240
T. C. Jagadale and S.-W. Chu
10.1 Introduction
10.1.1 Why Label-Free Super-Resolution Light Microscopy?
Light microscopes facilitate the magnified imaging of bio-specimen based on the
principle that the rays from the specimen converge onto a corresponding single
point at the image plane. However, beyond a critical magnification wherein size
of structures becomes comparable to the wavelength of light, diffraction takes in
and the structures cannot be well resolved. This is known as diffraction limit in
light microscopy, which had been explained theoretically by Airy [1] (1835) and
subsequently formulated by Ernst Abbe in 1873 [2] and Rayleigh in 1896 [3]. It
states, in a loose sense, that it is impossible to resolve two structures that are closer
to each other than half of an excitation wavelength in lateral plane (x, y) and even
further apart in the longitudinal plane (z).
To overcome the resolution limit, non-optical imaging approaches such as electron microscopy and scanning probe microscopy [4, 5] have been developed to offer
nano-scale resolution. However, compared to light microscopy, electron microscopy
does not allow in vivo bio-imaging, and scanning probe microscopy is limited to surface observations. In addition, the mature fluorescence labelling in light microscopy
provides precise molecular imaging with high contrast. Currently, the majority of
research in life sciences is done using light microscopy since light is transparent
to cells and provides non-invasive in vivo imaging into tissues, but with only submicrometre resolution [6]. Therefore, it has been a historically long pursuit to develop
resolution enhancing techniques of light microscopy.
In the era of Abbe and Rayleigh, resolution is defined under the geometry of widefield microscopy, in which the entire specimen is exposed to light, and the image
can be viewed directly from an objective. The best wide-field microscope could
reach the resolution of ~250 nm laterally and ~800 nm axially [7]. To extend the
classical diffraction limit, Marvin Minsky in 1957 patented the concept of confocal
microscopy [8], in which the sample is scanned with a focussed beam and a pinhole
is used to block the out-of-focus light, improving lateral resolution by a factor of
√
2. In addition, confocal microscopy exhibits optical sectioning capability, so the
axial contrast is greatly improved.
Another optical sectioning technique, multi-photon microscopy [9, 10], which is
based on simultaneous absorption of two or more photons at longer wavelength (thus
less scattering), significantly increases penetration depth in thick samples. The intrinsic nonlinearity sharpens the focal excitation volume and leads to a corresponding
reduction of point spread function (PSF) width by a factor of
√
2 for two-photon and
√
3 for three-photon microscopy. Nevertheless, due to long excitation wavelengths,
the lateral resolution is not as good as confocal microscopy. Presently, the main
application of multi-photon microscopy is for deep-tissue imaging up to 1 mm, with
sub-micrometre spatial resolution.
One major evolution in optical sectioning technique recently is the invention
of selective plane illumination microscopy (SPIM), or more generally light sheet
T. C. Jagadale and S.-W. Chu
10.1 Introduction
10.1.1 Why Label-Free Super-Resolution Light Microscopy?
Light microscopes facilitate the magnified imaging of bio-specimen based on the
principle that the rays from the specimen converge onto a corresponding single
point at the image plane. However, beyond a critical magnification wherein size
of structures becomes comparable to the wavelength of light, diffraction takes in
and the structures cannot be well resolved. This is known as diffraction limit in
light microscopy, which had been explained theoretically by Airy [1] (1835) and
subsequently formulated by Ernst Abbe in 1873 [2] and Rayleigh in 1896 [3]. It
states, in a loose sense, that it is impossible to resolve two structures that are closer
to each other than half of an excitation wavelength in lateral plane (x, y) and even
further apart in the longitudinal plane (z).
To overcome the resolution limit, non-optical imaging approaches such as electron microscopy and scanning probe microscopy [4, 5] have been developed to offer
nano-scale resolution. However, compared to light microscopy, electron microscopy
does not allow in vivo bio-imaging, and scanning probe microscopy is limited to surface observations. In addition, the mature fluorescence labelling in light microscopy
provides precise molecular imaging with high contrast. Currently, the majority of
research in life sciences is done using light microscopy since light is transparent
to cells and provides non-invasive in vivo imaging into tissues, but with only submicrometre resolution [6]. Therefore, it has been a historically long pursuit to develop
resolution enhancing techniques of light microscopy.
In the era of Abbe and Rayleigh, resolution is defined under the geometry of widefield microscopy, in which the entire specimen is exposed to light, and the image
can be viewed directly from an objective. The best wide-field microscope could
reach the resolution of ~250 nm laterally and ~800 nm axially [7]. To extend the
classical diffraction limit, Marvin Minsky in 1957 patented the concept of confocal
microscopy [8], in which the sample is scanned with a focussed beam and a pinhole
is used to block the out-of-focus light, improving lateral resolution by a factor of
√
2. In addition, confocal microscopy exhibits optical sectioning capability, so the
axial contrast is greatly improved.
Another optical sectioning technique, multi-photon microscopy [9, 10], which is
based on simultaneous absorption of two or more photons at longer wavelength (thus
less scattering), significantly increases penetration depth in thick samples. The intrinsic nonlinearity sharpens the focal excitation volume and leads to a corresponding
reduction of point spread function (PSF) width by a factor of
√
2 for two-photon and
√
3 for three-photon microscopy. Nevertheless, due to long excitation wavelengths,
the lateral resolution is not as good as confocal microscopy. Presently, the main
application of multi-photon microscopy is for deep-tissue imaging up to 1 mm, with
sub-micrometre spatial resolution.
One major evolution in optical sectioning technique recently is the invention
of selective plane illumination microscopy (SPIM), or more generally light sheet