MICROSCOPIC METHOD AND HIGH-RESOLUTION MICROSCOPE

Description

[0001] The invention relates to a microscopic method and microscope with increased resolution. US 2009 / 0219607 A1 relates to such a device and such a method for improved microscopy of biological nanostructures.

[0002] A classic application of light microscopy for examining biological specimens is luminescence microscopy. Here, specific dyes (so-called phosphors or fluorophores) are used to label samples, such as cell parts. As mentioned, the sample is illuminated with excitation radiation, and the resulting luminescence is detected by suitable detectors. Typically, the microscope uses a dichroic beam splitter in combination with blocking filters, which separate the luminescence from the excitation radiation, allowing for separate observation. This method makes it possible to visualize individual, differently stained cell parts under the microscope. Of course, several parts of a specimen can also be stained simultaneously with different dyes that bind specifically to different structures within the specimen.This method is called multiple luminescence. It is also possible to measure samples that luminesce on their own, i.e., without the addition of dye.

[0003] Luminescence is used here, as is generally the case, as an umbrella term for phosphorescence and fluorescence, thus encompassing both processes. Where fluorescence is mentioned here, it is meant to be representative of the whole and not to be understood as a limitation.

[0004] For sample examination, laser scanning microscopes (LSMs) are also used. These microscopes image only the plane located in the focal plane of the objective lens from a three-dimensionally illuminated image using a confocal detection arrangement (then called a confocal LSM) or a nonlinear sample interaction (so-called multiphoton microscopy). An optical section is obtained, and the recording of several optical sections at different depths of the sample allows a suitable data processing device to subsequently generate a three-dimensional image of the sample, composed of these various optical sections. Laser scanning microscopy is therefore suitable for examining thick specimens.

[0005] Of course, a combination of luminescence microscopy and laser scanning microscopy is also used, in which a luminescent sample is imaged at different depth levels using an LSM.

[0006] US 5043570 describes an attempt to increase resolution by oversampling. This does not lead to a significantly improved resolution below the diffraction limit of the microscope. A resolution-enhancing method is described in M. Barth et al., "BOOSTING THE OPTICAL TRANSFER FUNCTION WITH A SPATIALLY RESOLVING DETECTOR IN A HIGH NUMERIC APERTURE CONFOCAL REFLECTION MICROSCOPE", Optik, Wissenschaftliche Verlagsgesellschaft, Vol. 96, No. 2, April 1994, pages 53-58, ISSN 0030-4026.

[0007] In principle, the optical resolution of a light microscope, including a light microscope (LSM), is diffraction-limited by the laws of physics. Special illumination configurations are known for achieving optimal resolution within these limits, such as 4π arrays or arrays with standing wave fields. These configurations can significantly improve the resolution, particularly in the axial direction, compared to a conventional LSM. Furthermore, non-linear depopulation processes can increase the resolution by a factor of up to 10 compared to a diffraction-limited confocal LSM. One such method is described, for example, in US 5866911. Various approaches for these depopulation processes are known, such as those described in DE 4416558 C2, US 6633432, and DE 10325460 A1.

[0008] Another high-resolution microscopy technique is addressed in US 5867604, in which an object with a periodic structure is scanned.

[0009] Another similar method for increasing resolution is addressed in EP 1157297 B1. This method aims to exploit nonlinear processes using structured illumination. The document cites fluorescence saturation as the nonlinearity involved. The described method claims to achieve a shift in the object space spectrum relative to the transfer function of the optical system through structured illumination. Specifically, this shift in the spectrum means that object space frequencies V0 are transmitted at a spatial frequency V0 - Vm, where Vm is the frequency of the structured illumination. Given a maximum spatial frequency that the system can transmit, this enables the transmission of object space frequencies that lie above the maximum frequency of the transfer function by the shift frequency Vm. This approach requires a reconstruction algorithm for image generation and the processing of multiple images to create a single image.Another disadvantage of this method is that the sample is unnecessarily exposed to radiation in areas outside the detected focus, as the required structured illumination permeates the entire sample volume. Furthermore, this method cannot currently be used with thick samples because extrafocally excited fluorescence reaches the detector as a background signal, thus drastically reducing the dynamic range of the detected radiation.

[0010] A method that achieves a resolution beyond the diffraction limit, independent of laser scanning microscopy, is known from WO 2006 / 127692 and DE 102006021317 A1. This method, abbreviated as PALM (Photo Activated Light Microscopy), uses a labeling agent that can be activated by an optical activation signal. Only in the activated state can the labeling agent be excited by excitation radiation to emit specific fluorescence. Non-activated molecules of the labeling agent emit no or at least no noticeable fluorescence, even after being irradiated with excitation radiation. The activation radiation thus switches the labeling agent into a state in which it can be excited to fluorescence. Other activation methods, e.g., thermal activation, are also possible. Therefore, this is generally referred to as a switching signal.In the PALM method, the switching signal is applied in such a way that at least a certain proportion of the activated labeling molecules are spaced far enough apart from neighboring activated molecules to be separated, or subsequently separable, according to the optical resolution of the microscopy. The activated molecules are thus largely isolated. After recording the luminescence radiation, the center of the resolution-limited radiation distribution of these isolated molecules is determined, and from this, the position of the molecules is calculated with higher accuracy than is actually possible with optical imaging. This increased resolution through computational determination of the center of gravity of the diffraction distribution is also referred to as "superresolution" in English-language literature.It requires that at least some of the activated labeling molecules in the sample are distinguishable, i.e., isolated, at the optical resolution used to detect the luminescence radiation. Then, for such molecules, the location can be determined with increased resolution.

[0011] To isolate individual labeling molecules, the PALM method utilizes the fact that the probability of a labeling molecule being activated after receiving a switching signal of a given intensity, e.g., a photon of the activation radiation, is the same for all molecules. By adjusting the intensity of the switching signal, and thus the number of photons incident on a unit area of ​​the sample, it is possible to ensure that the probability of activating labeling molecules present in a given area of ​​the sample is so low that there are sufficient regions in which only distinguishable labeling molecules emit fluorescence radiation within the optical resolution. By appropriately selecting the intensity, e.g., the photon density, of the switching signal, it is achieved that, as far as possible, only labeling molecules that are isolated with respect to the optical resolution are activated and subsequently emit fluorescence radiation.For these isolated molecules, the center of the diffraction-induced intensity distribution, and thus the position of the labeling molecule, is then calculated with increased resolution. To image the entire sample, the isolation of the labeling molecules from the subset is repeated by introducing activation radiation, subsequent excitation, and fluorescence imaging until as many labeling molecules as possible have been included in a subset and isolated within the resolution of the image.

[0012] The PALM method has the advantage that neither activation nor excitation requires high spatial resolution. Instead, both activation and excitation can be performed using wide-field illumination.

[0013] As a result, the labeling molecules are statistically activated in subsets by appropriately selecting the intensity of the activation radiation. Therefore, generating a complete image of a sample, in which the positions of all labeling molecules can be computationally determined with a resolution exceeding, for example, the diffraction limit, requires the evaluation of a large number of individual images. This can involve up to 10,000 individual images. Consequently, large amounts of data are processed, and the measurement takes a correspondingly long time. Acquiring a single complete image alone requires several minutes, which is essentially determined by the readout rate of the camera used. The position determination of the molecules in the individual images is performed using complex computational procedures, such as those described in Egner et al., Biophysical Journal, pp. 3285–3290, Vol. 93, November 2007.The processing of all individual images and their assembly into a high-resolution overall image, i.e., an image in which the locations of the labeling molecules are specified with a resolution beyond the diffraction limit, typically takes 1-2 hours.

[0014] WO 2007 / 009812 A1 deals with high-resolution luminescence microscopy, where a non-linearity of a dye is exploited by illuminating a special pattern that covers several surface elements of the sample simultaneously. US 2009 / 0147354 A1 relates to an illumination device in microscopy for generating an illumination pattern. WO 2004 / 053558 A1 relates to light sheet microscopy.

[0015] Further articles on high-resolution methods include: Hell, SW (2007): "Far-Field Optical Nanoscopy", Science 316, 1153 - 1158, and on SAX (Saturated Excitation) Microscopy Fujita et al., Phys. Rev. Lett. (2007), as well as Yamanaka et al., J. Biomed. Opt. (2008)).

[0016] State-of-the-art high-resolution methods have several disadvantages: For the STED method, it is the availability of dyes and the required high laser intensity. RESOLFT / GSD requires a high number of switching cycles. PAL-M / STORM has a slow image generation rate. With SAX, the dyes are severely bleached.

[0017] The invention relates to a method and a device for achieving microscopic resolution below the diffraction limit of the microscope according to the independent claims.

[0018] The invention is defined in claims 1 and 8. Preferred embodiments are the subject of the dependent claims.

[0019] The invention makes it possible to generate high-resolution images even for samples more complex than those consisting of single molecules. This is achieved by exploiting the spatial changes in gray distributions.

[0020] The process includes the following essential steps: a. The illumination (an illumination pattern) is shifted for detection with an accuracy exceeding the achievable optical resolution. b. During the shift, multiple images are acquired at the highest achievable optical resolution, each corresponding to a specific shift position of the illumination pattern. c. A high-resolution image is calculated from the detected signals, for example, a fluorescence signal, of the individual acquisitions. d. The resolution can be further improved by using nonlinear excitation and appropriate fluorescence markers.

[0021] The core idea of ​​the invention is illustrated by an illumination spot of a laser scanning microscope in Fig. 1 a) - d) The image shows the scan field of a laser scanning microscope (LSM), see also Fig. 2, is moved across the sample by suitable means (table movement, second scanner). L1 is the light spot of the LSM, which is moved across the sample in a defined scan grid. Detection values ​​are assigned to each position of, for example, a point light beam, in order to generate an image of the sample. Due to the diffraction limit of the optical imaging, the individual scan points have a minimum distance (e.g., 20 nm). Two fluorescent objects f1, f2 are shown as examples in an example sample region Pr within the sample. Fig. 1In 1a), f1 and f2 are not detected because the scanned area of ​​L is located outside the sample area Pr in which f1 and f2 are located. In 1b), shifting the scan field (the illumination grid generated by the LSM) relative to the detection area by an amount below the resolution limit, for example, 10–20 nm, shows that object f1 lies within the detection area of ​​the laser spot L and provides a detection signal. In 1c), further shifting of L1 results in the detection signal originating from f2. In 1d), objects f1 and f2 are again not detected by L. Fig. 1 The distributions of the respective illumination intensity B and detection intensity D are also shown.

[0022] The calculation method for separating / reconstructing the signal of individual objects by signal deconvolution is discussed further below. Since the illumination spot in an LSM can be adjusted with greater accuracy than the diffraction-limited resolution of the microscope, the resolution can be increased by shifting the illumination grid relative to the detection area.

[0023] In Fig. 2Figure 1 shows an embodiment of the invention. It consists of a laser scanning microscope, the basic principle of which has been described in detail in various publications, for example in DE 19702753. In this case, however, according to the present invention, the illumination light BL is directed via a first scanner SC1, lens L, deflecting mirror SP1, and a partially reflective mirror SP2 towards a second scanner SC2 (the X / Y scanner of the LSM) and focused by the latter onto a sample P via a scanning optic SCO, tube lens TL, and objective O. On the return path, the sample light passes via SC2, SP2, filter F, a pinhole optic PO, and a pinhole PH to a detector DE for confocal detection.

[0024] This combines a fast scanner SC2 (LSM) with a slow scanner SC1.

[0025] The SC1 scanner shifts the LSM's illumination grid against detection, as shown by Fig. 1As already shown, at each step of the SC1 scanner, the sample is scanned with SC2. This creates a Gaussian illumination function ( Fig. 1 The sensor is slowly moved across the sample relative to the detection point. The movement and corresponding image acquisition occur, for example, in 10 nm increments.

[0026] The number of individual positions depends on the required resolution.

[0027] The use of multiple confocal detectors simultaneously is also particularly advantageous, with one detector always operating in sync with the excitation (standard confocal) and others independently observing the sample confocally. The advantage lies firstly in the utilization of all fluorescence light, and secondly in the improvement of the signal-to-noise ratio (SNR), which can also be factored into the analysis, since multiple measurement points are detected for the same illumination.

[0028] In Fig. 3A wide-field embodiment, not implementing the invention, is shown, featuring a grating for illumination modulation. Instead of a grating, an analogous optical light pattern or interference pattern can be used. The sample P is illuminated in a wide field (with a light source not shown) via a line grating G and the objective lens O. A beam splitter ST is provided to separate the illumination and detection light. The sample light is detected with spatial resolution via filters F and interchangeable tube lenses TL using a CCD receiver DE. Illumination lines are generated on the sample by the line grating G. By moving the grating relative to the sample, a stationary scan field, represented by the illumination lines, is shifted with high precision, for example, by means of a piezoelectric actuator PZ that moves the grating. The magnitude of the shift is less than the spacing between the individual illumination lines on the sample.The modulated illumination radiation, which is determined by the known grating modulation, is changed by the grating shift in . Fig. 1 An analogous relationship to detection modulation is established, and a high-resolution image can be calculated.

[0029] Fig. 4 Figure 1 shows a non-scanned detection suppression method, which does not implement the invention. In this method, laser light is moved scanning across the sample via a dichroic mirror Fd towards a detector arrangement, preferably with two detectors CCD1 and CCD2 in transmission and reflection of a dichroic mirror Fd1, which can, for example, detect different wavelengths. It is also advantageous to arrange CCD1 and CCD2 in different z-planes to increase the z-resolution (direction of the optical axis).

[0030] As mentioned above, the resolution can be increased by shifting the scan field and the corresponding signals on detectors CCD1 and CCD2 by small amounts. This shift can be achieved, for example, by slightly moving the detection along an X-direction perpendicular to the optical z-axis, but also, for example, by moving the sample or by using a second scanner, as illustrated by... Fig. 2 described.

[0031] Fig. 5Figure 1 shows an arrangement that does not implement the invention, in which a scanned image is projected onto an area detector, such as a CCD receiver or a PMT, APD, or CMOS array. The pinhole of the scanning microscope is replaced here by reading out only specific detector pixels. Each detector subunit detects quasi-confocally. The displacement for detection is simulated here by a "shift" in the sense of switching between preferably adjacent detector elements (pixels). Displacements with respect to detection are set by suitable switching of the detector pixels, corresponding to those described above, without the need to generate any direct mechanical movement. Pixel offset advantageously creates the effect of a displacement of a detection field relative to an illumination field. This is analogous to non-scanned detection as in Figure 1. Fig. 4This is possible, for example, and advantageous instead of further movement, such as of the sample. A prerequisite for increasing the resolution is, of course (taking into account the magnification of the microscope's imaging optics), that the effective pixel size at the detector location is only a fraction of the PSF's full width at half maximum (FWHM), for example, one quarter. Increase in vertical (z) resolution:

[0032] A high resolution according to the invention can also be achieved analogously in the vertical (z) direction by creating and detecting a distance between the focal plane and the illumination plane during the scanning of the sample in a known manner in z-stacks by means of an additional displacement in the z-direction. For example, in Fig. 2A lens that can be moved in the z-direction, shown schematically using the tube lens TL with arrows in the z-direction, is provided to bring the focus, which is adjusted in the z-direction, for example by a piezo lens, into intermediate positions between the individual z-positions during image acquisition when recording z-stacks.

[0033] Generalizing this principle leads to the following method for reconstructing sample information that has an extent smaller than the diffraction limit. Basic considerations:

[0034] The intensity at each pixel can be specified in the following form: <menclose notation="box"> D x 0 ∝ ∫ − ∞ ∞ H x − x 0 ⋅ I x ⋅ c x dx < / menclose>

[0035] This involves c ( x ) to determine the concentration of fluorophores, H is the "point spread function" of the detection, I(x) is the illumination intensity at location x, which in the case of a "scanning" microscope represents the "point spread function" of the illumination.

[0036] In the case of a pixelated camera detector, the integral becomes a Riemann sum. If the excitation spot or pattern is now moved with high accuracy along the x-direction across the sample, the following system of equations is obtained: <menclose notation="box"> D 1 , j = ∑ j I 1 , j · c j · H 1 , j D 2 , j = ∑ j I 2 , j · c j · H 2 , j ⋮ D n , j = ∑ j I n , j · c j · H n , j < / menclose> with n as the index for the different lighting settings and j as the index for the detector (camera) pixel.

[0037] Since the lighting pattern and the "point spread function" HSince the underlying principles are well known, solving the highly overdetermined system of equations is relatively straightforward, for example, by pseudo-matrix inversion (the "Penrose Moore Inverse") – analogous to the "Zeiss Unmixing" algorithm. The signal-to-noise ratio of the oversampled region limits the achievable resolution. The laser scanning microscope is ideal in this respect, as the spots intrinsically reach the diffraction limit. The number of necessary iterations will be small. In general, for each measurement, the number of iterations can be optimized against the modulation depth using the signal-to-noise ratio to achieve the best possible resolution improvement.

[0038] It should be noted that the increase in resolution only occurs in the "scan" direction. Similar to the structured illumination method, scanning the sample in at least three directions allows for the reconstruction of higher-resolution image information from the measurement data. Mathematical formalism:

[0039] The one-dimensional image is given by the following integral: D x p = ∫ x ′ O x ′ E x ′ − p H x − x ′ dx ′ where O ( x ) the object represents, E ( x ) is the stimulation profile ( p is the phase, or offset, of the excitation profile in different "scans", and H ( x ) is the "Point Spread Function" (PSF) of the detection. Representation of the measurement signal in Fourier space:

[0040] D ( x,p ) can regarding the x -coordinate or Fourier-transformed with respect to the p-coordinate. Structured lighting:

[0041] If D ( x,p ) regarding x After a fourier transform, one obtains: D f ω p = FT x ′ O x ′ E x ′ − p H f ω

[0042] A product A in spatial space becomes a convolution in Fourier space: D f ω p = ∫ ω O f ω ′ E f ω − ω ′ , p dω ′ H f ω or D f ω p = O f ω ⊗ E f ω p H f ω

[0043] The discrete representation of E f< ( ω, p) In Fourier space, it looks like this: E f ω p = ∑ n e n p δ ω − ω n

[0044] This ultimately gives us: D f ω p = ∑ n e n p O f ω − ω n H f ω

[0045] If the excitation profile has a simple exponential curve, the following applies: E f ω p = Aδ ω + B ω g p δ ω + ω g + B − ω g p ω − ω g and then equation (1) becomes D f ω p = AO f ω + B − ω g p O f ω − ω g + B ω g p O f ω + ω g H f ω

[0046] Equation (2) mathematically embodies the classical concept of structured lighting, in which E f< ( ω,ρ ) defines the detection bands and the OTF detection H f< ( ω ) determines the weighting of the bands. Scanning lighting:

[0047] If D ( x,p ) regarding p After a Fourier transform, the following applies: D f x ω = FT x ′ O x ′ H x − x ′ E f ω

[0048] A product A in spatial space becomes a convolution in Fourier space: D f x ω = ∫ ω O f ω ′ H f x , ω − ω ′ dω ′ E f ω or D f x ω = O f ω ⊗ H f x ω E f ω

[0049] H f< ( x, ω) can be represented in Fourier space as a discrete series of frequency components: H f x ω = ∑ n h n x δ ω − ω n

[0050] The formula for the detected image is then: D f x ω = ∑ n h n x O f ω − ω n E f ω

[0051] In equation (3) compared to equation (1), excitation and detection are reversed.

[0052] For simple exponential excitation E f ω = Aδ ω + Bδ ω + ω g + Bδ ω − ω g Three frequencies contribute to the signal. However, since the detection OTF H f< ( x , ω Since the signal consists of a (quasi-)continuous spectrum of frequencies, all frequencies contribute to the signal, which is the sum of ω g and the highest frequency component that differs from the noise in H f< ( x , ω ) result. Local area (camera):

[0053] The signal detected at a CCD pixel is: D x k , p = ∫ x k − Δ x x k + Δ x D x , p dx = ∫ x ′ O x ′ E x ′ − p dx ′ ∫ x k − Δ x x k + Δx H x − x ′ dx = ∫ x ′ O x ′ E x ′ − p H ^ x k , x ′ dx ′

[0054] We are considering two pixels detected at the location xk and xk + Δ x: D x k p = ∫ x ′ O x ′ E x ′ − p H ^ x k , x ′ dx ′ D x k + Δ x , p = ∫ x ′ O x ′ E x ′ − p H ^ x k − Δ x , x ′ dx ′

[0055] As an object to be reconstructed O ( x ), let us consider a set of points that are as evenly distributed as possible. hi , i= 1,..., m: O x = ∑ i = 1 m a i δ x − h i where δ ( x ) the Dirac delta function is, AI are the image signals to be reconstructed. In the given model, the "over-sampling" can be quantified as m / 2. With (4), the detected pixels take on D ( xk ,p ) and D ( xk + Δ x,p ) the following form D x k p = ∑ i = 1 m a i E h i − p H ^ x k h i D x k + Δ x , p = ∑ i = 1 m a i E h i − p H ^ x k + Δ x , h i

[0056] Furthermore, if the excitation profile is a cosine and the PSF H ( x ) a Gaussian profile, then we obtain: D x k p = ∑ i = 1 m a i cos ω h i − p H ^ x k h i D x k + Δ x , p = ∑ i = 1 m a i cos ω h i − p H ^ x k + Δ x , h i with H ^ x k , x ′ = ∫ x k − Δ x x k + Δ x H x − x ′ dx = ∫ x k − Δ x x k + Δ x exp − x − x ′ 2 / 2 σ 2 dx

[0057] The standard deviation σ can be approximately described as σ = 0.21 λ / N / A , where λ the emission wavelength and N / A The numerical aperture is.

[0058] The two sets of linear algebraic equations (5) can be compared with respect to the unknown a 0 and a 1 to be solved. It should be noted that equations (5) are additionally parameterized by the phase of the excitation profile. Therefore, more linearly independent equations can be derived for different p This increases the information content of the set of equations (5) and ultimately leads to a more accurate reconstruction with better resolution. The system of equations can ultimately be expressed in matrix form: D = SA where D = D x k p 1 , D x k + Δ x , p 1 , … , D x k p n , D x k + Δ x , p n T S = cos ω h i − p 1 H ^ x k h i cos ω h i − p 1 H ^ x k + Δ x , h i … cos ω h i − p n H ^ x k h i cos ω h i − p n H ^ x k + Δ x , h i i = 1 , … , m A = a i i = 1 , … , m T and n The number of scans is given. The elements of matrix D are measured values, i.e., they have a certain error Δ. D These errors are carried over into the solution. A, so that A with an error Δ A is afflicted.

[0059] The error of A is known to be associated with the error of D linked via ΔD D ≤ κ S ΔA A where k ( S ) = ∥ S ∥∥ S -1< ∥ is the conditioning number of matrix S and ∥.∥ is the norm of the matrix.

[0060] The higher the conditioning number, the less resistant the system is to errors in the input data. One way to estimate the conditioning number is by using the singular values ​​of the matrix S: κ S = s max S / s min S where s max ( S ) and s min ( S ) the maximum and minimum singular value of S are.

[0061] The conditioning number is therefore a measure of how well the high-resolution information can be reconstructed. The derivation makes it clear that the improvement in resolution depends on the interplay of two factors: 1. The signal-to-noise ratio, similar to standard deconvolution. This means that the equations described above may no longer be recognized as linearly independent due to noise and therefore can no longer contribute to the solution. The solution loses uniqueness. 2. The steepness of the "illumination PSF." The greater the modulation, the more likely linearly independent information is obtained that can contribute to solving the system of equations.

[0062] The resolution improvement should be at least twice the confocal resolution, since scanning illumination, like structured illumination, pushes the resolution limit further with its frequencies. However, since the highest frequency that can be imaged into the sample with light through the objective lens is also diffraction-limited, the factor of two results, as above and also derived for structured illumination. However, here the starting point is the confocal resolution, not the resolution of the wide-field microscope as with structured illumination.

[0063] Initial experiments have shown resolutions of <90nm. Modulated lighting:

[0064] The described method (scanning via excitation and detection) has the disadvantage, due to the drop in PSF on both the detection and excitation sides, that with only a small overlap of excitation and detection, the measured intensity is low, and consequently, so is the signal-to-noise ratio. Illumination modulation, such as increasing the laser intensity via an AOTF or direct modulation of the laser, and in the case of pulsed illumination by adjusting the pulse rate or amplitude, or adjusting the exposure time, etc., can completely eliminate this effect and significantly improve the SNR for these data points. In this context, changing the intensity or modulation can represent an adjustable control variable in a control loop to achieve an optimal detection signal.

[0065] It is particularly advantageous to keep the signal-to-noise ratio (SNR) constant throughout the entire process, or even to increase it in areas with low overlap between excitation and detection, in order to significantly improve the deconstruction (solving the linear system of equations described above) and thus the image resolution. The system of equations for the pixels with low overlap is small because most of the detected points are not illuminated and therefore contribute nothing. Therefore, the more accurately these intensities can be measured, the better this small system of equations with only a few unknowns can be solved.

[0066] More generally, modulation by improving the SNR makes it possible to truly utilize the measured linearly independent information and not lose it in the noise.

[0067] These considerations indicate that combining the technique with DIM (Dynamic Illumination Microscopy or CLEM) is particularly advantageous, as a preset optimal SNR can be achieved point by point by controlling the illumination.

[0068] Furthermore, a combination with the following nonlinear methods is conceivable and advantageous: 1. STED / RESOLFT, since the PSF is already steeper by a factor of 2-3 or more, thus further increasing the resolution. 2. Use of photoactivatable dyes, since the PSF of activation and fluorescence excitation multiply, resulting in a steeper effective PSF. 3. All methods such as dye saturation, etc., which can also be used for nonlinear structured illumination (R. Heintzmann, TM Jovin, and C. Cremer. Saturated patterned excitation microscopy - A concept for optical resolution improvement. J. Opt. Soc. Am. A, 19(8):1599-1609, 2002). 4. Multiphoton microscopy (EP 500717 B2).

[0069] Possible advantageous and surprising implementations of the invention are, without claiming to be exhaustive: Sliding an aperture in the intermediate image: The method also works in simpler systems where the same method can be applied on the illumination side by sliding an aperture in the intermediate image. The edge of the illumination serves here as an illumination weighting and is used for the unfolding. Wide-field illumination with a sharp edge represents, in the broadest sense, a structuring of the illumination. The application of the method is also particularly advantageous in reflection (without fluorescence), since the signal-to-noise ratios are usually significantly better in reflection (for example, point scanners in materials microscopy).

[0070] Spinning Disc: Using this method in a spinning disc setup is also advantageous. In this setup, illumination is achieved solely through the disc, and detection is performed on a camera without a pinhole. Confocality can be achieved through software / image processing on the camera (virtual pinhole). Here, the accuracy of the scanning for detection and sampling can be incorporated into the disc manufacturing process, which is extremely beneficial and eliminates potential sources of error during calculation.

[0071] SPIM: The SPIM technology is described, for example, in DE 10257423 A1 and WO 2004 / 0530558 A1. The use of the inventive method results in an axial resolution gain for SPIM when scanning the light sheet. The lateral resolution increase can be achieved, for example, by structuring the light sheet, the phase of which is scanned through the sample. This results in all the typical advantages of SPIM plus increased resolution.

[0072] CARS, SHG, RAMAN: Since fluorescence and linearity of excitation are in no way prerequisites for this method, techniques such as CARS, Second Harmonic Generation, or stimulated Raman also benefit. The sample is pushed through the sample. All the typical advantages of SPIM are achieved, plus increased resolution.

[0073] CARS, SHG, RAMAN: Since fluorescence and linearity of excitation are in no way prerequisites for this method, there is also an advantage for methods such as CARS, Second Harmonic Generation or stimulated RAMAN.

Claims

1. Reflection or fluorescence microscopy method for increasing the resolution of a laser scanning microscope (LSM) when detecting an illuminated sample (P), wherein - a light spot (L1) is generated on the sample (P) by means of a point light beam, and the light spot (L1) is repeatedly shifted relative to the sample (P) in several steps, - an region (Pr) of the illuminated sample (P) is confocally imaged onto a detector device by the microscope with an optical resolution, several detection signals are recorded and a high-resolution image is calculated, - the light spot (L1) is shifted by means of a slow first scanner (SC1) and a fast second scanner (SC2), and the confocal imaging is performed only via the fast second scanner (SC2) , - the confocally imaged region (Pr) is shifted relative to the light spot (L1) with a step size that is smaller than the optical resolution, and during this displacement, the plurality of detection signals is recorded with the highest achievable optical resolution of the microscope, each of which detection signals corresponds to a specific relative displacement position relative to the light spot (L1), wherein at each step of the first scanner (SC1), the confocal image scans the sample (P) by means of the second scanner (SC2), and - the high-resolution image is calculated from the detection signals.

2. Method according to claim 1, wherein the sample (P) is excited by the light spot (L1) to emit fluorescent radiation.

3. Method according to claim 2, wherein the sample (P) is provided with nonlinearly excitable fluorescence markers and these are excited nonlinearly to emit fluorescence radiation, wherein one of the following microscopy methods is used: STED, RESOLFT, use of photoactivatable dyes, methods involving saturation of dyes, and multiphoton microscopy.

4. Method according to one of the above claims, wherein the step size is less than half the resolution.

5. Method according to one of the preceding claims, wherein the light spot (L1) scans the sample (P).

6. Method according to one of the preceding claims, wherein the light spot (L1) is created by projection of a wide-field illuminated pinhole disc.

7. Method according to one of the preceding claims, wherein the point light beam realizes a Gaussian illumination function.

8. Microscope for performing a method according to one of the preceding claims, comprising: - means for generating a light spot (L1) on the sample (P) using a point light beam, - means for confocally imaging a region (Pr) of the illuminated sample (P) with an optical resolution onto a detection unit for detecting sample light, - means for repeatedly generating a displacement of the light spot (L1) in several steps relative to the sample (P) and for generating a displacement of the confocally imaged region (Pr) relative to the light spot (L1) with a step size that is smaller than the optical resolution, wherein the means for generating the displacement of the light spot (L1) comprise a slow first scanner (SC1) and a fast second scanner (SC2) and the confocal imaging is performed only via the fast second scanner (SC2), - detection and storage means for recording a plurality of detection signals with the highest achievable optical resolution during the displacement of the confocally imaged region (Pr) relative to the light spot (L1), each detection signal corresponding to a specific relative displacement position relative to the light spot (L1), wherein at each step of the first scanner (SC1), the confocal imaging scans the sample (P) by means of the second scanner (SC2), and - means for processing the individual images and generating a high-resolution image from the individual images.

9. Microscope according to claim 8, which has a wide-field illuminated Nipkow disc for generating the light spot (L1).