written mostly for the purpose of self-education (but it still works fine).