Can anyone please suggest a protocol for SPIM/Light sheet microscope alignment of the light sheet to the focal plane of the detection with fluorescent beads??
So I’m building a SPIM, but there’s still a little bit of gyration of the rotation arm of the 4D stage (to about 1 to 2 degrees off the vertical axis). To correct for this, what I did during imaging before acquiring a stack for each view was to roughly align a reference point on the sample (such as the top of the sample which can be seen at the different angles) to crosshairs on live view. I’m thinking imaging with beads in addition to this will further make the reconstruction better. what are your thoughts on this, please?
I can send you a copy of the last article if you need that.
For alignment there are two schools of thought so to speak:
A) If the focus of the lightsheet is in focus of the lens the brightness of the bead will be highest. The bead will also appear the smallest in xy. If completely out of focus the bead will be large and have rings.
B) The other way is to take a z-stack and then look at the PSF of the bead in the x-z or y-z direction (Orthogonal view). The PSF should appear as a symmetric hourglass (if the lens is also in good shape). Since the human eye has problems in picking up brightness differences easily I always preferred the latter geometric assessment. The peak in brightness and contrast might be better suited for an automatic alignment. Keep in mind that the optical properties will degrade the deeper you go in any sample or sample embedding (agarose, fep tube). Thus alignment using beads in the center of the field of view and at the top - closest to the objective lens - of the sample embedding will allow you to have a good optical alignment.
In my experience the multiview reconstruction using beads is giving robust reconstructions that can be automated. Depending on your sample a sample based registration can be beneficial since the views can be distorted due to scattering and bending of the light inside the sample. But it might require more manual optimization.
Using the multiview reconstruction application (https://imagej.net/Multiview-Reconstruction) you can combine registrations from different sources. Thus first use a bead based registration to get a good first registration and apply a sample based registration on top of that.
The shape of the PSF is independent of the illumination side.
Any point source of light will be convolved by the PSF when acquired through an optical system.
The size in xy will be determined by the objective mostly. In z it will be a product of the resolution in z and the width of the lightsheet (optical sectioning).
What you will see with a single side is that one side of the object, may it be the sample or an agarose column will be bright and of better image quality. Since the light will be scattered through the sample.
Look at the Papers from Ernst Stelzer and Jan Huiskens lab they have pictures explaining this.
There are plugins for extracting the PSF and measuring properties. But these will be mostly useful if you want to monitor the optical quality of your system over time. For alignment it is sufficent to first bring the beads into focus. So that they are sharp and bright. Then acquire a z-stack and look if the hourglass of the PSF is symetrical. Using Fijis orthogonal views for example: https://imagej.nih.gov/ij/docs/guide/146-28.html#sub:Orthogonal-Views
Do you have an OpenSPIM? There are some more tricks with getting the lightsheet properly straight and properly centered. I can put you in touch with the author and also current users of that system.