Distorted PSF due to refractive index mismatch
In fluorescence microscopy, the main cause of Spherical Aberration are mismatches between the Refractive Index of the Lens Immersion Medium and Specimen Embedding Medium. Huygens sophisticated PSF model corrects for this mismatch during the deconvolution process.
Here we show the changes in the Point Spread Function (PSF) along the sample depth when using a 1.3 Numerical Aperture oil lens with watery medium, in a Confocal Microscope with Emission Wavelength of 520 nm.
Depths are 0, 5, 10, 15, 20 and 25 µm. The coverslip is at the bottom, as in an inverted microscope. The central XZ slices are shown, with intensities in false color (high contrast) representation. They are 2 µm wide.
Only the first PSF, at zero depth, is symmetric, but it is still elongated in comparison with the PSF of a matching condition due to a geometrical distortion (see Fishtank Effect). The Huygens software considers the zero depth position as being just above (upward microscope) or below (downward microscope) the coverslip. When imaging deeper into the sample, this 'zero depth position' is the image plane closest to the coverslip. Since Huygens version 3.5, this position can be manually adjusted with the Coverslip Position editor.
This should be avoided during acquisition by using a lens immersion medium with refractive index matching that of the specimen embedding medium (in this case, a water inmersion lens should be used).
The Huygens Software considers that the lowest plane in a 3D stack is Z = 0, the surface of the sample at the coverslip. The higher you go in the image, the deeper you travel into the sample. You can think about this as if the image was obtained with an inverted microscope. If you plan to correct for Spherical Aberration with the software, you may need to mirror your dataset to adapt it to this geometrical condition (see Spherical Aberration for more details).
The PSF asymmetry shown above is the typical one for a microscope setup in which the Lens Refractive Index is larger than the Medium Refractive Index (and the coverslip at the bottom, as Huygens expects it). The smoothed parts are at the bottom of the cones of light, and the peaks and even secondary maxima are at the top.
When you see the asymmetry the other way around is because
- the image is mirrored (thus better flip it before deconvolution) or
- the refractive index of the Lens Immersion Medium is smaller than the Specimen Embedding Medium's (thus simply use their values in the deconvolution: the image is already in the good orientation).
You can use the PSF generator in the Nyquist Calculator to simulate Spherical Aberration conditions at different depths inside the sample.
See Refractive Index Mismatch, and Mismatch In Recorded Beads.