The present disclosure generally relates to improving axial resolution in microscopy, and in particular to systems and methods for improved axial resolution in instant structured illumination microscopy using photoswitching and standing-wave illumination techniques.
One known method of increasing the accessible axial spatial frequencies (and thus the resolution) in conventional, widefield fluorescence microscopy is to use standing-wave illumination. In this method, two counter-propagating coherent beams are superposed at the imaging focal plane. Interference between the beams results in sharp, periodic illumination fringes with periodicity given by λ/(2 n cos Θ) where λ is the wavelength of illumination, n the index of the media and Θ the ‘crossing angle’ of the beams, i.e. the angle relative to the vertical illustrated in
In thin samples (thickness<λ), introducing a standing-wave illumination pattern as described above can yield valuable subdiffractive information. Moving the standing-wave pattern relative to the sample (i.e. altering the phase of the standing-wave pattern) causes alternating sample regions within the focal plane to glow, thereby allowing axial features finer than the axial spread of the point spread function to be discerned. However, for samples that are substantially thicker, three problems arise. First, out-of-focus fluorescence can swamp in-focus signal. Second, the repeating axial nature of the high frequency interference pattern implies an ambiguity about ‘which fringe is which’, i.e. fringes within the point spread function (PSF) create ringing artifacts in the reconstructed images (an alternative explanation of this problem is that the high frequencies of the illumination are aliased into the passband of the microscope). Third, there is an intermediate frequency ‘gap’ that exists in the reconstructed images because the frequency f of the standing wave lies outside the band limit of the microscope's optical transfer function (OTF).
Frequency gaps when using a standing wave illumination are shown in
These latter two issues are not solved in traditional 4pi microscopy, but are addressed in a I5S system, which uses a complex three-beam interference pattern, interference of both excitation and emission light, and multiple images per focal plane to ‘fill in’ the missing axial spatial frequencies and reassign them to their proper location in frequency space.
However, the I5S system introduces the following problems. First, the I5S system has so far required two-objective interferometry and a complex beam setup which makes the system difficult to align and build due to the need to maintain the optics along two separate paths (one for each objective) aligned to a spatial precision much better than λ. Second, the I5S system requires fifteen images per focal plane to achieve improved axial resolution improvement which significantly slows down the imaging process and thus far limits imaging to fixed cells. Third, no confocal pinhole is employed by the I5S system such that in densely labeled specimens Poisson noise from out-of-focus light will limit contrast in the focal plane. Finally, the beam illumination scheme of the I5S system is highly specialized since the same illumination scheme is used both for creating the axial resolution improvement and the lateral resolution improvement. Because the resolution improvement is coupled, this method is not easily adapted to confocal geometries.
It is with these observations in mind, among others, that various aspects of the present disclosure were conceived and developed.
Corresponding reference characters indicate corresponding elements among the view of the drawings. The headings used in the figures do not limit the scope of the claims.
It is well known that the axial resolution of conventional widefield fluorescence microscopy is limited to a range between ˜500-700 nm. Systems and methods that can further improve axial resolution are of great interest in fluorescence microscopy, as such improvements would allow for greater detail to be observed in biological samples. Various embodiments related to systems and methods that enable axial resolution down to ˜100 nm by acquiring only four extra images at each focal plane for a total of five images per focal plane (instead of one image) are disclosed herein that address these deficiencies. Given the modest number of additional images required to improve axial resolution of images, embodiments of the present system and method and can be applied for sustained volumetric imaging (‘4D imaging’) in live cells, which is currently not possible with other microscopy techniques. Furthermore, the present system and method is flexible and can be combined with other super-resolution microscopes that allow further improvements in lateral resolution for those types of microscopy systems. In some embodiments, the microscopy system includes a spatial light modulator positioned conjugate to the sample being illuminated for activating the sample with a standing wave. In some embodiments, a method and related system is disclosed for supplying an axial illumination structure of intermediate and finest periodicity for the standing wave. In some embodiments, a triple beam-splitting device is used to generate three mutually coherent light beams from a single light beam that interferes at the sample to produce lower spatial frequency axial fringes necessary for achieving higher axial resolution. Referring to the drawings, embodiments of a microscopy system using photoswitching and standing wave illumination techniques are illustrated and generally indicated as 100, 200, 300 and 400 in
The present system and method is directed to decoupling standing-wave illumination from fluorescence excitation and readout using a photoswitching technique, and utilizing a compact standing-wave reflector and illuminator arrangement. Together, these elements allow axial super-resolution at much higher speeds than previously possible. Although the present system and method can be applied to a large class of microscopes (e.g. spinning-disk confocal microscopes and widefield microscopes), the present disclosure describes, by way of example, the inventive concept being applied to instant structured illumination microscopy (iSIM4) since combining iSIM with photoswitching and standing-wave illumination techniques enables confocal, 3D super-resolution microscopy having ˜100 nm axial resolution and high frame rates consistent with live-cell imaging.
Using Photoswitching Technique to Decouple Standing Wave Illumination from Fluorescence Excitation and Readout
By using a reversibly switchable fluorescent molecule such as rsEGP2 and employing a standing-wave illuminator/reflector arrangement (described further below), fluorescence excitation and readout may be performed using a large variety of confocal (or other) microscope geometries (whose excitation optics and thus illumination remain virtually unchanged relative to the base microscope) as the axial resolution enhancement may be ‘added on’ to the underlying microscope. Additionally, by using an activation wavelength in addition to the typical fluorescence excitation wavelength, axial resolution is slightly improved since λactivation<λexcitation.
Referring to
As further shown, for example, an illumination beam 102A (dark blue rays) is transmitted through the objective lens 104 and coverslip 114 and mirror 106 positioned parallel to the coverslip 114 reflects the collimated beam 102B back (lighter rays). Interference between the two beams produces a standing wave pattern 116 (red lines) in the region of beam overlap. In some embodiments, piezoelectric device 110 affixed to the mirror 106 translates the mirror 106 that provides fine control of the phase of the standing-wave pattern 116.
As discussed above, a single on-axis standing-wave pattern that produces a bright/dark fringe spacing substantially less than λactivation (e.g. the 76-nm spacing mentioned above) enables higher resolution at the expense of a spatial frequency gap, which in turn leads to artifacts in the reconstructed image. However, if additional patterns with coarser spacings that lie in the intermediate frequency gap are used, the frequency gap can be ‘filled in’ as shown in
The question now is how to generate and apply additional patterns of the appropriate periodicity. According to the present system and method one simple way of altering the fringe spacing is to vary θ. For example, for 405 nm illumination (i.e. λactivation), n=1.33, θ=0 degrees implies a periodicity of 152 nm (fringe spacing 76 nm) and θ=60 degrees implies a periodicity of 304 nm (fringe spacing 152 nm). In order to quickly vary θ at the sample plane, the following illumination setup as described below was conceived.
As shown in
In some embodiments, the first and second galvanometric scanners (G1) 205 and (G2) 207 provide independent control of the position and angle of the collimated light 204 at the back focal plane 214, and thus change angle or position, respectively, in the sample plane. By varying the angle of the first galvanometric mirror scanner (G1) 205 appropriately, patterns of periodicity ranging from λactivation/2n to λactivation/(2n cos θMAX) can be created by the microscopy system 200 at the sample plane 218, where θMAX is the maximum half angle allowed by the objective lens (e.g. 64.5 degrees for a 60×, 1.2 NA water lens). By varying the angle of the second galvanometric mirror scanner (G2) 207 appropriately, the patterns may be translated at the sample plane 218, ensuring that these patterns illuminate the sample 220.
An additional advantage of this ‘single objective’ setup of the microscopy system 200 with a mirrored reflector is that the alignment of the microscopy system 200 is likely far more stable and resistant to mechanical/thermal drift than a classic 2-objective setup (as is used e.g. in I5S or 4pi microscopy systems): since a common optical path is employed for both direct and reflected beams only the sample-to-mirror distance must be kept stable to within λ. Nevertheless, the setup may benefit from an autofocus or ‘focus lock’ module (home-built or commercially available) that may be added to the objective or sample stage in some embodiments.
Finally, some embodiments for an acquisition and processing scheme capable of combining the photoswitching techniques and the illumination/reflector setup of microscopy system 200 outlined above are described in greater detail below.
The resulting five images per focal plane were found to be sufficient for markedly increasing the axial resolution of the underlying microscope, as we have verified with simulations as illustrated in
Simulations illustrating progressive improvement in axial resolution are reproduced in
In a third embodiment of the microscopy system for utilizing the photoswitching and standing-wave illumination techniques, designated 300, is shown in
The SLM 304 provides an easy and flexible method for introducing both intermediate and finer (e.g. 300 nm, 150 nm patterns in
Various SLM patterns are shown in
The acquisition procedure performed by the microscopy system 300 will be very similar to the two-galvanometer setup:
Step 1: The sample 316 is labeled with a reversibly switchable fluorescent marker such as rsEGP2.
Step 2: the sample 316 is activated with a standing wave of intermediate periodicity by using the SLM 304 to display sharp sinusoidal illumination.
Step 3: The sample 316 is imaged using the base optical microscope arrangement, e.g. the instant SIM.
Step 4: Steps 2) and 3) are repeated at two other phases of the standing wave, achieved by displaying the appropriate patterns on the SLM 304.
Step 5: The sample 316 is activated with a standing wave of maximum periodicity (i.e. collimated incident and reflected beam at Θ=0 degrees), by changing to a uniform pattern on the SLM 304.
Step 6: The sample 316 is imaged using the base optical microscope, e.g. the instant SIM.
Step 7: Steps 5) and 6) are repeated for an additional phase of the standing wave, achieved by translating the piezoelectric actuator/mirror.
Step 8: Steps 2)-7) are repeated as necessary at different focal planes in the sample, e.g. for acquiring a 3D imaging stack.
Step 9: Images are combined and deconvolved with Richardson-Lucy deconvolution to improve axial resolution.
In a fourth embodiment of the microscopy system for generating a sharp axial illumination structure for achieving axial super-resolution, designated 400, is shown in
The polarization state of the first, second and third split light beams 403A, 403B, and 403C may be controlled using a first half wave plate 404 and second half wave plate 408. The first, second, and third split light beams 403A, 403B, and 403C at the back focal plane 436 provide illumination with a sharp axial structure as shown in
The acquisition procedure performed by the microscopy system 400 will be very similar to a two-galvanometer microscopy setup or a spatial light modulator (SLM) microscopy setup:
Step 1: The sample 440 is labeled with a reversibly switchable fluorescent marker, such as rsEGP2.
Step 2: The sample 440 is activated with a standing wave of intermediate periodicity by allowing the first, second and third light beams 403A, 403B, and 403C to propagate through the microscopy system 400, thereby enabling sinusoidal illumination at the sample 440.
Step 3: The sample 440 is imaged using a base optical microscope (not shown), such as an instant selective illumination microscopy.
Step 4: Steps 2 and 3 are repeated at four other phases of the standing wave which is achieved by rotating the galvanometer mirror 426 appropriately.
Step 5: The sample 440 is activated with a standing wave of maximum periodicity, for example collimated incident and reflected at Θ=0 degrees, using the optical chopper 436 to block the outer two laser beams, e.g., first and second light beams 403A and 403C.
Step 6: The sample 440 is imaged using the base optical microscope (not shown), such as an instant selective illumination microscopy.
Step 7: Steps 5 and 6 are repeated for an additional phase of the standing wave, thereby achieved by translating the mirror 419 by using a piezoelectric actuator (not shown).
Step 8: Steps 2 through 7 are repeated as necessary at different focal planes in the sample 440, for example by acquiring a three dimensional imaging stack.
Finally, the captured images are combined and deconvolved using a Richardson-Lucy deconvolution to improve axial resolution.
It was noted that since a nonlinear transition (photoswitching) is used in the microscopy systems 100, 200, 300 and 400 disclosed herein, in theory ‘unlimited’ resolution is possible by ‘saturating’ either ON or OFF states. Achieving saturation is simple in principle by turning up the 405 nm laser would be one way of saturating the ON state, leading to higher harmonics in each axial slice; however, the price that must be paid to read out this resolution improvement would be the acquisition of more raw images, but it is in principle possible given sufficiently photo-stable samples.
In one aspect, the techniques for photoswitching and standing wave illumination described herein may be applied to other microscopy systems to improve axial resolution. For example, the aforementioned techniques may be used with any type of widefield fluorescence or confocal microscopy systems to improve axial resolution.
It should be understood from the foregoing that, while particular embodiments have been illustrated and described, various modifications can be made thereto without departing from the spirit and scope of the invention as will be apparent to those skilled in the art. Such changes and modifications are within the scope and teachings of this invention as defined in the claims appended hereto.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/US2019/039551 | 6/27/2019 | WO | 00 |
Number | Date | Country | |
---|---|---|---|
62789210 | Jan 2019 | US | |
62693750 | Jul 2018 | US |