The first step is acquiring the image data, then the images must be manipulated to give the information the experimenter needs.

Data acquisition can be performed using the ASIdiSPIM plugin in Micro-Manager, using custom-written LabView programs, or using other software that is not yet documented on this wiki (both commercial or home-built). The acquisition software will produce image files.

Combining the two views of diSPIM data into a single dataset with isotropic resolution requires preprocessing (e.g. background subtraction, ROI cropping, coarsely registering the two views, and correcting for stage-scanning distortion) before fusion (i.e. registration and joint deconvolution). Fortunately open-source software solutions are available. The Shroff group has recently expanded on their previous work with two ImageJ macros along with executable libraries: diSPIM Preprocessing¹⁾ and GPU diSPIMFusion for processing. A toolkit, momomagick, implements this algorithm in Python to process many images from the command-line interface allowing for GPU calcuration. MIPAV GenerateFusion is their CPU based approach to processing. Fiji Multi-view Reconstruction plugin was originally developed for the OpenSPIM community but can be applied to diSPIM data as well. Improving the data analysis of light sheet data is an area of active development by multiple academic groups, and hopefully companies that sell image analysis software will add registration/fusion abilities to their offerings.

Further analysis is often required depending on the experiment, for instance doing automated cell tracking over a time-series or stitching together 3D volumes into a 3D volume larger than can be acquired in one pass.