Spiral Phase Microscopy

Severin Fürhapter; Alexander Jesacher; Christian Maurer; Stefan Bernet; Monika Ritsch-Marte    Division for Biomedical Physics, Innsbruck Medical University, A-6020 Innsbruck, Austria

I INTRODUCTION

The invention of the light microscope allowed a first glimpse into the world of micron- and smaller-sized objects that are otherwise not resolvable with the human eye. First microscopes used the brightfield mode, where a specimen is illuminated and the transmitted or reflected light is imaged by a microscope objective. This method still plays an important role in microscopy. Because the human eye is unable to recognize phase changes, a brightfield microscope is only suitable for specimens that show an amplitude contrast. An object is called an amplitude object if it absorbs parts of the incoming light due to pigments within the sample. The fact that the majority of the examined biological samples consist largely of water leads to poor contrast from the surrounding medium.

In fluorescence microscopy, biological cells are stained so that specific parts can be examined. The labeling of cells is a complex process that needs extensive preparation. The examiner must know in advance which parts of a sample are to be imaged, and on this basis, a marker must be selected. In many cases, the dyes used are harmful and destroy the sample. These shortcomings led to the development of a variety of microscopy methods whose aim was to enhance the contrast and to unveil parts of transparent specimens that are not visible in brightfield mode. Established methods in optical microscopy that solve this problem are, for instance, darkfield, phase contrast, differential interference contrast, Hoffman contrast, or Dodt contrast imaging.

In order to enhance contrast in light microscopy, the origin of the contrast must be understood (Born and Wolf, 1980). An excellent compendium that describes the principles of contrast in microscopy is, for example, given in Microscopy Primer (2006), http://micro.magnet.fsu.edu/primer/techniques/contrast.html and a general overview of imaging methods for living samples is given in Tadrous (2002) and Stephens and Allan (2003). When a microscopic sample is illuminated (e.g., by a white light source), some of the light passes through the sample without being absorbed or scattered. The remaining part of the light is diffracted from the sample and acquires a phase shift in comparison to the undiffracted light. The microscope objective projects all light beams into the image plane, where the undiffracted light evolves to a plane wave. The diffracted light focuses at different positions in the image plane, and there interference with the plane wave occurs, resulting in an intensity image of the sample.

Darkfield microscopy is one method to increase image contrast. There the zeroth order of the illumination beam is blocked such that only light diffracted, refracted, or reflected at the specimen is coupled into the microscope objective, where it can contribute to the formation of the image. The result is an illuminated object in front of a dark background. The sample is illuminated by a hollow cone of light, which is blocked by a ring in the darkfield objective, or the illumination light completely misses the collecting lens of the objective (ultra-darkfield method). This method works well for objects with low contrast and is suitable to enhance edges and contours. Since the direct illumination beam is blocked, and thus intensity is lost, this microscopy technique requires a bright light source.

An object is called a phase object if it does not absorb light and only modifies the phase of the incoming light field. The following microscope techniques are based on the fact that they convert phase differences (Barone-Nugent et al., 2002) into amplitude variations that are visible to the human eye.

Phase contrast microscopy (Zernike, 1934; Zernike, 1935; Zernike, 1955; Noda and Kawata, 1992; Barty et al., 1998; Paganin and Nugent, 1998; Liang et al., 2000; Bellair et al., 2004; Paganin et al., 2004), which was first introduced by Frits Zernike, images small differences in refractive index or thickness variations between several parts of the cell. The original central phase contrast technique is based on a filter that is placed in a Fourier plane of the imaging pathway, creating a phase difference between the diffracted and undiffracted wavefront. For small phase variations, Zernike could show (Zernike, 1942a; Zernike, 1942b) that there exists a difference of a quarter wavelength between the diffracted and the undiffracted light field in a phase sample. This phase variation cannot be seen by the human eye, which is sensitive only to intensities. By shifting the phase of the undiffracted light by another quarter wavelength, these phase variations on the sample can be transformed into amplitude variations in the image plane. If the resulting phase shift between the diffracted and the undiffracted light is half a wavelength, both light fields interfere destructively, and the method is called positive phase contrast. The specimen appears dark against a bright background. Conversely, if the diffracted and undiffracted light are in phase after the phase filter, the method is called negative phase contrast. The resulting images have bright specimen details on a dark background. The success of his method earned Zernike the Nobel Prize in Physics in 1953. An advantage of this method is that living samples can be examined. As a disadvantage, “halo effects” (i.e., bright areas around dark objects or dark areas around bright objects) appear when thicker probes are analyzed.

Differential interference contrast (DIC) was introduced by Georges Nomarski (Nomarski, 1955; Padawer, 1968; Allen et al., 1969; Pluta, 1989; Cogswell et al., 1997; Van Munster et al., 1997; Preza, 2000; Franz and Kross, 2001; Arnison et al., 2004) and utilizes phase gradients in the sample for contrast. Linearly polarized light is passed through a first modified Wollaston (or Nomarski) prism, which splits the light into two parts with a 90-degree difference between their polarizations. Behind the Wollaston prism, the two rays have a small shear in their directions, less than the optical resolution of the microscope. After passing the condenser, the light traverses the sample, and differences in refractive index or thickness affect each beam differently. Subsequently, the two beams are collected by the objective, recombined by a second Wollaston prism, and finally interfere behind a second polarizer. This procedure detects the phase difference between the sheared image waves. The result is an image with a pseudo-three-dimensional (3D) relief. The image is not isotropic, which means that only phase changes of one specific direction are detected, determined by the orientation of the Nomarski prisms. As opposed to phase contrast imaging, there are no halo effects in DIC. A disadvantage of the method is that birefringent specimens cannot be examined; thus no plastic tissue culture containers can be used. This microscopy type is also not suited to determine the exact height of a sample, because it only displays gradients within the optical thickness. It also cannot distinguish between elevations and depressions: A 180-degree rotation of the sample apparently converts a “hill” into a “valley.”

Hoffman modulation contrast (Hoffman and Gross, 1975; Hoffman, 1977), developed by Robert Hoffman, consists of a standard brightfield microscope where an aperture slit before the condenser and a modulator in the Fourier plane is inserted. The modulator is a spatially modulated optical amplitude filter that converts phase gradients into brightness variations and is inserted into the rear focal plane of a microscope objective. The cross-section of the modulator contains three parallel zones: the first zone has a transmission of 1%, 15% passes the middle zone, and the third zone, which is the largest, has a transmission of 100%. The three zones influence different gradient directions: zones 1 and 3 correspond to opposite gradient directions the intensity of which is reduced to 1% or can pass through completely, respectively. The middle zone, through which the mainly undiffracted light passes, allows the transmission of 15% of the incoming light field. The resulting...