Microscopy

Microscopy

"I stumbled upon the idea of formalin fixed tissue much later and realized with satisfaction, that formalin fixation does not impair the birefringence of myelinated nerve fibers. Therefore, we can study nerve fibers hardened and conserved in formalin … with polarization microscopy and observe possible pathological alterations."

 (K. Brodmann, J Psych Neurol, 1903, p.212)

After the brain tissue to be examined has been prepared, the sectioning process takes place in a large-section cryostat at −50 °C. 50-micrometer-thick sections of the brain tissue are produced. Each of these sections is transferred to a glass slide and initially stored at −80 °C until further processing.

A high-resolution camera mounted above the cryostat captures an image of the brain surface before each cutting operation. These so-called blockface images are of crucial importance for the subsequent reconstruction of the brain.

To ensure a constant pixel size and focus during image acquisition, the blockface camera is automatically moved toward the brain block after each image via a motorized z-axis by the respective section thickness (e.g., 50 µm). Two LED panel lights, positioned to the left and right of the camera, provide the most homogeneous illumination possible.

To the top

The unstained histological brain sections are scanned using various polarimetric setups that we have developed and established as 3D Polarized Light Imaging (3D-PLI). The optical configuration of these microscopes essentially consists of two rotatable linear polarizing filters, a quarter-wave retardation plate, and a green-wavelength LED light source. The measurement principle is based on the fact that myelinated nerve fibers are optically birefringent. When polarized light passes through this birefringent tissue, its polarization state changes depending on the orientation of the myelin sheaths (and thus the nerve fibers). This change in polarization is detected by the microscopes and enables the reconstruction of the three-dimensional orientations of the nerve fibers through pixel-precise analysis.

Using this method, individual fibers and fiber tracts can thus be visualized, and their three-dimensional course can be examined across many consecutive brain sections.

Our polarimetric technologies have been continuously refined and adapted to the requirements of automated, high-throughput, large-scale polarization microscopy. Our latest 3D-PLI microscope with tilt option (LMP3D, Taorad) enables precise adjustment of the illumination angle in addition to the existing measurement principle. The measured signal differences in these additional tilt measurements allow for an even more accurate interpretation of the measurement and a clearer determination of the inclination of the nerve fibers. Thus, in addition to the fiber direction in the section plane (direction), the angle of the fibers relative to the section plane (inclination) can also be reconstructed.

Finally, signal and image processing based on high-performance computing, as well as simulation approaches, enable a reliable interpretation and visualization of the fiber architecture.

To the top

In Scattered Light Imaging (SLI), brain tissue is illuminated from various directions and the scattered light is captured by a camera. In simple terms, nerve fibres can be regarded as long, cylindrical structures. Very little light is scattered along their longitudinal axis, whereas significantly more is scattered perpendicular to it.

Varying the illumination angle produces angle-dependent scattered light profiles. From these profiles, the orientation of the nerve fibres in the white matter can be reconstructed. To do this, various illumination patterns are displayed on an LED screen. One advantage of SLI is that even crossing nerve fibres can be detected, as the scattered light signals superimpose linearly.

However, as the measured light is very faint and large amounts of data are generated, we are developing optimised illumination patterns. Similar to a QR code, information is encoded into the illumination by illuminating the sample from several directions simultaneously. This increases the measured light intensity, thereby reducing the measurement time and improving the effective angular resolution of the nerve fibre orientations.

Publication: Optimizing the lighting

To the top

Unlike conventional, intensity-based microscopy, digital holographic microscopy (DHM) enables the measurement of the complete complex-valued light field. In addition to intensity, this also allows the phase of the light field to be determined, which contains additional information about the sample.

Holography utilises the interference between a light wave modulated by the sample and an unmodulated reference wave. Assuming a largely transparent sample, amplitude and phase information can be reconstructed from this.

In our setup, we use in-line holography with a so-called double-sideband filtering method. This enables the simplest and most stable measurement possible, as holographic methods are highly sensitive to external influences. This is achieved using a Spatial Light Modulator (SLM) and linear polarizers arranged behind the sample.

As DHM measures both the amplitude and the phase of the light field, digital filters can be applied directly to the reconstructed light field. This allows, for example, dark-field or phase-contrast images to be generated computationally. As with the optical modulation in conventional dark-field or phase-contrast microscopes, this filtering takes place in Fourier space.

This allows the contrast to be specifically adapted to certain tissue components. Structures that are only faintly visible in conventional intensity images can thus be highlighted more clearly and made more useful for further analyses, such as cell segmentation or cell classification.

Furthermore, the recording of the complete light field—that is, both amplitude and phase—enables the digital calculation of light propagation in 3D. To achieve this, the Helmholtz equation, which describes the propagation of electromagnetic waves, is solved numerically. By applying autofocus criteria, it is then possible to determine the depth at which macroscopic structures, such as thick fibre bundles, are located.

In this way, spatially overlapping structures can be better separated from one another. This is particularly helpful in resolving intersections of fibre bundles and examining the three-dimensional organisation of brain tissue in greater detail.

To the top