HIGH-RESOLUTION MICROSCOPE AND METHOD FOR TWO- OR THREE-DIMENSIONAL POSITION DETERMINATION OF OBJECTS

Description

Problem description (state of the art)

[0001] In principle, the optical resolution of a light microscope, including that of an LSM, is diffraction-limited by the laws of physics. Special illumination configurations are known for achieving optimal resolution within these limits, such as 4Pi arrays or arrays with standing wave fields. These can significantly improve the resolution, especially in the axial direction, compared to a conventional LSM.

[0002] Nonlinear depopulation processes can further improve resolution compared to a diffraction-limited confocal LSM. One such method is described, for example, in US Pat. No. 5,866,911 A. Various approaches for depopulation processes are known, for example, as described in DE 44 16 558 C2, US Pat. No. 6,633,432 B2, or DE 103 25 460 A1.

[0003] In recent years, various methods have been developed and applied to overcome the diffraction limit in fluorescence microscopy (PALM Structured Illumination WO 2006 / 127692 A2; EP 1 157 297 B1). The document HUANG B ET AL: "Three-Dimensional Super-Resolution Imaging by Stochastic Optical Reconstruction Microscopy" (SCIENCE, Vol. 319, February 8, 2008 (2008-02-08), pages 810-813) shows a microscope for the STORM method.

[0004] A method for high-resolution fluorescence microscopy that is currently developing particularly advantageously is based on the highly precise localization of individual molecules. It is known that localization, i.e., the determination of the location of a single fluorescent molecule, is not subject to diffraction limitations.

[0005] This location determination can be achieved with highly sensitive wide-field cameras with an accuracy down to the nm range if sufficient photons of the molecule can be detected. In localization-based, high-resolution microscopy, an image is compiled from the molecular positions thus obtained. The crucial factor here is that only a subset of the sample's molecules is in the fluorescent state, so that, on average, the "nearest-neighbor" distance of the active molecules is always greater than the PSF (point spread function) of the microscope. This is achieved by using optically or chemically switchable fluorophores: in a densely labeled area of ​​a sample, stochastic subsets of the fluorophores in the observed area are switched to the fluorescent state by irradiating them with a suitable conversion wavelength.The density of the spots is adjusted to enable continuous localization of the molecular positions. This optical switching technique is used, for example, in PhotoActivated Localization Microscopy (PALM). This basic technique has been described in detail in various variations in the literature [1-6].

[0006] The high-resolution methods (PALM, STORM, D-STORM, etc.) differ mainly in the choice of fluorophores and the type of optical switching process.

[0007] What all methods have in common, however, is the localization of the molecules by imaging them onto a highly sensitive camera (e.g., EMCCD). The quasi-point light source (molecule) is imaged onto multiple camera pixels using the microscope's point spread function (PSF). The exact position of the molecule in the x / y plane can then be determined either by fitting the known PSF (Gaussian) or by determining the center of gravity, or by a combination of both (Gaussian mask).

[0008] Typical localization accuracies are (depending on experimental conditions) between 5 and 30 nm; this is approximately the lateral resolution of this method.

[0009] In practice, the requirement for molecules that are not too close to each other on the one hand and for the structures under investigation to be reproduced as completely as possible on the other hand means that many individual images (typically 10,000-20,000) of the sample must be taken.

[0010] This leads to considerable image acquisition time and to problems in adjusting the switching intensity, especially when the sample or the structures of interest are very inhomogeneously labeled: in order not to lose information, the switching intensity must always be adjusted to the most densely labeled areas of the sample.

[0011] In principle, the localization-based high-resolution method described above is limited to surfaces or two dimensions, since the localization of individual dye molecules in the third spatial direction (z-direction) is much more complex. The number of individual molecules required to reproduce the structures increases accordingly in the three-dimensional case.

[0012] A further problem for three-dimensional high-resolution imaging in the depth of a sample is the limited penetration depth: as is also known from classical fluorescence microscopy and laser scanning microscopy, the increasing scattering of the excitation radiation in the depth of the sample leads to an increase in the background signal while simultaneously reducing the actual useful signal.

[0013] In addition to the unwanted photobleaching of sample areas outside the focal plane, the PALM method also causes unwanted switching of the fluorophores by the activation radiation in the layers not currently being measured in the depth of the sample.

[0014] The state of the art generally requires high-resolution fluorescence imaging in three dimensions with high penetration depth and "sectioning" (i.e., measuring a layer while avoiding excitation / bleaching and, in particular, switching the photoconvertible fluorophores in layers above or below) and an increase in the acquisition speed. Summary of the invention

[0015] The invention is defined in independent claims 1 and 2.

[0016] According to the invention, it was recognized that a three-dimensional resolution-enhanced microscope can be realized through synergies of the following technologies and arrangements in the arrangements and methods described in more detail below: High-resolution determination of the axial (Z) position of molecules: Astigmatism / cylindrical lens ([9])

[0017] In this approach, a weak cylindrical lens is inserted into the detection beam path, resulting in an astigmatic PSF. Accordingly, the image of the molecule is distorted elliptically when the molecule is located above or below the symmetry point of the PSF. Information about the molecule's z-position can then be extracted from the orientation and strength of the distortion.

[0018] According to the invention, the anamorphic optics (cylindrical lens) are advantageously used to determine the vertical molecular position by shape and / or size detection and are implemented in a microscope using the following methods and arrangements. Detection in two planes (biplane detection) (Toprak et al., Bewersdorf et al.[7,8])

[0019] Here, a 50 / 50 beam splitter is inserted into the detection beam path, which splits (duplicates) the image into two partial images. These two images are projected either onto two identical cameras or side by side onto a camera chip. An optical path length difference is introduced into one of the two partial beam paths such that the two partial beam paths result in two object planes that are separated by approximately half to one z-PSF (700 nm) in the z-direction. The z-position for molecules lying between these two planes is then determined, for example, by subtracting the two partial images of the molecule, by fitting a three-dimensional PSF, or by comparable algorithms.

[0020] A splitting of the detection beam path and offset detection of the split beam paths, which have a path length difference, as well as the color-selective use for several fluorophores in several detection beam paths are realized with the other methods and arrangements according to the invention in a microscope. Penetration depth: multiphoton microscopy

[0021] The state of the art of multiphoton microscopy, in particular 2-photon (2P) excitation and its advantages are specifically used with the further described methods and arrangements according to the invention. The 2P switching of convertible dyes

[0022] The possibility of 2P excitation exists not only for the typical fluorescence excitations of known dyes, but also, in principle, for the switching transitions of so-called switchable fluorophores or photoswitches. These are fluorophores that can be switched to a fluorescent or non-fluorescent state, depending on their initial state, by irradiating them with a switching wavelength (cf. PALM, STORM, etc.). The switching wavelength can also be the same as the excitation wavelength. Such photoswitches are essential for the aforementioned high-resolution methods.

[0023] In various publications

[10] it has been reported that some switchable fluorophores can also be switched with 2-photon excitation, e.g. the proteins Dronpa, EosFP, Kaede, Kikume, PA-GFP (references 2P-Switching).

[0024] The 2P activation (switching to an activatable state) is used in particular in the point scanning mode, but also in the line scanning mode, preferably with wide-field fluorescence excitation and wide-field detection, individually or with the further described arrangements and methods according to the invention.

[0025] They are particularly advantageous in conjunction with the marking of predetermined regions (ROI), which are captured and marked, for example, in a pre-image, such as certain cells or other biological areas of interest, in particular by: AOTF control, SLM control or DMD control of the exposure to switching radiation and / or excitation radiation.

[0026] This is two-dimensional and three-dimensional (in the image stack), also in conjunction with the other listed inventive methods and arrangements in a microscope. Temporal focusing of the switching radiation

[0027] With so-called temporal focusing, the depth discrimination effect, which arises in point-scanning 2P microscopy due to the quadratic intensity dependence of the excitation in combination with strong focusing, can also be achieved in wide-field imaging. For this purpose, for example, in

[11] short laser pulses are spectrally split using a grating; this grating is then imaged via the microscope objective. As a result, the different spectral components of the pulses take different optical paths and only meet again in the focal plane, where they then form the original short laser pulse. Thus, the peak power of the pulse is only maximum in the focal plane, which, in conjunction with the aforementioned quadratic intensity dependence of the 2P excitation, leads to depth discrimination, but now in the wide-field.

[0028] This can be combined with the ROI function as mentioned above via SLM or DMD control. 2P multispot imaging

[0029] With a rasterized microlens array or rotating microlens disk, e.g. in

[14] and references therein, it is realized with the other listed methods and arrangements in a microscope. Switching and excitation via a line scanner

[0030] Is realized with the other listed methods and arrangements in a microscope.

[0031] In contrast to the arrangements and methods described below, the current state of the art has not yet been able to achieve the synergies and advantages of the invention: PALM and related techniques provide a two-dimensional high-resolution fluorescence image, but only at the coverslip interface under TIRF excitation. Bi-plane or astigmatism approaches can further increase the z-resolution; however, the measurement of thick samples is particularly problematic because only slices of the order of a PSF can be measured. Slices above or below can be measured sequentially as image stacks, as is common in microscopy; however, since this involves (laser) wide-field excitation and switching, the slices above and below are also switched and / or bleached during the measurement of the current slice and can therefore not be measured or can only be measured to a limited extent. Temporal focusing for switching the fluorophores

[12] has been demonstrated, and a sectioning comparable to 2P was observed. However, true 3D high resolution cannot be achieved with this method.2P switching using a point-scanned cw laser combined with wide-field detection has been demonstrated

[13] . While sectioning is possible, 3D resolution is not possible. For all localization-based high-resolution methods based on photoswitches, the intensity of the switching laser per image must currently be adjusted to the highest labeling density within the observed sample area. Literature:

[0032] [1] Betzig it. al., Sciene 313, 1642-1645 (2006) [2] Hess et al., PNAS 104, 17370-17375 (2007) [3] Hess et al., Biophys J. 91,4258-427 (2006) [4] Shroff et al., PNAS 104, 20308-2031(2007) [5] Rust et al., Nat Methods 3, 793-796 (2006) [6] Egner et al., Biophys J. 93, 3285-3290 (2007) [7] Toprak et al., Nano Lett. 7, 3285-3290 (2007) [8] Juette et al., Nature Methods 5, 527 (2008) [9] Huang et al., Science 319, 810 (2008)

[10] Marriot et al., PNAS 2008, Ivanchenko et al. BioPhysJ 2007, Watanabe et al. OptExp. 2007, Schneider et al., BioPhysJ 2007. Microscopy (3rd Edition) DE 19829981A1 : Regions of Interest DE 19930532A1 : SLM DE 10259443 A1 : Pulse association in the sample DE 19835072A1 : DMD US 5867604 : Structured illumination

[0033] The invention is further characterized by the features of the independent claims.

[0034] It is implemented in a high-resolution microscope and method for two- or three-dimensional position determination of objects, in particular individual fluorophores, preferably for spatially high-resolution fluorescence microscopy of a sample labelled with marker molecules which can be activated or switched with a signal in such a way that they can only be excited to emit fluorescent radiation in the activated or switched state, the method comprising the following steps: 1) Applying the signal to the sample in such a way that only a subset of the labeling molecules present in the sample are activated, wherein there are subregions in the sample in which activated labeling molecules are at least a distance from the activated labeling molecules closest to them that is greater than or equal to a length resulting from a predetermined optical resolution, 2) Excitation of the activated molecules to emit fluorescent radiation, 3) Detection of the fluorescent radiation with the predetermined optical resolution, and 4) Generation of a single image from the luminescence radiation recorded in step 3), wherein the geometric locations of the marker molecules emitting fluorescent radiation are determined with a spatial resolution increased beyond the predetermined optical resolution, wherein the steps are repeated several times and the several individual images thus obtained are combined to form an overall image.

[0035] The method according to the invention involves selective activation or switching of sample regions. In the course of the method according to the invention, multi-plane detection is performed in the Z direction, in which either an astigmatic point-image blurring function is generated or partial images with a spatial offset in a detection plane are generated. In certain embodiments, the high-resolution microscope according to the invention can also be used for confocal depth discrimination with wide-field illumination.

[0036] The background of the invention is explained in more detail below with reference to the schematic drawings.

[0037] First, they show: Fig. 1 a schematic representation of an activated labeling molecule in a resolution-limited volume, Fig. 2 a schematic representation of the imaging of various activated and non-activated labeling molecules on a spatially resolving detector, Fig. 3 a flow diagram for image generation in the PALM method, Fig. 4 the flow diagram of the Fig. 3 Explanatory representations of the detector of the Fig. 2 depicted marker molecules, Fig. 5 a schematic representation of a microscope for PAL microscopy (state of the art).

[0038] Fig. 1shows a schematic representation of a marker molecule 1 that has been excited to fluorescence. Of course, fluorescence detection requires a multitude of excitations, since each excitation produces exactly one fluorescence photon, and radiation detection requires the integration of many fluorescence photons. Due to physical principles, the fluorescence radiation emitted by the marker molecule 1 can only be detected in a microscope with a limited optical resolution. Even if the microscope reaches the diffraction limit of optical resolution, the photons of the fluorescent marker molecule 1 are still scattered due to diffraction and thus detected in a diffraction disk 2. Thus, in principle, the microscope outputs the geometric dimensions of the marker molecule 1, which are Fig. 1 schematically drawn as a black circle, represents a larger object that is Fig. 1by the diffraction disk 2. The size of the diffraction disk 2 depends on the quality of the microscopy equipment used and is defined by the half-width of the point dispersion function of the optical image. In fact, it is not a two-dimensional object, but rather a diffraction volume into which the fluorescence photons enter. In the two-dimensional representation of the Fig. 1 However, this appears as a disc. The term "diffraction disc" is therefore used here very generally to refer to the maximum resolution volume that the optics used can achieve. However, the optics used do not necessarily have to operate at the diffraction limit, although this is preferable.

[0039] In order to more precisely localize the marker molecule 1 within the diffraction disk 2, the PALM method described above is used. This activates individual marker molecules. In this description, the term activation generally refers to the activation of certain luminescence properties of the marker molecules, i.e., both the activation of the luminescence excitability and a change in the luminescence emission spectrum. s , which corresponds to the activation of certain luminescence properties. In the embodiment described here, activation is achieved by optical activation radiation. However, other non-optical activation mechanisms are also possible.

[0040] The activation now occurs in such a way that there are at least some activated molecules whose center of gravity is not in the diffraction disk of other activated molecules, ie which can at least just be distinguished within the optical resolution.

[0041] Fig. 2 shows schematically an exemplary situation on a detector 5, which integrates the photons spatially resolved. As can be seen, there are areas 3 in which the diffraction disks 2 of neighboring marker molecules overlap. Here, as in the left area 3 of the Fig. 2 However, as can be seen, only those marker molecules that were previously activated are relevant. Non-activated marker molecules 1' do not emit the specific fluorescence radiation that is captured by the matrix detector 5, and thus play no role.

[0042] In regions 4, e.g., the region 4 located in the center of the matrix detector 5, marker molecules 1 are positioned such that their diffraction disk 2 does not overlap with any diffraction disk 2 of another activated marker molecule 1. The right-hand region of the matrix detector 5 shows that regions 3 in which diffraction disks 2 of activated marker molecules 1 overlap can be adjacent to regions 4 in which this is not the case. The right-hand region 4 also illustrates that the proximity of an activated marker molecule 1 to a non-activated marker molecule 1' is irrelevant for detection, since such a marker molecule 1' does not emit the fluorescent radiation detected by the matrix detector 5, i.e., it does not fluoresce.

[0043] To capture an image that is detailed beyond the optical resolution specified by the device, which is a high-resolution image in the sense of this description, the Fig. 3 schematically illustrated steps are used.

[0044] In a first step S1, a subset of the marker molecules is activated by means of a switching signal; they are thus switched from a first state in which they cannot be excited to emit the specific fluorescent radiation, to a second state in which they can be excited to emit the specific fluorescent radiation. Of course, the activation signal can also cause selective deactivation, i.e. an inverse procedure can also be used in step S1. It is important that after step S1, only a subset of the marker molecules can be excited to emit the specific fluorescent radiation. The activation or deactivation (for the sake of simplicity, only the case of activation is described below) depends on the marker molecules used. For a dye such asDRONPA, PA-GFP or reversibly switchable synthetic dyes (such as Alexa / Cyan constructs) are activated by optical radiation, so the switching signal is switching radiation.

[0045] The under the Fig. 3 depicted Fig. 4 shows the state after step S1 in part (a). Only a subset of the marker molecules I_n is activated. The marker molecules of this subset are represented by a fully highlighted black dot. The remaining marker molecules were not activated in this step. They are shown in part (a) of the Fig. 4 denoted by l_n+1.

[0046] Activated marker molecules can then be excited to emit fluorescent radiation in a second step (S2). Fluorescent proteins known from the state of the art, such as PA-GFP or DRONPA, are preferably used as fluorescent dyes. Activation of such molecules occurs with radiation in the 405 nm range, excitation to fluorescent radiation occurs at a wavelength of approximately 488 nm, and the fluorescent radiation is in a range above 490 nm.

[0047] In a third step S3, the emitted fluorescence radiation is detected, for example by integrating the recorded fluorescence photons, so that the photons in the sub-image (b) below the Fig. 4shown situations on the matrix detector 5. As can be seen, the diffraction disks of the activated label molecules I_n do not overlap. The size of the diffraction disks is determined by the optical resolution of the image on the matrix detector 5. In addition, in part (b) of the Fig. 4 (Theoretical) diffraction disks of fluorescent molecules belonging to the non-activated group I_n+1 are shown. Since these non-activated labeling molecules do not emit fluorescence radiation, no fluorescence radiation contained in their (theoretical) diffraction disks interferes with the detection of the fluorescence radiation of the subset I_n of activated labeling molecules.

[0048] In order to ensure that as few diffraction disks as possible overlap in the subset I_n so that the labeling molecules can no longer be distinguished, the activation energy is set so that the subset I_n only represents a comparatively small proportion of the total number of labeling molecules, so that statistically many labeling molecules can be distinguished in relation to the volume that can be resolved with the optical arrangement.

[0049] In a fourth step S4, the position of the fluorescent marker molecules is calculated from the diffraction distribution of the fluorescence disks, whereby the resolution with which the position of the activated marker molecules is known is sharpened beyond the resolution of the optical arrangement, as shown in the partial image (c) of the Fig. 4 shows.

[0050] As an alternative to a computational determination, it is fundamentally possible to amplify the recorded fluorescence radiation non-linearly and thus sharpen the resolution beyond the optical arrangement with reduced effort. The nonlinear amplification can, for example, be described according to the function S = A • FN< (Equation 1) or S = A • exp F / w< (with w = 10 -N< (Equation 2)), where F is the amplitude of the fluorescence signal, A is a normalization factor and N is an integer greater than 1. A strong nonlinear dependence of the parameter S on F is particularly advantageous, for example high values ​​for N in Equations 1 or 2. Of course, other functions can also be used. In principle, the nonlinearity is preferably chosen so that the half-width of the diffraction disk corresponds to a desired spatial resolution for the location of the labeling molecules.In addition to nonlinear amplification, nonlinear attenuation can also be used. In this case, fluorescence signals of low amplitude or intensity are attenuated, while strong signals remain largely unattenuated. Of course, a combination of nonlinear amplification and attenuation can also be used.

[0051] A fifth step (S5) assembles the marker molecules, whose position has been specified, into a single image whose spatial resolution is increased beyond the optical resolution. However, it only contains information about the previously activated subset of marker molecules.

[0052] In a sixth step S6, the individual image is combined into a complete image in a known manner. The process then returns to step S1, where the previously fluorescent molecules must be deactivated again. Depending on the type of marker molecule, deactivation can be achieved by separate radiation or by decaying the activated state. It is also possible to bleach previously imaged marker molecules using excitation radiation.

[0053] With each run, a further individual image is obtained that contributes to the overall image. In the next run, a different subset of the marker molecules is activated, for example, the Fig. 4 represented subset I_n+1.

[0054] By repeatedly passing through steps S1 to S6, the overall image is constructed from individual images from the individual passes, which indicate the locations of the marker molecules with a spatial resolution that is sharper than the resolution of the optical image. Through a corresponding number of iterations, a high-resolution overall image is thus successively constructed. The reduction of the diffraction disk preferably occurs in all three spatial dimensions when multiple image stacks spaced apart in the z-direction are acquired. The overall image then contains the locations of the marker molecules with high resolution in all three spatial directions.

[0055] Fig. 5shows a schematic of a microscope 6 for high-resolution imaging of a sample 7. The sample 7 is labeled, for example, with the dye DRONPA (cf. WO 2007009812 A1). For activation and fluorescence excitation, the microscope 6 has a radiation source 8 which has individual lasers 9 and 10 whose beams are combined via a beam combiner 11. The lasers 9 and 10 can, for example, emit radiation at 405 nm (activation radiation) and 488 nm (fluorescence excitation and deactivation). Dyes are also known (e.g. the dye called DENDRA (cf. Gurskaya et al., Nature Biotech., Vol. 24, pp. 461-465, 2006)) in which activation and fluorescence excitation can occur at one and the same wavelength. In this case, one laser is sufficient.

[0056] An acousto-optical filter 12 serves for wavelength selection and for rapidly switching or attenuating individual laser wavelengths. An optical system 13 focuses the radiation via a dichroic beam splitter 14 into the pupil of an objective 15, so that the radiation from the radiation source 8 is incident on the sample 7 as wide-field illumination.

[0057] Fluorescence radiation generated in sample 7 is collected via the objective lens 15. The dichroic beam splitter 14 is designed to allow the fluorescence radiation to pass through a filter 16 to a tube lens 17, so that the fluorescent sample 7 is imaged onto the detector 5.

[0058] To control the operation of the microscope 6, a control device is provided, here embodied as a computer 18 with a display 19 and keyboard 20. Process steps S2 to S6 are performed in the computer 18. The frame rate of the matrix detector is crucial for the total measurement time, so a matrix detector 5 with the highest possible frame rate is advantageous for reducing the measurement time.

[0059] The reference symbols in Fig. 6-11 mean the following: Pr: Sample O: Lens D, D1, D2, D3, D4: dichroic beam splitters L1-L2: Light sources Bt1, Bt2: Image splitter module (multi-level detection) K1, K2: Area detector (camera) S: Scan module (schematic) with X / Y scanners Sx: One-dimensional scanner (in X direction) TL: Tube lens SO: Scan lens SLM: Spatial Light Modulator G: Grating SML: Microlens array SE: Individual sensors of a sensor arrangement B1, B2: Adjustable slit diaphragms Ld1, Ld2: Line detector ZL: Anamorphic optics such as cylindrical lens SF: Beam former for line focus (anamorphic optics)

[0060] What most drawings have in common is that, as in Figures 6-10Described using the PALM method as an example, a first laser L1 is provided for switching (activating) the dye, and lasers L2, L3 are provided for fluorescence excitation / deactivation of dyes in the sample Pr in the wide field, and one or more cameras (preferably CCDE) are provided for wide field detection. Figures 6 and 8 illustrate the embodiments according to claim 1 and claim 2, respectively. Further details are shown in Figures 12-14. Figure 7 explains the function of the device according to Fig. 6 . Figures 9-11 relate to examples that are not part of the present invention.

[0061] Included in the disclosure are further methods for the time-dependent activation / deactivation of dyes to generate high-resolution microscopic images as described above in the prior art.

[0062] Multi-level detection is, for example, an image splitter module according to Fig. 12-14 or an anamorphic image as well as designs of the state of the art ([7,8]).

[0063] "ROI" - "regions of interest" refers to areas that have been preselected automatically or manually, for example based on an overview image, and which can be selectively exposed to radiation. Bi-plane detection scheme with 2P switching by point scanner

[0064] Figure 6 shows a depth-selective 3D high-resolution fluorescence microscope according to the invention (schematic) in a 2-channel version (for 2 different, simultaneously observed fluorophores with two cameras K1, K2.

[0065] A point-scanning scanning module S follows a laser L1 in the illumination direction for switching via a 2P excitation, this can be a ps laser diode or even a cw diode laser in the range 780-830nm (or another wavelength suitable for the 2P switching process). The laser L1 is coupled into the illumination beam path via dichroic D1.

[0066] Lasers for wide-field excitation L2 and L3 in different wavelengths are coupled into the illumination beam path via D. Fluorescence detection is carried out via D2 in transmission towards the cameras K1, K2, with a splitting into two color channels for different wavelengths of two fluorophores via D3. For each color, the image can be split using image splitter modules (Bt1 and Bt2) as shown in Fig. 12-14 described.

[0067] Therefore, fluorescence excitation by laser wide-field illumination and wide-field detection by (sensitive) cameras as is usual with PALM and comparable localization-based high-resolution methods are advantageously carried out here.

[0068] The detection is advantageously carried out using a biplane (multi-plane) detection scheme [7,8], Fig.12-14 for highly precise localization of molecules also on the z-axis.

[0069] The image becomes advantageous (see Fig. 12-14 ) by the image splitter modules (Bt1 / 2 in Figure 6 ) are split in such a way that two partial images are created, each with half the intensity, whose conjugate object planes are offset by, for example, approximately half an axial PSF.

[0070] These partial images are imaged side by side onto the camera sensor (K1 / K2) and can be evaluated according to [7,8] to determine the z-position of the individual fluorophores from the 2 partial images.

[0071] A particularly advantageous embodiment of this image splitter module is the subject of Fig. 12-14 with the following advantages: Telecentricity no change in image scale as a function of z-position parallel incidence on detector (→ no z-dependent distortion of the PSF) adjustable z-splitting and thus possibility of adaptation to different objectives and embedding media possibility of setting the splitting to zero, whereby the reference image plane can be on the sample surface; the adjustable splitting then occurs into the sample (and not into the coverslip, as in known designs) possibility of removing the splitting in order to utilize normal camera imaging with a full sensor.

[0072] Particularly advantageous is the one with the described image splitter modules according to Fig. 12-14achieved depth of field in a similar range as the minimum layer thickness of the two-photon excitation (for example 700 nm) or can be achieved with the arrangement according to Fig.12-14 can be variably adjusted in this way. This results in an optimal overlap between the activated (=switched) layer and the high-resolution measured layer.

[0073] The switching (photoconversion) of the fluorophores from their non-excitable to their excitable state (prerequisite for PALM) is carried out via a focused, point-scanned excitation beam, the wavelength of which is chosen so that the switching process occurs through a 2-photon absorption process while the fluorescence excitation and detection of the thus switched fluorophores occurs in the wide field.

[0074] Due to the stochastic activation of the photoswitches in the PALM process, the rasterized activation does not necessarily have to be synchronized with the image acquisition of the camera.

[0075] 2P switching (conversion) advantageously leads to the "sectioning" known from 2-photon microscopy—that is, the selective excitation of only the molecules in the focal plane. This allows a z-resolution comparable to confocal microscopy to be achieved without the need for a confocal pinhole. This allows the method to be easily combined with the wide-field camera detection required for PALM.

[0076] In addition, in contrast to confocal detection, out-of-focus switching is avoided, which is particularly important for the acquisition of high-resolution z-stacks.

[0077] It is known that the photoconversion (switching) of the photoswitches typically used for PALM microscopy requires only extremely low intensities. Therefore, the 2P effect can also be achieved with inexpensive continuous wave lasers, which is advantageous over expensive short-pulse lasers for 2P microscopy.

[0078] For fluorophores or applications or sample preparations that require higher power for the 2P switching process, a laser diode operated in picosecond mode can be used (example example), or other cost-effective ps laser systems.

[0079] The typical 2P switching wavelengths for common PALM fluorophores are 760 - 850 nm, a range that is well covered by low-cost semiconductor lasers.

[0080] In Figure 7 In a), the switching intensity of the point-scanned switching laser is adjusted to the labeling density in the sample. In the more sparsely labeled area I, the switching intensity is increased to match the rate of localized molecules to that in the more densely labeled area II. Figure 7b) shows a schematic image stack in xz section.

[0081] By selectively switching on the scanning activation laser only in previously manually or automatically defined regions of interest, unwanted switching above and below the slice being acquired is avoided. This is particularly advantageous when acquiring z-stacks consisting of multiple such slices.

[0082] With the point-scanned switching beam, ROls (Region Of Interest - see DE 198 29 981 A1) can be defined within which the fluorophores are activated; with TIRF-PALM, however, all areas of the sample are always activated. For example, with the method described here, the switching intensity within a frame can be optimally adapted to the labeling density of the sample (cf. Figure 7aThis allows for a reduction in image acquisition time, which is particularly important for acquiring 3D image stacks. Furthermore, it also allows for the localization rates of regions in the sample labeled with different fluorophores to be aligned.

[0083] A further advantage is that unwanted switching of molecules in sample regions above and below the z-slice being imaged can be minimized (in addition to the above-mentioned intrinsic sectioning of the 2P excitation) by switching on the activation radiation only in the sample regions with colored structures ( Figure 7 b) ).

[0084] An advantageous method for a sequential process for number-controlled feedback for 2P switching can be carried out.

[0085] It enables a faster achievement of a comparable localization density in different sample areas for inhomogeneously stained samples and thus an acceleration of the acquisition time: After the camera image has been acquired (optional: real-time localization), the number or number density of the localized molecules per ROI is determined and a comparison {area ROI:number} with the PSF is carried out as follows: If {area ROI:number} >> (greater than) the PSF, the intensity of the 2P switching beam for this ROI is increased. If {area ROI:number} ~= (within the range of) the PSF, the intensity of the 2P switching beam for this ROI is decreased.

[0086] The corresponding intensity modulation is output to the scanning 2P switching beam and a next camera image is recorded. A method for ROI-controlled feedback for 2P switching:

[0087] minimizes the unwanted switching of molecules in sample areas above and below the sample plane being measured by the following steps: Define one or more ROIs automatically or manually based on an initial image acquired with high switching and / or excitation intensity, or based on an image acquired with other contrast functions (DIC). Output the corresponding intensity modulation to the scanning 2P switching beam so that the switching radiation is "on" only within the specified ROIs. Within these predefined ROIs, the switching intensity can be adjusted to the counting rate of the molecules as described in 9.3.2.

[0088] The described control of the switching intensity is particularly advantageous for samples that are stained with different fluorophores that are measured simultaneously.

[0089] In this way, the potentially different switching rates of the different molecules can be adjusted by spatially adjusting the switching intensity, thus reducing the acquisition time for multicolor measurements. The ROIs can be defined based on sample characteristics or superimposed over the image as a regular grid with fields of appropriate size; the latter is particularly advantageous for the versions with wide-field switching illumination (temporal focusing, spinning microlens disc) mentioned below. Bi-plane detection scheme with 2P switching using line scanner

[0090] To achieve the activation of a stochastic subset of fluorophores per camera frame required for PALM, a very high scanning speed of a point scanner may be required, especially for fast image acquisition rates. Using a line for 2P switching reduces the scanning speed accordingly. The reduction in peak intensity associated with the line can be accepted due to the low switching intensities required.

[0091] Instead of a point scanner in Fig. 6 For 2P excitation, a line scanner can also be used (see also Fig.11 , not part of the invention).

[0092] The low intensities required to switch the dyes also make the use of a scanned line in conjunction with the 2P switching process possible.

[0093] This line can be created, for example, with an anamorphic lens and moved through the object using a one-dimensional scanner.

[0094] The advantage is that a faster scan is possible , because only 1D scanning is required.

[0095] Since the activation occurs stochastically and is detected in the wide field, the line does not need to be synchronized with the image acquisition. Bi-plane detection scheme with 2P switching using temporal focusing

[0096] Figure 8 shows a depth-selective 3D high-resolution microscope according to the invention (schematic) with L1 as activation laser; L2,L2 as excitation laser, a grating G: grating; an SLM: spatial light modulator (see, for example, DE 199 305 32 A1); D1-D4 are dichroic beam splitters or combiners, Bt1 and 2: image splitter modules, K1 and K2 are cameras.

[0097] In Fig. 8a the beam path from the light source to the sample is shown enlarged.

[0098] The illumination beam path is drawn with a continuous line and the figure is dashed.

[0099] It also applies analogously to the representation in Fig. 9 (not part of the invention).

[0100] The objective O is schematically represented by a lens.

[0101] The grating is positioned in an intermediate image and is imaged (dashed) into the next intermediate image, which contains the SLM. This, in turn, is imaged into the sample Pr.

[0102] In contrast, the laser's illumination beam (continuous line) is "offset." In the sharp intermediate images (grating, SLM, probe Pr), the illumination spot Spot is maximally out of focus.

[0103] Here again, an ROI functionality can be implemented as shown above, with the difference that the respective switching intensity is now set via the SLM in the intermediate image.

[0104] This arrangement advantageously does not require any moving parts (non-scanning) and achieves a similar flexibility as the previously described scanning arrangements.

[0105] Depth selection is now achieved via the 2P switching effect in conjunction with temporal focusing [11 and references therein]. For this purpose, a grating G is illuminated with the expanded beam of a short-pulse laser L1 with a wavelength suitable for 2P excitation of the switching process. The illuminated region of the grating G is then imaged onto the sample Pr. The grating G splits the spectral components of the laser pulse so that the original short pulse shape is only achieved at the focus of the image (and possibly intermediate images), and accordingly, the high peak intensities for the 2P activation process are present only there.

[0106] As shown in (11), in a telescopic beam path, the short pulses are only recovered in the sample Pr (the image plane of the telescope).

[0107] An SLM as a modulator for the split wavelengths is positioned in an intermediate frame of the image, where the original pulse shape is restored.

[0108] ROIs are created by selectively switching the SLM elements.

[0109] Here, the ROI functionality can now also be implemented in wide-field excitation (in addition to switching). For this, the laser wide-field excitation must also be routed through the SLM or receive its own SLM; then, the separate definition of ROIs for excitation and switching is also possible. This implementation is available in Figure 8 for the case of 2-color excitation and detection.

[0110] As an alternative to the SLM, a DMD (Digital Micromirror Device) array can also be used; the structure is analogous to Figure 8 , but the DMD array must now be used in reflection (again in the intermediate image).

[0111] Astigmatism / cylindrical lens detection scheme with 2P switching by point scanner The setup corresponds Figure 6 , 8 (both according to the invention), 9 (not part of the invention), but with cylindrical lens(es) ZL in the detection beam path instead of Bt1 and / or Bt2.

[0112] This is in Figure 6 , 8 , 9 optionally marked by an arrow. Bi-plane detection scheme with 2P switching using a rotating microlens disk

[0113] Figure 9 (not part of the invention) shows a depth-selective 3D high-resolution microscope (schematic) with a scanning microlens disk (SML). Otherwise, the designations are as in Figure 8 .

[0114] Illumination is provided by a 2P laser L1, for example, at around 800 nm (e.g., a ps laser diode). After beam adjustment to a scanning microlens array (SML), the foci of the (rotating) microlenses are imaged onto the sample via an intermediate image containing an SLM. This creates a rapidly scanned 2P point source for switching the fluorophores with ROI functionality via the SLM.

[0115] In addition, the laser wide-field illumination for excitation of the fluorophores is coupled via dichroic 1 (D1). Detection and z-information correspond to the above examples.

[0116] Advantages are: the speed and confocal depth discrimination under wide-field illumination, combined with ROI functionality through SLM or DMD. Confocal high-resolution microscope with line scanner

[0117] The localization-based high-resolution methods described in [ ] cannot, in principle, be performed with a standard LSM: sub-PSF-accurate localization must be achieved by simultaneous spatial oversampling by the camera pixels, since the molecules only fluoresce for a limited time or are activated at unpredictable times (stochastically). This rules out a method in which the molecule's center of gravity is determined sequentially (raster method).

[0118] With the wide-field imaging practiced so far, the desirable depth selection through the confocal imaging of an LSM is no longer possible.

[0119] The arrangement presented here consists of a special line sensor ( Figure 10 , not part of the invention) and corresponding microscope ( Figure 11, not belonging to the invention) allows the advantageous combination of confocal line scanning and localization-based high resolution and thus enables a PALM-like high-resolution microscopy method in conjunction with the depth selection of a confocal microscope.

[0120] To increase the penetration depth, the line illumination for switching and / or excitation can also be achieved by a multi-photon process.

[0121] This approach differs from Fig. 6 because depth selection is achieved through confocal line imaging.

[0122] Figure 10 : a) shows a principle of localization-based confocal high-resolution microscopy with a line sensor, 10 b) a sensor with an interlocked pixel structure, 10 c) an example for the realization of an interlocked pixel structure with existing pixel geometries by masking.

[0123] Figure 10 a)(not part of the invention) explains the principle using a 2-line line sensor (for example the sensor of a line scan camera as often used in machine vision).

[0124] Lateral localization (along the row direction) can be achieved, as is well known, by center-of-gravity calculation or fitting a suitable function (1D Gaussian), taking into account the interlacing pixels of both rows. Localization orthogonal to the row direction can be achieved by calculating the difference signal of the corresponding pixels in the opposite pixel rows. The principle here corresponds to a position-sensitive detector (PSD). To ensure the PSD function, the confocal slit aperture must be opened to slightly more than one Airy unit.

[0125] The different localization accuracies in the x- and y-direction that may arise with this approach can be adapted by alternative sensor designs. Figure 10 b) (not part of the invention) shows a possible sensor structure with two interlocking combs.

[0126] To determine the y-coordinate, the determined x-position must first be evaluated and taken into account, since the interlocked pixel structure means that the PSD function (y-center of gravity determination) depends on the x-position of the molecule to be localized.

[0127] The line mapping should be chosen so that the PSF width exactly corresponds to the width of a pixel line; however, it is centered on the center of the interlocking comb structure.

[0128] Figure 10 c)(not part of the invention) shows an alternative implementation proposal based on conventional, square pixel geometries. Here, for example, a normal sensor can be rotated 45 degrees to achieve this pixel orientation. In principle, an area sensor can also be used, which is masked except for the desired line for detection. This mask can also be implemented as a variable aperture.

[0129] Several, preferably two, rows of pixels are used for evaluation. Figures 10 a) - c) each show a molecule in dashed lines, which is detected by several detector elements.

[0130] For this purpose, the existing slit aperture of a line scanner is opened slightly to capture two rows simultaneously. In 10 a ) and b ) the slit image captured in this way is shown.

[0131] In Fig. 10 c)For example, the captured area is shown somewhat brighter than the hidden area above and below it.

[0132] By mutual subtraction of advantageously four detector elements, which are located next to each other and each detect a signal, the center of gravity of the detected molecule can be determined exactly (by the magnitude and sign of the subtraction result).

[0133] In addition, there is depth information from shape and size, for example with the method cited above.

[0134] Figure 11 (not part of the invention) shows schematically a possible design of a depth-selective high-resolution microscope based on this sensor principle.

[0135] Figure 11shows a depth-selective 2D high-resolution confocal microscope (schematic) with a line scanner, exemplary for two-color detection. Switching and excitation lasers (L1-L3) are formed into a line shape using a beam shaper (SF) and scanned in one spatial direction (scanner Sx). During detection, the line is confocally imaged onto line detectors (Ld1, Ld2) through corresponding slit apertures, as described previously in the text. D1-D4 are dichroic beam splitters and combiners, respectively.

[0136] The illumination line is generated by the beam-shaping optics SF (cylindrical lens) and scanned in one spatial direction (scanner Sx). The fluorescence is confocally focused via slit diaphragms B1 and B2 (here an example with two color channels) onto the Fig. 10 sketched line detectors Ld1 and Ld2.

[0137] In Fig. 12-14 are advantageous variable image splitter modules Bt1, Bt2 (see e.g. Fig. 6 ) for z-splitting.

[0138] The reference symbols mean in detail: E1, E2: Object planes BM: Image splitter module O: Lens Dic1: Main color splitter L1: Light source EF: Emission filter TL: Tube lens SP: Deflecting mirror B: Aperture (telecentric aperture) L1, L2: Lens groups Dic 2: Color splitter for optional suppression BS: Beam splitter cube P1: Double deflection prism, adjustable perpendicular to the optical axis P2: Deflection prism DE: Wide field detector S1, S2: Sensor halves e1, e2, e3: Image planes

[0139] Fig. 12 shows a wide-field beam path with an expanded light source L1 and a spatially resolved area detector DE, for example a CCD camera.

[0140] The light from the light source L1 reaches (reflects) a sample Pr via Dic 1 and the objective O.

[0141] The reflected and fluorescent sample light passes through the objective O towards the detection.

[0142] At the splitter Dic 1 and through the filter EF, the desired light is selected, here a suppression of the reflected light, i.e. only the fluorescent light reaches the detection direction.

[0143] The sample light enters the inventive arrangement of BS, P1 and P2 via SP, L1, L2 and then onto the detector DE.

[0144] E1 and E2 are different object planes in the sample Pr.

[0145] Fig. 12a shows an enlarged view of the variable image splitter module BM according to the invention. An image of the sample Pr is split into two partial images on the detector DE via the beam splitter cube BS. The prism P1 is motor-driven vertically to the optical axis to adjust the splitting of the two object planes E1, E2. The shifted partial image is reflected back into the continuous beam path via P2 (fixed), spatially offset from the continuous beam path, i.e., laterally offset to DE.

[0146] The invention is further explained using image planes e1-e3 of one and the same object plane in Pr, for example E1: The dashed position of P1 indicates the shortest possible position of BM; the resulting image plane e3 would be located behind the detector DE.

[0147] The image plane e2 of the undeflected partial image lies on the sensor (partial area S2 of the sensor). By moving the prism P1 outward to the lower position in Fig. 12a, the path is extended and the image plane e1 moves forward (opposite to the light path).

[0148] This means that an object (molecule) located in object plane E2 will be imaged sharply on sensor half S2 and blurred on sensor half S1. The same applies to an object located in object plane E1, or between E1 and E2 (E1 would be sharp on S1 and E2 would be sharp on S2; everything else would be blurred at the same prism position).

[0149] In Fig. 12b, enlarged point images of molecules 1, 2, 3 are shown, each from the object planes E1, E2 on the sensor halves S1 and S2.

[0150] It can be seen that molecules arranged in different planes in E1 and E2 (1 in E1 and 3 in E2) are sharply detected in S1 (molecule 1) and S2 (molecule 3), while molecule 2 is equally blurred because it is obviously located between E1 and E2.

[0151] The size of the molecules on the detector sections can therefore be used to determine their exact position in the Z direction in the sample.

[0152] The drawing also shows the ray paths of two molecules, both located in object plane E2.

[0153] The aperture B defines the image section reduced to half (corresponding to the size of the sensor halves S1 and S2) and prevents light from outside this area from spilling over into the two partial beam paths.

[0154] The splitting into two partial images takes place at the 50 / 50 beam splitter cube BS.

[0155] The task of focusing the deflected partial image into an image plane different from the directly imaged partial image is fulfilled by prism P1. This prism P1 extends the focal length of the corresponding beams compared to the corresponding path in air; therefore, it is preferably made of a high-refractive-index glass to achieve the largest possible working area. These beams are reflected back parallel to the direct beam path via prism P2 and directed onto the image sensor DE in section S1.

[0156] By selecting the length of the prism, the focal length extension after the last optics / lens L2 in the prism P1 can ensure that, despite the inevitably longer path length of the deflected partial image, both partial images from different object planes E1, E2 are simultaneously sharply imaged onto the detector DE.

[0157] By moving the prism P1 perpendicular to the optical axis of the direct beam path, the z-shift of the focal planes of the two partial images (from different object planes) in the sample can be adjusted.

[0158] For example, the splitting can be advantageously adapted to different lenses.

[0159] This is now possible without refocusing the system, as the system has been designed to allow a shift of "zero" and, especially when observing at the coverslip surface, to move the second focal plane into the sample (and not into the coverslip).

[0160] In another arrangement, in which the glass path of prism P1 is omitted and only two deflection mirrors that can be moved together perpendicular to the optical axis are used, it would also be advantageous to image two different object planes sharply onto the camera. However, due to the lateral path, zeroing would not be achieved, and as the mirrors move sideways, they would move further and further into the "cover glass." To vary the second object plane in the sample chamber, the object plane splitting would then have to be adjusted via the mirrors and then refocused on the objective side so that the cover glass is again sharply imaged onto the detector via the deflection. However, this described arrangement is nevertheless within the scope of the invention disclosed here.

[0161] It is possible to replace the prism P1 with another prism with different glass paths.

[0162] The imaging of the object onto the detector by lens L2 in conjunction with the optics L1 to the left of the beam splitter Dic 2 is advantageously telecentric overall.

[0163] An adjustable, preferably rectangular aperture B in an intermediate image serves to define a rectangular image area split over P1, P2 and to suppress fluorescence and stray light from the field areas outside this new area.

[0164] For multi-color experiments, a second emission wavelength can be reflected via the color splitter Dic 2 and, using a second, preferably identical z-splitting module and another detector, this beam path can also be used for high-resolution localization.

[0165] Possible longitudinal chromatic aberrations of the objective O or other chromatic aberrations that influence the z-localization of the individual molecules can be compensated by adjusting the splitting for the second color channel.

[0166] Fig. 13a ) shows a further advantageous design of a module for splitting a camera image into 4 partial images of equal intensity for z-high resolution.

[0167] In Fig. 13 a) On the left, a design with separate mirrors and beam splitters is shown. A first splitter T1, here with a 25 / 75 split in transmission and reflection, directs part of the light onto the surface Q1 of the detector De4.

[0168] A second portion is reflected by T1 and directed via a mirror S1 toward a second splitter T2 with a ratio of 66 / 33. This splitter transmits part of the light toward surface Q2 of De4 and reflects part toward a mirror S3, which redirects the light toward the 50 / 50 splitter T3. Part passes through T3 onto surface Q3 of De4, and part is redirected via S3 toward surface Q4 of De4.

[0169] By splitting the z-planes in the monolithic design by one "cube length" of a beam splitter cube each, four uniformly shifted image or object planes are created, with quadrant Q1 of De4 seeing the image in direct throughput, quadrant Q2 seeing an image shifted by one "standard distance" (= one beam splitter cube length in the monolithic design on the right), quadrant Q3 seeing an image shifted by two distances, and quadrant Q4 seeing an image shifted by three distances.

[0170] The splitting ratios of the 3 beam splitters shown here without restriction ensure that all 4 quadrants preferably capture the same intensity.

[0171] In Fig. 13 b) is a monolithic design with 3 beam splitter cubes for T1-T3 and 3 mirrored prisms for S1-S3 shown from the side, from above and in perspective.

[0172] The arrangement in Fig. 13a ) and b) is also advantageously suitable for a telecentric detection beam path as shown by Fig. 12 described.

[0173] The advantages of the arrangement according to the invention in Fig. 13 In particular, the advantage is that four images from four sample planes provide four support points for the z-determination, thus enabling more precise z-determination in a larger working area.

[0174] The square sensor format of the highly sensitive EMCCD cameras continues to be used even more efficiently.

[0175] In Fig. 14 an advantageous combination of an execution according to Fig. 13 with movable prisms according to Fig. 12 .

[0176] On the right is a view from the sensor direction, on the left a top view.

[0177] Instead of the mirrors S1-S3 in Fig. 14 Prisms 1-3 are provided here, which define a light path parallel to the optical axis through a medium of higher optical density. Compared to Fig. 13 Additional deflection elements 1-3 are used here for re-reflection into the respective partial beam path.

[0178] Each of the three prisms 1-3, which are adjustable perpendicular to the optical axis (light direction), is for one quadrant Q2-4 of the sensor De4 ( Fig. 13 ), the fourth quadrant Q1 is illuminated by direct transmission. Each prism could now be set to a different plane displacement.

Claims

1. High-resolution microscope for two-dimensional or three-dimensional position determination of objects of a sample (Pr) for spatial high-resolution fluorescence microscopy of the sample (Pr), which is marked with marking molecules which are activatable or switchable using a signal in a manner such that they are excitable to emit fluorescence radiation only in the activated or switched state, comprising: - a first laser (L1) for activating or switching the marking molecules, wherein the first laser (L1), which serves as an activation or switching source, is a short-pulse laser for activating or for switching by means of a 2-photon absorption process; wherein - arranged in the illumination direction downstream of the first laser (L1) is a scanning module (S), which is designed as a point scanner or as a line scanner, - wherein the scanning module (S) is provided as a means for selectively activating or switching sample regions; and - wherein the microscope has means which are used to introduce the signal onto the sample (Pr) by using the scanning module (S) in a manner such that only a subset of the marking molecules present in the sample (Pr) is activated, wherein in the sample (Pr) partial regions exist in which activated marking molecules have at least a distance from the activated marking modules which are their closest neighbours that is greater than or equal to a length which results from a predetermined optical resolution, - two lasers (L2, L3), of different wavelengths, for fluorescence excitation of activated marking molecules as widefield illumination; - and a multi-plane detection in the Z-direction is carried out by virtue of the fact that in the imaging beam path a surface receiver (Kl) for detecting the fluorescence radiation with the predetermined optical resolution is present in the widefield such that an individual image is generable from the recorded fluorescence radiation, and either - an anamorphic optical unit in the form of a cylindrical lens (ZL) is arranged which brings about an astigmatic point spread function (PSF) and an image of the location of the fluorescence radiation is elliptically distorted if the imaged location of the fluorescence radiation is above or below the point of symmetry of the PSF, wherein on the basis of an orientation and degree of the distortion a piece of information about the position of the location in the Z-direction is extractable; - or at least one image splitter (Bt1) with a beam splitter for splitting the detection beam path into at least two partial beam paths is provided, whose effect causes a recorded image to be split into two partial images which have a spatial offset relative to one another in the detection plane and are imaged next to one another on the surface receiver (Kl), - wherein the microscope is configured to ascertain the geometric locations of the marking molecules emitting fluorescence radiation with a spatial resolution which goes beyond the predetermined optical resolution, to repeat the steps of activation or switching and excitation and detecting several times and to stitch together the plurality of individual images thus obtained into one overall image.

2. High-resolution microscope for two-dimensional or three-dimensional position determination of objects of a sample (Pr) for spatial high-resolution fluorescence microscopy of the sample (Pr), which is marked with marking molecules which are activatable or switchable with a signal in a manner such that they are excitable to emit fluorescence radiation only in their activated or switched state, comprising: - a first laser (L1) for activating or switching the marking molecules, wherein the first laser (L1), which serves as an activation or switching source, is a short-pulse laser for activating or for switching by means of a 2-photon absorption process; wherein - widefield illumination is provided using a grating (G) for activating or switching with the laser (L1), wherein the grating (G) is present in an expanded beam of the short-pulse laser (L1) and an illuminated region of the grating (G) is imaged into the sample (Pr) under illumination by means of the short-pulse laser (L1), wherein the grating (G) splits spectral components of the laser pulses from the short-pulse laser (L1) and radiation components split in a telescopic beam path are combined in the image plane of the telescope or in the sample (Pr) such that the original pulse shape is achieved in the focus of the image representation of the grating (G),wherein in an intermediate image either a spatial light modulator (SLM) or a DMD array (digital micromirror device array) is arranged; - wherein the spatial light modulator (SLM) or the DMD array are provided as means for selectively activating or switching sample regions; and - wherein the microscope has means which cause the introduction of the signal onto the sample (Pr) using the grating (G) in a manner such that only a subset of the marking molecules present in the sample (Pr) are activated, wherein partial regions exist in the sample (Pr) in which activated marking molecules have at least a distance from the activated marking molecules which are their closest neighbours which is greater than or equal to a length which results from a predetermined optical resolution, - two lasers (L2, L3), of different wavelengths, for fluorescence excitation of activated marking molecules as widefield illumination; - and multi-plane detection in the Z-direction is carried out by virtue of the fact that in the imaging beam path a surface receiver (Kl) for detecting the fluorescence radiation with the predetermined optical resolution in the wide field is present so that an individual image is generable from the recorded fluorescence radiation, and either - an anamorphic optical unit in the form of a cylindrical lens (ZL) is arranged which brings about an astigmatic point spread function (PSF), and an image of the location of the fluorescence radiation is elliptically distorted if the imaged location of the fluorescence radiation is above or below the point of symmetry of the PSF, wherein using an orientation and degree of the distortion a piece of information about the position of the location in the Z-direction is extractable; - or at least one image splitter (Bt1) with a beam splitter for splitting the detection beam path into at least two partial beam paths is provided whose effect causes a recorded image to be split into two partial images which have a spatial offset relative to one another in the detection plane and are imaged next to one another onto the surface receiver (Kl), - wherein the microscope is configured to ascertain the geometric locations of the marking molecules emitting fluorescence radiation with a spatial resolution that goes beyond the predetermined optical resolution, the steps of activating or switching and of exciting and detecting are repeated several times and the plurality of individual images thus obtained are stitched together into an overall image.