3-D Deconvolution

The 3D Deconvolution module adds a 3-D deblurring and an image restoration algorithm to SlideBook™. The nearest neighbor and constrained iterative deconvolution algorithms have been optimized for speed and performance. The module includes an interactive guide for measuring point spread functions (PSF) and a PSF database. The database can store multiple PSFs for each configuration, allowing easy selection of the appropriate PSF.

Nearest Neighbors Deconvolution

Our Nearest Neighbors deconvolution is a rapid way to deblur fluorescence data. The algorithm uses the plane above and below the plane of interest to compute and subtract the fraction of the data that is out-of-focus information.

Constrained Iterative Deconvolution

Our Constrained Iterative (CI) deconvolution is a true image restoration tool. Based on the algorithm developed by David Agard at UCSF, our CI deconvolution can quantitatively reassign out-of-focus information in 3-D data while improving both axial and lateral data resolution.

The module can use either measured point-spread functions (PSFs) or computed PSFs when measured PSFs are unavailable. The module includes a PSF collection guide and a PSF database. Given SlideBookâs knowledge of image collection optics, the correct PSF can automatically be applied to CI deconvolution without user assistance.

With SlideBook 5, CI deconvolution includes extensive speed improvements taking full advantage of advances in computational algorithms, multiple processors and hardware acceleration for both Macintosh and Windows platforms. CI deconvolution has been improved so that it can handle deconvolution of very large data sets (up to 350 MB per wavelength for a system with 2 GB RAM). In addition, CI deconvolution now includes a number of advanced user options to compensate for a variety of imaging conditions, from single cells to tissue sections, so that an optimal deconvolution can be achieved for a particular experiment.