Smart illumination for 3D-imaging of biological tissues FOCUS
Figure 1. A three-dimensional structure is captured by acquiring a stack of 2D slices at successive positions along the optical( z) axis( 4 of ~ 50 shown). Effective optical sectioning means each slice is sensitive only to a narrow axial thickness, Δz; if Δz is large, out-of-focus fluorescence from neighboring planes blurs the image. while access to the third dimension( z) is indirect.
To create a 3D image with a conventional microscope, a stack of 2D images must be acquired while translating the sample along the optical axis. In principle, this method should work: each slice samples a different depth, and the stack encodes the volume( see Fig 1).
In practice, however, we encounter the sectioning problem. As the excitation light propagates along the optical axis, it also excites fluorophores above and below the focal plane that is optically conjugated to the camera. These fluorophores emit photons that are not part of the desired image plane; they travel back along the same axis and impinge on the camera as a blurred background. In link with this background, widefield microscopes discriminate objects poorly along the optical axis: a uniform fluorescent layer perpendicular to the optical axis cannot be localized, and two homogeneous layers at different depths are indistinguishable. The thicker the specimen, the stronger this unwanted background becomes.
This is why the confocal microscope, introduced as a solution for optical sectioning, has become such an important tool in biology.
CONFOCAL MICROSCOPY— RESTORING OPTICAL SECTIONING A confocal microscope forms an image one point at a time( Fig. 2). A tightly focused excitation spot scans the sample. At the image plane conjugated to the focal plane, the in-focus fluorophores excited by the beam focus yield a small intense spot while the out-of-focus fluorophores, excited by the diverging beam, form a large low-intensity smear. By placing a pinhole strictly conjugated to the excitation focus spot, one ensures that most of the light emitted by the in-focus fluorophores is collected while most of the light of the out-of-focus fluorophores is discarded. Emission produced above or below the focus being largely blocked at the pinhole( Fig. 2), the microscope now records a slice with far less out-offocus background. By repeating the scan at successive depths, a 3D stack is built with far better axial discrimination than the conventional-widefield- microscope.
While this introduces sectioning, it also comes at a price in the form of necessary trade-offs for 3D imaging. The pinhole rejects most of the emitted photons, many of which carry useful information, so the overall throughput is low. In order to maintain the signal, users often have to slow down the scan in order to accumulate more photons per pixel, which reduces the imaging temporal resolution; alternatively, they can open the pinhole, which sacrifices sectioning. Ultimately, the low usable signal per pixel( photon budget) forces the user to increase the excitation power. However, biological tissues can be very sensitive to light exposure. Excited fluorophores can generate reactive oxygen species( ROS) that can alter signalling or trigger cell death. In practice, the cumulative dose increases with the number of planes, dwell time and re-imaging frequency, meaning that 3D confocal imaging imposes stress that scales with volume and time.
In subsequent sections, we
Photoniques 134 I www. photoniques. com 37