96
B. Simon and O. Haeberlé
Microscopy (SLIM) [114–117], which is at the basis of white-light diffraction tomography or Spatial Light Interference Tomography (SLIT). The optical train is based on
a phase contrast microscopy, in which the observed sample is not illuminated using
full condenser aperture, but through an annular aperture, so that only inclined rays
illuminate the specimen.
In conventional phase contrast microscopy, those rays from the illumination,
which have passed through the sample without being diffracted (but which are already
phase retarded by passing through the sample), are focused in the back-focal plane
of the objective, where an annular phase filter adds a supplemental (but fixed) phase
shift. This phase shift is not added to the diffracted rays, which are not focused in the
back-focal plane of the objective, or Fourier plane. In the image plane, interference
between the diffracted and non-diffracted parts of the field creates a contrasted image
between the background and the sample.
In the SLIM approach, an optical module is added, to relay the Fourier plane
onto an active optical element: a spatial light modulator, which projects a controlled
sequence of four phase masks, with 0, 0.5π, π, and 1.5π delay (Fig. 4.7b). The
final phase image (Fig. 4.7c) is computed from the four frames. Finally, scanning
the observed sample in z, and properly processing of the data then allows for reconstruction of 3-D transparent samples in a tomographic mode.
This technique is also commercially distributed [118], with the advantage of being
available as an add-on module, which can be fitted to a standard microscope body.
The PhiOptics website describes applications of the SLIM and SLIT techniques.
Note also that a similar approach, but for DIC microscopy instead of phase contrast
microscopy, has also been developed, under the GLIM acronym (Gradient Light
Interference Microscopy), which allows for observation of larger, scattering samples
at the tissue scale instead of the cellular scale [119].
4.5 TDM with Sample Rotation
Tomographic Diffractive Microscopy with illumination rotation has demonstrated its
usefulness for many biological applications, and is even commercially available (see
previous section). It however suffers from the so-called “missing cone” problem along
the optical axis, leading to poor optical sectioning, if simple Fourier reconstructions,
based on the Born or Rytov approximations, are used. Things are well improved if
more elaborate reconstruction methods and/or a priori knowledge about the sample
are used, to numerically fill the missing cone, leading to improved optical sectioning,
but anyway, final images are characterized by a strong anisotropic resolution resulting
from the shape of the OTF, which is much wider in x-y directions than along the
optical axis. Same limitation holds for white-light TDM, because of the peculiar
shape of its OTF.
At macroscopic scale, it is common to perform tomography by rotating the sample
or by rotating the source-sensor ensemble around the sample; so, trying to adapt the
technique at microscopic scale is a natural approach [68–73], despite the strong
B. Simon and O. Haeberlé
Microscopy (SLIM) [114–117], which is at the basis of white-light diffraction tomography or Spatial Light Interference Tomography (SLIT). The optical train is based on
a phase contrast microscopy, in which the observed sample is not illuminated using
full condenser aperture, but through an annular aperture, so that only inclined rays
illuminate the specimen.
In conventional phase contrast microscopy, those rays from the illumination,
which have passed through the sample without being diffracted (but which are already
phase retarded by passing through the sample), are focused in the back-focal plane
of the objective, where an annular phase filter adds a supplemental (but fixed) phase
shift. This phase shift is not added to the diffracted rays, which are not focused in the
back-focal plane of the objective, or Fourier plane. In the image plane, interference
between the diffracted and non-diffracted parts of the field creates a contrasted image
between the background and the sample.
In the SLIM approach, an optical module is added, to relay the Fourier plane
onto an active optical element: a spatial light modulator, which projects a controlled
sequence of four phase masks, with 0, 0.5π, π, and 1.5π delay (Fig. 4.7b). The
final phase image (Fig. 4.7c) is computed from the four frames. Finally, scanning
the observed sample in z, and properly processing of the data then allows for reconstruction of 3-D transparent samples in a tomographic mode.
This technique is also commercially distributed [118], with the advantage of being
available as an add-on module, which can be fitted to a standard microscope body.
The PhiOptics website describes applications of the SLIM and SLIT techniques.
Note also that a similar approach, but for DIC microscopy instead of phase contrast
microscopy, has also been developed, under the GLIM acronym (Gradient Light
Interference Microscopy), which allows for observation of larger, scattering samples
at the tissue scale instead of the cellular scale [119].
4.5 TDM with Sample Rotation
Tomographic Diffractive Microscopy with illumination rotation has demonstrated its
usefulness for many biological applications, and is even commercially available (see
previous section). It however suffers from the so-called “missing cone” problem along
the optical axis, leading to poor optical sectioning, if simple Fourier reconstructions,
based on the Born or Rytov approximations, are used. Things are well improved if
more elaborate reconstruction methods and/or a priori knowledge about the sample
are used, to numerically fill the missing cone, leading to improved optical sectioning,
but anyway, final images are characterized by a strong anisotropic resolution resulting
from the shape of the OTF, which is much wider in x-y directions than along the
optical axis. Same limitation holds for white-light TDM, because of the peculiar
shape of its OTF.
At macroscopic scale, it is common to perform tomography by rotating the sample
or by rotating the source-sensor ensemble around the sample; so, trying to adapt the
technique at microscopic scale is a natural approach [68–73], despite the strong