4 sub images. The first is a base image labeled traditional. The next is more focused and labeled cubic phase plate. The third is Cubic phase plate and linear filtering. The last image is the clearest and labeled cubic phase plate  and nonlinear filteringOur approach to extending theĀ depth of fieldĀ utilizes the concept of point spread function (PSF) engineering to greatly increase the axial range clearly imagedĀ in a single exposure. The basic concept behind the extended depth of fieldĀ is that a specially designed phase plate is inserted at the back aperture of the microscope objecive. The phase plate creates a uniquely shaped PSF that is focus invariant over a depth of field of up to 10 times greater than that of a standard microscope objective. A deconvolution algorithm is then used to remove the "blur" induced by the phase plate, restoring the sharpness of the image over the full focus-invariant region of the PSF.ā€‹

Our group is working to advance EDF technology in two main areas: Design new image processing algorithms that will speedily extract this information without being adversely affected by background noise, and create new phase plate shapes that produceĀ focus invariant PSF over a larger axial range. The figureĀ below shows images of the algaeĀ OscillatoriaĀ obtained with an EDF microscope and processed with our new nonlinear filtering approach implemented via a computational neural network.ā€‹

The nonlinear filter has the capability to restore the resolution of an EDF image without amplifing the background noise. Under traditional imaging, large portions of the algae lie beyond the depth of field of the 20x/0.5NA objective and thus appear blurred. By placing a cubic phase mask in the back aperture of the objective, the depth of field is significantly extended and more of the algae appears in focus. The blur induced by the cubic phase mask was removed by usingĀ a linearĀ and a nonlinearĀ filter. The nonlinear filter enhances the image resolution with out amplifying the noise or the creating processing artifacts seen in the linear filter.