86
B. Simon and O. Haeberlé
methods have found their preferred domain of application. But, while very powerful
for phenomenon detection, or morphological measurements, they are all limited in the
sense that interpretation of the observed contrast is difficult in terms of quantitative
measurements of the optical properties.
Nowadays, fluorescence microscopy is often the reference method, because it
allows for specific labeling of cellular structures, therefore cellular functions, with
ever-increasing set of natural or especially chemically designed fluorescence labels.
Being indeed a dark field technique, fluorescence microscopy has an extreme sensitivity, leading to single molecule detection and tracking, and in recent years, fluorescence microscopy has also evolved into fluorescence nanoscopy, with an unsurpassed
resolution obtainable with modern techniques like STED microscopy or pointillism
microscopies [1–5]. Conversely, fluorescence microscopy also has the ability to study
at the cellular, and even subcellular scale, large samples such as organoids, and even
entire model organisms as large as Danio rerio [6].
In some cases, fluorescence labeling may, however, induce unfavorable effects
like photo-toxicity. Labeling itself may be difficult for some samples. Speed is also
rather limited, especially for 3D imaging, because fluorescence signal is often weak
(compared to transmission microscopy for example). In some cases, use of fluorescence is also to be avoided. For example, for long time-lapse studies, fluorophores are
diluted in dividing cells, leading to vanishing signal over time, and for some studies
on natural cell lines, use of genetically modified organisms is simply not possible.
These limitations explain that, in recent years, there has been a growing interest for developing new imaging techniques not requiring specific sample labeling,
like second or third harmonic generation (SHG and THG) microscopy, Raman or
Coherent Anti-Stokes Raman Scattering (CARS) microscopies [7–11]. These methods have proven their interest, CARS being, for example, very effective for lipid
detection, but having as drawback that they require expensive instrumentation for
signal generation/detection.
As possible alternate for efficient label-free microscopy, the coupling of fast-,
high-sensitivity, and high-resolution electronic sensors with the exploding computational and memory capacities of modern computers has recently open new possibilities for revisiting conventional transmission optical microscopy.
4.2 Digital Holographic Microscopy
In conventional transmission microscopy (classical wide-field, dark-field, phasecontrast, Differential Interference Contrast, Hoffman modulation…), the image is
formed by a complex interaction of the incoherent illuminating light with the specimen. The recorded image is an intensity-only image, which contrast is very efficient
to detect structures, and therefore is for example very helpful for morphological
evolution studies, but which does not easily deliver quantitative information on the
optical characteristics of the observed specimen. In particular, refraction and absorp-
B. Simon and O. Haeberlé
methods have found their preferred domain of application. But, while very powerful
for phenomenon detection, or morphological measurements, they are all limited in the
sense that interpretation of the observed contrast is difficult in terms of quantitative
measurements of the optical properties.
Nowadays, fluorescence microscopy is often the reference method, because it
allows for specific labeling of cellular structures, therefore cellular functions, with
ever-increasing set of natural or especially chemically designed fluorescence labels.
Being indeed a dark field technique, fluorescence microscopy has an extreme sensitivity, leading to single molecule detection and tracking, and in recent years, fluorescence microscopy has also evolved into fluorescence nanoscopy, with an unsurpassed
resolution obtainable with modern techniques like STED microscopy or pointillism
microscopies [1–5]. Conversely, fluorescence microscopy also has the ability to study
at the cellular, and even subcellular scale, large samples such as organoids, and even
entire model organisms as large as Danio rerio [6].
In some cases, fluorescence labeling may, however, induce unfavorable effects
like photo-toxicity. Labeling itself may be difficult for some samples. Speed is also
rather limited, especially for 3D imaging, because fluorescence signal is often weak
(compared to transmission microscopy for example). In some cases, use of fluorescence is also to be avoided. For example, for long time-lapse studies, fluorophores are
diluted in dividing cells, leading to vanishing signal over time, and for some studies
on natural cell lines, use of genetically modified organisms is simply not possible.
These limitations explain that, in recent years, there has been a growing interest for developing new imaging techniques not requiring specific sample labeling,
like second or third harmonic generation (SHG and THG) microscopy, Raman or
Coherent Anti-Stokes Raman Scattering (CARS) microscopies [7–11]. These methods have proven their interest, CARS being, for example, very effective for lipid
detection, but having as drawback that they require expensive instrumentation for
signal generation/detection.
As possible alternate for efficient label-free microscopy, the coupling of fast-,
high-sensitivity, and high-resolution electronic sensors with the exploding computational and memory capacities of modern computers has recently open new possibilities for revisiting conventional transmission optical microscopy.
4.2 Digital Holographic Microscopy
In conventional transmission microscopy (classical wide-field, dark-field, phasecontrast, Differential Interference Contrast, Hoffman modulation…), the image is
formed by a complex interaction of the incoherent illuminating light with the specimen. The recorded image is an intensity-only image, which contrast is very efficient
to detect structures, and therefore is for example very helpful for morphological
evolution studies, but which does not easily deliver quantitative information on the
optical characteristics of the observed specimen. In particular, refraction and absorp-