Method and apparatus for imaging samples using manipulated excitation radiation

Abstract

Method for imaging a sample (14) in which - a first excitation radiation (5) is focused into a volume of the sample (14), - a first detection radiation (15) caused by the first excitation radiation (5) is detected and evaluated with regard to the formation of its wavefront, - based on the evaluation results, a second excitation radiation (11), which is focused into a volume of the sample (14), is manipulated by adjusting a spatial phase distribution of the second excitation radiation (11) in a pupil plane in order to correct determined deviations of the wavefront, - an area (20) of the sample (14) to be imaged is scanned by means of the second excitation radiation (11), and - a second detection radiation (16) caused by the second excitation radiation (11) is recorded as image data, characterized in that - the second excitation radiation (11) is directed into the sample volume in at least two partial beams (11T) into each of a spot (22) illuminated by the respective partial beam (11T) in a focal position, wherein the spots (22) are separated from each other in an object plane (13) determined by the focal position and to be detected - the second detection radiations (16) caused by the partial beams (11T) are detected separately.

Description

[0001] The invention relates to a method and apparatus for imaging samples by means of manipulated excitation radiation according to the respective preamble of the independent claims.

[0002] Laser scanning microscopy has become one of the most frequently used tools in biomedical research. It enables the acquisition of three-dimensional images of samples without the need for sectioning. Furthermore, it offers users a wide range of methods and applications that extend beyond mere imaging. A sustained trend in recent years has been the functional investigation of living tissues or entire organisms.

[0003] Laser scanning microscopes (LSMs) are required to achieve the already excellent image quality in the future even at greater sample depths and to overcome strongly scattering tissue layers.

[0004] To increase the penetration depth of the radiation into the tissue, nonlinear excitation in the near-infrared spectral range is usually used [W. Denk et al., “Two-photon laser scanning fluorescence microscopy”, Science 248, 73 (1990)].

[0005] Due to the longer wavelength of the excitation radiation, it is scattered less than shorter-wavelength radiation. Therefore, a compact focus can be created with longer-wavelength excitation radiation even at greater depths within the sample. The nonlinearity of the excitation ensures that when fluorescence events are excited, fluorescence emission is emitted only from the focus or from a small focal volume. This limitation also has the advantage that, when imaging the fluorescence as detection radiation, no pinhole is required to suppress out-of-focal components of the detection radiation in order to achieve optical sectioning. Scattering of the fluorescence photons is only problematic if they no longer remain within the aperture of the microscope system and therefore cannot be detected. The quality of the detection photon field (PSF) then plays only a minor role.Another advantage of nonlinear optical excitation is that very little fluorescence is excited outside the focal point. Therefore, the dye is not bleached in these areas, and phototoxic effects are minimal. Furthermore, long-wavelength optical radiation causes very little DNA damage.

[0006] It has been known for some time that the image quality of a nonlinear laser microscopy (LSM) can be significantly improved if the wavefront aberrations of the excitation radiation caused by scattering in the sample can be anticipated and ultimately compensated for using adaptive optical methods [MAA Neil et al., “Adaptive aberration correction in a two-photon microscope”, J. Microsc. 200, 105 (2000)]. This involves reducing the excitation volume and thus achieving greater modulation contrast. To obtain the necessary information for adjusting the wavefront manipulation, image-based approaches are frequently used, which iteratively approach image enhancement [D. Débarre et al., “Image-based adaptive optics for two-photon microscopy”, Opt. Lett. 34, 2495 (2009)]. A problem with this approach is that a large number of images must be acquired to achieve the desired image quality.Such methods are therefore usually slow and can damage the sample through repeated illumination with excitation radiation.

[0007] Another approach focuses on measuring the wavefront. This can be done interferometrically [M. Rueckel et al., “Adaptive wavefront correction in two-photon microscopy using coherence-gated wavefront sensing”, PNAS 103, 17137 (2006)] or with a Shack-Hartmann sensor (SHS) [JW Cha et al., “Shack-Hartmann wavefront-sensor-based adaptive optics system for multiphoton microscopy”, J. Biomed. Opt. 15, 046022 (2010)]. Shack-Hartmann sensors are camera systems with a microlens array (MLA) positioned in front of them. When light with a plane wavefront is shone perpendicularly onto the Shack-Hartmann sensor, a pattern is created that is formed by light spots centered on the subapertures of the MLA. Deviations of the light spots from the centers of the sub-apertures are a measure of the local gradient of the wavefront. This technology thus provides the incident wavefront for every camera image.Therefore, at best, the wavefront information update rate can be on the order of the frame rate of the SHS camera.

[0008] Solutions for correcting the wavefronts of an illumination radiation are also known from US 2016 / 0209646 A1, DE 11 2015 001 640 T5 and US 2015 / 0362713 A1.

[0009] In Wang et al. [K. Wang et al., “Rapid adaptive optical recovery of optimal resolution over large volumes”, Nat. Meth. 11, 625 (2014)] and US 9,500,846 B2, it was shown that the application of an EM-CCD-based SHS to the fluorescence emission after multiphoton excitation can be used to determine the wavefront disturbance caused by the sample and to manipulate the excitation wavefront in such a way that a significant gain in image quality can be achieved in subsequently acquired images, even for comparatively large penetration depths in the sample.

[0010] However, the well-known method of wavefront correction for excitation radiation is relatively slow and operates at high excitation radiation intensities. Therefore, it is not suitable for imaging and observing radiation-sensitive and / or rapidly changing samples.

[0011] The invention is based on the objective of proposing a method for imaging samples in which the sensitivity of the image acquisition is increased compared to the prior art and which at the same time enables a frame rate that allows observation of dynamic processes of the sample.

[0012] The problem is solved by a method and a device according to the independent claims. Advantageous embodiments of the invention are the subject of the dependent claims.

[0013] The method for imaging a sample involves focusing a first excitation beam into a volume of the sample. A first detection beam, generated by the first excitation beam, is captured and analyzed to determine its wavefront characteristics. Based on these results, a second excitation beam is manipulated to correct any detected wavefront deviations. This manipulated second excitation beam is then focused into the sample to acquire image data. Optionally, a second detection beam generated by the second excitation beam can also be manipulated. Using this manipulated second excitation beam, a region of the sample to be imaged (hereinafter also referred to as a "patch") is scanned, and the second detection beam is captured, resulting in image data of that region.Manipulation of the second excitation radiation is achieved by adjusting its spatial phase distribution in a pupil plane. Image data, as used here, refers to measured values ​​that serve as input data for generating images of the areas to be imaged.

[0014] The method is characterized by the fact that the second excitation beam, more precisely its beam, is directed into the sample volume in at least two partial beams, each illuminating a spot of light (hereinafter also referred to as a "spot") at a focal position. The spots lie in an object plane defined by the focal position and are to be detected, and are separated from one another. The second detection beams produced by each partial beam are detected separately.

[0015] The spots generated in each object plane by the focused partial beams of the second excitation beam are advantageously created to be large enough and spaced so that they are clearly separated and neither overlap nor touch. However, all spots must be located within the area or patch of the object plane currently being sampled.

[0016] The patches are determined by virtually dividing the object plane of the sample to be scanned into a number of, preferably adjacent, areas (patches). For each patch, the wavefront is analyzed, and the second excitation beam is manipulated based on the result of this analysis when the respective patch is scanned. The patch is then scanned using its partial beams, specifically manipulated for that patch. An image of the object plane to be scanned is obtained from the entirety of the patches. A two- or three-dimensional image of the sample, or parts thereof, is created by combining image data from several object planes shifted relative to each other along the optical axis of a lens (Z-direction) to form a spatial representation of the sample (also referred to as a Z-stack).

[0017] This approach allows for local correction of resulting wavefront errors. Since wavefront errors can differ in the various patches of the object plane, a global correction across all patches would be significantly less accurate than the described area- or patch-wise correction using wavefront manipulation. Furthermore, a high-quality correction of the (excitation) spot enables a high penetration depth of the excitation radiation and the formation of a sufficiently diffraction-limited focus.

[0018] The choice of spot spacing can depend on the size of the patch and the sample, the wavelengths of the excitation radiation used, and the specific configuration of the optical arrangement employed for carrying out the method. For example, a spot spacing of 240 µm can be selected in an intermediate scan image of a microscope beam path used to carry out the method according to the invention. With a second excitation radiation wavelength of 800 nm, this corresponds to ten Airy diameters. The spots are thus clearly separated from one another. To scan, for example, 64 lines per patch with 8x8 patches per image, a single patch could occupy a size of 1.9 x 1.9 mm in the intermediate image and would therefore be twice as large per dimension as, for example, a 4x4 spot pattern. With an 8x8 grid of these patches, almost the entire scan field could be covered.A sample volume or object plane would be scanned with a field number of 22.

[0019] The field number indicates the diameter of the field in the intermediate image of the optical path in millimeters. This intermediate image does not necessarily have to be physically present and refers to the so-called reference tube. The magnification specifications on the respective objectives used apply to this reference tube. For example, a field of view is scanned by an optical system with a field number of 22 if, for example, an objective with 20x magnification is used and a field with a diagonal of 22 mm / 20 = 1.1 mm is scanned in the sample plane (object plane).

[0020] The first excitation radiation can differ from the second excitation radiation and, for example, have a different wavelength and / or intensity. The respective excitation radiations can be provided by different light sources. It is also possible for the first and second excitation radiations to be the same. In such a case, it is advantageous that only one light source is required. It is in accordance with the invention that radiation from the light source is used to excite the first detection radiation and to evaluate the wavefront, thus constituting the first excitation radiation. Furthermore, the second detection radiation can be generated using radiation from the same light source, based on the previously generated partial beams as the second excitation radiation.

[0021] In order to efficiently manipulate the partial beams together, in one embodiment of the inventive method, they are superimposed on each other in a plane optically conjugate to the objective pupil and projected onto a wavefront manipulator. The partial beams can then be manipulated together using this manipulator. Apart from a spot-dependent tilt, the exact same wavefront correction is applied to all spots, since all spots originate from a common pupil. This is justified because all spots within the so-called isoplanatic patch or area scan with the previously determined correction.

[0022] In an advantageous embodiment of the method, the partial beams can create spots in the object plane, which are represented as an array with rows and columns. The scanning of a respective patch using the array can be performed along the direction of the rows and / or the columns, or along any direction.

[0023] Such an array can, for example, be implemented as a 4x4 array. In each row and each column, four spots are thus created in the respective object plane. If these arrays are shifted along, for example, a row, each spot in that row can be used to excite, for instance, a fluorescence radiation as a second detection radiation, which can then be detected by suitable detectors. The excitation can therefore be performed with a lower excitation intensity per time point compared to scanning with only one spot, in particular only a quarter. 4x4 parallelization can be used to achieve a fourfold reduction in image acquisition time compared to prior art methods, for example, to only 0.25 seconds instead of the previous one second (Wang et al., 2014). At the same time, the excitation intensity per spot can advantageously be reduced by a factor of four.The invention therefore allows for faster and more sensitive acquisition of image data.

[0024] Compared to known methods, particularly those of Wang et al., 2014, the method according to the invention achieves significant improvements. For example, applying a wavefront averaged within plausibly chosen patches to the scan of the entire patch results in a significant signal and image improvement. Furthermore, the system can be designed without significant latency. It is possible to perform a wavefront analysis with a wavefront detector, such as a Shack-Hartmann sensor (SHS), within a period of 14 ms. Simultaneously, a currently configured wavefront manipulation can be adjusted. For example, the phase pattern of a spatial light modulator (SLM) as a wavefront manipulator can be adjusted within this timeframe.

[0025] Wavefront correction is essentially determined by the frame rate of the wavefront detector camera (e.g., the SHS camera) and the update rate of the wavefront manipulator, e.g., a SLM. Currently, EM-CCD-based SHSs (e.g., Alpao's SH-EMCCD-fast-468) with a 2 kHz frame rate and 23x23 support points are already available for wavefront determination. Raw data is acquired in less than 1 ms. Future technologies such as SPAD cameras (single photon avalanche diode) and quantum image sensors (QIS; e.g., CMOS sensors) could potentially be even faster with a higher pixel count. Fast SLMs currently achieve updates in less than two milliseconds (e.g., Meadowlark HSP5120L). Therefore, the limitation of (image) data acquisition will no longer be due to the acquisition process itself.

[0026] For example, assuming a scan frequency of 2 kHz (corresponding to the "Speed ​​15" setting on the LSM 780; Carl Zeiss Microscopy GmbH), 64 image lines can be scanned within 16 ms. This makes synchronizing wavefront manipulation and patch sampling easily possible using the LSM's frame trigger. Dividing the field of view (FoV) into 8x8 patches would thus result in an image with a resolution of 512x512 pixels. The scan time for an entire frame would be a quarter of a second (compared to one second without 4x4 parallelization).

[0027] Both prior art methods and the method according to the invention require two scans or sampling of a respective area or patch. First, the mean wavefront within the patch is determined using the first detection beam in order to derive a correction term for the wavefront manipulation of the second excitation beam—and optionally, the second detection beam. Then, the actual image scan is performed using the manipulated and corrected wavefront of the second excitation beam. The total scan time for the method according to the invention is, for example, half a second per image. In contrast, the total scan time according to the prior art method is two seconds, and thus four times as long.

[0028] The method according to the invention can also be configured such that at least one partial beam has an intensity that differs from the intensities of the other partial beams. Accordingly, the spot in question is also formed by radiation with an intensity that differs from the intensity of at least some of the other spots. Such a configuration is advantageous, for example, for so-called HDR (high dynamic range) imaging.

[0029] The first excitation beam can be directed over the respective patch, and the desired wavefront data can be acquired. Subsequently, the patch is scanned using the partial beams of the appropriately manipulated second excitation beam, and image data is acquired. In a further embodiment of the method, the first and second excitation beams can be provided and directed onto the sample simultaneously. For example, in addition to an array of partial beams of the second excitation beam, a beam of the first excitation beam can be directed into a patch to be acquired in the future with a future focus position (object plane offset along the optical axis [z-direction]) or onto a patch to be acquired in the future in the same object plane. The patch to be acquired in the future is only acquired by the second excitation beam in a subsequent acquisition step, which is executed after the current acquisition step.This design advantageously allows for faster image acquisition, since the evaluation results of the wavefront of the future focus position (object plane) or of the area to be acquired in the future are already available at the end of the current acquisition step. The future acquisition step can be executed immediately following the current acquisition step, which enables a further doubling of the speed of the method according to the invention.

[0030] To carry out the method according to the invention, a device for imaging a sample is advantageously used, which has an illumination beam path and a detection beam path.

[0031] The illumination beam path comprises at least one light source for providing a first excitation beam and a second excitation beam. This at least one light source is, for example, a pulsed laser light source, which is particularly suitable for use in multiphoton excitation. Furthermore, a lens is provided for focusing the first and second excitation beams onto a plane of the sample to be scanned. The lens preferably also serves to detect the first and second detection beams. To direct the first and second excitation beams onto the relevant areas or patches and to scan the patches, at least one beam deflection unit with at least one scanner is provided. Manipulation of the second excitation beam is carried out by means of a wavefront manipulator.The first and second detection beams are rescanned by the beam deflection unit, meaning their deflection in the object plane is canceled out in the detection beam path.

[0032] A beam splitter is arranged in the detection beam path, serving to separate the first and second excitation beams from the detection beams. The first excitation beam produces the first detection beam, and the second excitation beam produces the second detection beam.

[0033] The second detection beam is captured by a detector located in the detection beam path. This detector is, for example, an image detector, which captures the second detection beam as data, specifically as image data. In addition, a wavefront detector is provided for capturing wavefronts of the first detection beam, as well as an evaluation and control unit for analyzing the captured wavefront data of the first detection beam and for controlling the wavefront manipulator based on the evaluation results.

[0034] A device according to the invention is characterized in that optical means for generating at least two partial beams of the second excitation beam are provided in the illumination beam path, and the image detector is designed for the simultaneous, separate acquisition of the image data of the detection beam of the partial beams of the second excitation beam.

[0035] The partial beams in the illumination beam path can be generated using a spot generator positioned between the laser light source and the wavefront manipulator. The spot generator can produce a dot pattern, for example, a 4x4 array. An additional SLM can be used for this purpose.

[0036] The spot generator can be configured to be switched on and off. In the switched-off state, only one beam enters the illumination beam path, which in this operating state serves as the beam of the first excitation radiation. During the scanning of a respective patch of the sample with the second excitation radiation, the spot generator is switched on. The existing beam of the first excitation radiation is either switched off or, possibly in a modified form, incorporated as one of the partial beams of the second excitation radiation into the pattern of spots to be generated.

[0037] Alternatively, one of the partial beams can be defocused with respect to the object plane currently being detected and instead focused as the first excitation beam on an object plane to be detected in the future. Such defocusing of this partial beam used as the first excitation beam can be generated using the wavefront manipulator.

[0038] In another embodiment of the device, partial beams can also be generated from a beam of the second excitation radiation using classical beam splitting with dielectric layers or beam splitting in waveguides.

[0039] For simultaneous detection of the second detection radiations caused by the partial beams, for example a multianode photon multiplier (MA-PMT) or a SPAD camera (single photon avalanche diode) can be used as an image detector.

[0040] To manipulate the second excitation beam with only one wavefront manipulator, an optical unit (relay optics) is advantageously present in the illumination beam path. This unit superimposes the partial beams in a plane optically conjugate to the objective lens's pupil and images them together onto the wavefront manipulator. Using relay optics, two conjugate planes can be generated. The objective lens pupil plane is imaged onto the manipulator plane, for example, the SLM plane. This allows the phase distribution of the second excitation beam to be manipulated there as if it were happening in the pupil. The image inversion caused by the relay optics is not a problem, but it must be taken into account with regard to the spatial distribution of the phase pattern.

[0041] In one embodiment of the device according to the invention, the beam deflection unit comprises a quasi-static scanner for deflecting the first and second excitation beams in a first direction and a second quasi-static scanner for deflecting the first and second excitation beams in a second direction orthogonal to the first direction. Additionally, a resonant scanner can be provided for deflecting the first and second excitation beams in either the first or the second direction. This embodiment is advantageous because, due to the acceleration of wavefront analysis with the wavefront detector, spatial averaging of the wavefront over the patch is no longer possible with a quasi-static galvanometer scanner. The scanners are preferably imaged onto one another and arranged in a plane conjugate to the objective pupil.

[0042] With an embodiment of the beam deflection unit according to the invention, comprising an additional resonant scanner, a larger section of the object plane can be scanned during the integration time of, for example, an SHS camera. The resonant scanner performs the line scan. The other of the two quasi-static scanners, which also deflects in the line direction, is used to select the respective patch and defines only an offset angle. Due to its higher sampling frequency, scanning with a resonant scanner inherently accelerates the image acquisition by, for example, a factor of 4. To avoid degrading the signal-to-noise ratio (SNR) due to the reduced pixel dwell time, the 4x4 spot pattern can be used entirely for data averaging.

[0043] In further embodiments of the invention, the patches can also be swept multiple times by the spots in the column direction to improve the SNR. As already described, the use of resonance scanners is also possible to move the spots across the patches, thereby achieving a 4x increase in data acquisition speed. This embodiment can also be combined with the previously described multispot acquisition, resulting in ultra-fast (16x) data acquisition and image capture.

[0044] To protect the sample from unnecessary illumination with excitation radiation, the intensity of the excitation radiation can be dampened, for example by means of an acousto-optic modulator (AOM), during the time between data acquisition with the wavefront detector and image acquisition.

[0045] The detection beam path must fulfill two functions. First, the initial detection beam must be directed to a wavefront detector. Second, the subsequent detection beam must reach an image detector. Various technical measures are possible to achieve these two functions. These measures either expose both detectors to portions of both detection beams or separate and direct the initial and subsequent detection beams.

[0046] For example, a variable beam splitter can be arranged in the detection beam path. This can be designed, for instance, as a filter wheel with a neutral-dividing gradient track on a constant radius. This allows the splitting ratio between image acquisition and wavefront detection to be optimized. The filter wheel can also be designed as a color splitter and exhibit a color gradient, for example, a spatial variation of the filter edge of a long-pass filter. The filter edge is preferably adjusted so that just enough of the long-wavelength end of the dye spectrum is directed to the wavefront detector to perform the wavefront evaluation correctly. The remaining portion of the signal, particularly that from the intensity maximum, is directed to the image detector for image acquisition.Furthermore, with multiply doped samples, it is also possible to adjust the color splitter so that parasitically excited dyes propagate to the wavefront detector while the desired signal is detected. Additionally, the waveforms can be applied to linearly displaceable substrates, or splitter matrices can be used. The aforementioned configurations of the detection beam path can advantageously be variably adjustable to allow manual and / or automatic adaptation to the current operating conditions of the device.

[0047] Feeding only the first detection beam to the wavefront detector and only the second detection beam to the image detector can be achieved, for example, by means of an aperture, an optical wedge, or a mirror in the detection beam path. If an aperture is used, it can be positioned in the detection beam path upstream of the image detector at the location of an intermediate image, particularly a relay optic, to suppress out-of-focal components of the first and second detection beams. The aperture diameter is approximately the same as the chosen spacing between the spots.

[0048] In a further embodiment of the device, an additional wavefront manipulator is provided in the detection beam path, positioned upstream of the image detector. This manipulator manipulates the second detection beam based on the evaluation results of the first detection beam. This increases the focusability of the fluorescence beam of the second detection beam and significantly reduces crosstalk between the second detection beams of the respective spots. This embodiment advantageously increases the parallelizable penetration depth of the imaging method and the microscope or device according to the invention.

[0049] All existing wavefront manipulators are connected to an evaluation and control unit, which in turn receives data from the wavefront detector. The evaluation and control unit can include an evaluation unit designed to read the wavefront data from the wavefront detector, process the data, and determine a phase pattern for wavefront correction. The determined phase pattern can be transferred by the evaluation unit to the wavefront manipulator, so the evaluation unit can function as a wavefront control system. The control unit can be, for example, a system PC. This PC reads the image data from the image detector and controls image acquisition and wavefront correction.

[0050] Furthermore, it is advantageous to arrange a pinhole in an intermediate image between the beam deflection unit and the image detector, particularly preferably in the intermediate image of the relay optics, which significantly suppresses out-of-focus light to increase contrast. The greater the penetration depth of the excitation radiation into the sample, the higher the excitation intensities must be to generate a signal at that depth. According to Lambert-Beer's law, the excitation intensity decreases exponentially with penetration depth, which is compensated for by increasing the excitation radiation power with precisely this functional profile as it increases with depth. However, this in turn means that two-photon excitation without focusing the radiation can also occur in the upper sample layers. These undesired fluorescence emissions must be suppressed.

[0051] The raw data and / or results of the wavefront analysis can be saved and used for subsequent processing of the acquired image data. For example, this data can be used for image unfolding. Spot-dependent tilt can also be taken into account. Furthermore, it is possible to catalog the determined correction data for different sample types and make it available for later applications. Similarly, currently determined correction data can be compared with existing data from a catalog. If a catalog of correction data is available, some images can be corrected using the catalog data, thus saving time and reducing sample stress.

[0052] The advantages of the invention lie in faster image acquisition combined with high penetration depth of the excitation radiation and high signal-to-noise ratios. The inventive method and device allow the recording of dynamic processes even in samples that are sensitive to illumination with excitation radiation and can potentially be damaged by it.

[0053] The invention is explained in more detail below with reference to illustrations and exemplary embodiments. These show: Fig. 1 a schematic representation of a scan head and a stand of a microscope as well as a scanning scheme of selected areas (patches) of an object plane according to the state of the art; Fig. 2 a schematic representation of a first embodiment of a device according to the invention with a variable beam splitter and a scanning scheme of selected areas (patches) of an object plane; Fig. 3a and Fig. 3b Schematic representations of arrays of spots in a patch, each of one object level; Fig. 4 a schematic representation of a second embodiment of a device according to the invention with an aperture in the detection beam path; Fig. 5 a schematic representation of a third embodiment of a device according to the invention with an optical wedge aperture in the detection beam path; Fig. 6 a schematic representation of a fourth embodiment of a device according to the invention with a mirror in the detection beam path; Fig. 7 a schematic representation of part of the beam path of an optical device with a further wavefront manipulator in the detection beam path as a fifth embodiment of a device according to the invention; and Fig. 8 a schematic representation of a sixth embodiment of a device according to the invention with two quasi-static and one resonant scanner.

[0054] In the following, reference symbols used in all figures denote the same technical elements, unless expressly stated otherwise in individual cases.

[0055] The prior art has revealed the construction of a microscope 1 which comprises as essential units a scan head 2 and a stand 3 ( Fig. 1) A first excitation beam 5 and, later in the process, a second excitation beam 11 from a light source 4 in the form of a pulsed laser, for example a femtosecond laser, are supplied to the illumination beam path of the scan head 2 and directed via a mirror 6 onto a wavefront manipulator 7 in the form of an SLM. From the wavefront manipulator 7, the first excitation beam 5 passes via a main color splitter 8 and a system of optical lenses 9 to a beam deflection unit 10. This unit contains, for example, two quasi-static scanners 10.1 and 10.2 (not shown), by means of which the first excitation beam 5 and the second excitation beam 11 are deflected, or can be deflected, two-dimensionally. The wavefront manipulator 7 is arranged optically conjugate to the pupil of a lens 12, as are the scan mirrors 10.1 and 10.2, which are advantageously imaged onto each other, and optionally 10.3 (see Fig. 8) The excitation beams 5 and 11 are directed by the objective lens 12 into an object plane 13 (see additional figure) of a sample 14 to be detected. The first excitation beam 5 causes a first detection beam 15 to be generated in the sample 14, and the second excitation beam 11 causes a second detection beam 16 to be generated, both of which are detected by the objective lens 12. The first and second detection beams 15 and 16 are, in particular, fluorescence radiation. From the objective lens 12 to the main color splitter 8, the illumination beam path and the detection beam path coincide. The detection beams 15 and 16, respectively, are scanned by the scanners of the beam deflection unit 10. At the main color splitter 8, the detection beams 15 and 16 are separated from the respective excitation beams 5 and 11, respectively, and directed to a polarizing beam splitter 17.One output of the polarizing beam splitter 17 is directed to a photomultiplier (PMT) as an image detector 18 for detection. The second output is fed to a wavefront detector 19 in the form of a Shack-Hartmann sensor (SHS) for wavefront analysis. The detection radiation 15, 16 is thus scanned. Since the detection radiation 15, 16 is unpolarized in the form of fluorescence radiation, 50% of the radiation is used for wavefront analysis, while the other 50% is detected as image data.

[0056] The additional figure illustrates the scanning scheme or procedure. The scan field (Field of View, FoV) is virtually divided into areas or patches 20, the size of which depends, for example, on the type of sample 14. This is often referred to as an isoplanatic patch 20. If the wavefront deformation changes very rapidly due to scattering in the sample 14, then the patches 20 are chosen to be very small; if the changes occur rather slowly—for example, in the case of relatively homogeneous, sparsely populated samples 14—then the patches 20 can be chosen to be larger.

[0057] In the Fig. 2 and Fig. Figure 3a shows a first embodiment of the device according to the invention. In the illumination beam path, a spot generator 21 is arranged between the light source 4 and the wavefront manipulator 7. This spot generator produces a number of partial beams 11T of the second excitation beam 11, which are displayed in the object plane 13 in the form of a dot pattern or array of spots 22. An array can, for example, have spots 22 in four rows and four columns (see Figure 3a). Fig. 3a and Fig. 3b). For the sake of clarity, the additional representation of the sampling scheme of Fig. Only spots 22 in a 2x2 array are shown as examples. The spot generator 21, and thus the generation of the partial beams 11T, can be switched on and off by means of a control unit 24.2 or an evaluation and control unit 24. The partial beams 11T superimpose on each other in a plane optically conjugate to the objective pupil. This plane is imaged onto the wavefront manipulator 7 by means of a relay optic 25 for the purpose of wavefront manipulation and, as a result, wavefront correction. Subsequently, the partial beams 11T of the now manipulated second excitation beam 11 are directed further towards the beam deflection unit 10, by means of which the spots 22 of the partial beams 11T are moved as a point matrix across the sample 14 during image acquisition.

[0058] The distances between the spots 22 in the FoV are chosen so that they are clearly separated from each other, but still significantly smaller than the dimensions of a patch 20 ( Fig. 3a).

[0059] In the first embodiment ( Fig. 2) is in the detection beam path instead of a polarizing beam splitter 17 (see Fig. 1) A variable beam splitter 23 is arranged. This is designed as a filter wheel with a neutral-dividing gradient track on a constant radius. This allows the division ratio between image data detection and wavefront correction data acquisition to be adjusted. The variable beam splitter 23 can also have a color gradient, which is formed, for example, by a spatial variation of a filter edge of a long-pass filter. The filter edge is adjusted so that just enough of the long-wavelength end of the dye spectrum is directed to the wavefront detector 19 to correctly perform the wavefront evaluation. The remainder of the signal, in particular that from the intensity maximum, is fed to the image detector 18 for image data detection. In this embodiment, the image detector is a multi-anode photomultiplier (MA-PMT).

[0060] In further embodiments, it is also possible with multiply doped samples 14 to adjust the variable beam splitter 23 so that parasitically excited dyes propagate to the wavefront detector 19, while the desired signal is detected with the image detector 18.

[0061] Even when carrying out the inventive method with a device according to the Fig. 2. The data acquisition for wavefront manipulation and the acquisition of image data are performed in two successive scan steps. For this purpose, the generation of additional partial beams 11T is deactivated when scanning patch 20 for wavefront evaluation. The first excitation beam 5 is directed onto patch 20 in only one beam or beam bundle, thus scanning it. This is particularly important when a Shack-Hartmann sensor (SHS) with a front-mounted microlens array (MLA) is used as the wavefront detector 19.

[0062] If the wavefront evaluation were performed using a large number of partial beams 11T, the different wavefront tilts of the many spots 22 would generate multiple signals per microlens. Therefore, wavefront evaluation and reconstruction by manipulation would not be unambiguous.

[0063] The acquisition of wavefront data based on the first detection radiation 15, its evaluation and conversion into control commands takes place in the interaction of the wavefront detector 19 and the evaluation and control unit 24. The latter can be functionally subdivided into an evaluation unit 24.1 and a control unit 24.2.

[0064] The evaluation unit 24.1 is connected to the wavefront detector 19 and the control unit 24.2 in a manner suitable for data transmission. The latter is connected to the wavefront manipulator 7, the beam deflection unit 10, the image detector 18, the spot generator 21, and the variable beam splitter 23 in a manner suitable for data transmission.

[0065] The evaluation unit 24.1, for example, features an FPGA (field programmable gate array) which uses the image data from the SHS camera to determine the deviations of the signal spots generated by the MLA from their expected target positions by means of center-of-gravity evaluation. These deviations yield the local wavefront tilt. The results for all support points supplied by the wavefront detector 19 are interpolated onto the usable grid of the wavefront manipulator 7, and the phase values ​​calculated in this way are subtracted from a phase pattern stored on the wavefront manipulator 7. If necessary, a correction factor is also taken into account, which rescales the measured phase pattern from the detection wavelength at the wavefront detector 19 to the excitation wavelength. The algorithm thus determines the duration of the wavefront correction.However, this process can be highly parallelized in an FPGA, thus also yielding a result within a few milliseconds. This offers the potential for a 2-3 factor acceleration in determining the wavefront correction. Of course, the calculation can also be performed on a PC or a GPU / CPU, provided these are fast and sufficiently powerful.

[0066] Fig. 3b In a further embodiment of the device and a further development of the method, the first excitation beam 5 is generated in addition to an array of partial beams 11T and forms a spot 22-5 of the first excitation beam 5 in the object plane 13 next to the exemplary 4x4 array shown. This beam of the first excitation beam 5 can be generated by means of a further light source 4 (not shown) or also by means of the wavefront manipulator 7. The beam can then be geometrically positioned so that the first detection beam 15 can be separated from the second detection beam 16 in the detection beam path with as little conflict as possible (see also Fig. 5 and Fig. 6).

[0067] It is particularly advantageous if the beam of the first excitation radiation 5 is additionally or alternatively defocused by the wavefront manipulator 7, so that the spot 22-5 already scans a patch 20 of an object plane 13 to be detected in the future and an evaluation of the wavefront of this patch 20 to be detected in the future is already carried out and stored.

[0068] In a second embodiment, the previously described arrangement has been supplemented by an aperture 26 in the intermediate image ZB in front of the wavefront detector 19. The aperture 26 is large enough to allow the aberrated spot image of the second detection beam to be transmitted to the wavefront detector 19, but small enough to block all other imaging of the spots 22 onto the wavefront detector 19. Based on the values ​​discussed above as examples, the upper limit of the aperture radius is five Airy diameters. Since only one spot 22 is now imaged onto the wavefront detector 19, there is no longer any need to use the spot generator 21 (see Fig. 2) to have to switch off. Therefore, a separate spot generator 21 can be completely dispensed with. The distribution of the excitation intensity to the partial beams 11T and the resulting spots 22 is achieved via appropriately programmed phase gratings on the wavefront manipulator 7, which are additively superimposed on the wavefront to be corrected.

[0069] In a third embodiment, the variable beam splitter 23 is completely omitted ( Fig. 5) Instead, a beam-deflectoring optical means 27 is present in the detection beam path near an intermediate image ZB. Through its action, a first detection radiation 15 originating from a partial beam 11T is deflected from the propagation direction of the first detection radiation 15 and directed to the wavefront detector 19. In the illustrated embodiment, the beam-deflectoring optical means 27 is an optical wedge. The deflected portion of the first detection radiation 15 can be directed onto the wavefront detector 19 by means of an optical lens 9.

[0070] Instead of the wedge, a mirror 6 can also be used for deflection in other possible versions of the device ( Fig. 6) The first detection radiation 15 is reflected back with an offset angle and the resulting pupil plane is offered to the wavefront detector 19.

[0071] In another embodiment of the device, a further wavefront manipulator 7.1 is arranged upstream of the image detector 18 in the detection beam path ( Fig. 7) Only a portion of the illumination beam path with the wavefront manipulator 7 and a portion of the detection beam path with the image detector 18 are shown. Upstream of the latter is the further wavefront manipulator 7.1, which can also be configured as an SLM. The further wavefront manipulator 7.1 is designed to manipulate the second detection beam 16 based on the evaluation results of the first detection beam 15 (see above). For this purpose, the further wavefront manipulator 7.1 is connected to the control unit 24.2 via a data transmission link (indicated) and can be controlled by it. The further wavefront manipulator 7.1 serves to correct the wavefront of the second detection beam 16 before it reaches the image detector 18.

[0072] In Fig. Figure 8 schematically shows scanners of a beam deflection unit 10. An exemplary beam, represented as an arrow, strikes a first quasi-static scanner 10.1, which deflects the beam in the object plane 13 in the direction of the y-axis (see Figure 8). Fig. 3a and Fig.3b) deflects the beam and causes a line feed. The beam then falls onto a second quasi-static scanner 10.2, which deflects the beam in the direction of the X-axis. The second quasi-static scanner 10.2 merely sets an offset angle and, in conjunction with the first quasi-static scanner 10.1, selects a patch 20. The actual line scan along the X-axis is performed by deflecting a resonant scanner 10.3 downstream of the second quasi-static scanner 10.2. The quasi-static scanners 10.1, 10.2, and the optional resonant scanner 10.3 are controlled by the control unit 24.2 (shown schematically). Reference sign 1 microscope 2 scan heads 3 Tripod 4 light sources 5 first excitation radiation 6 mirrors 7 Wavefront manipulator 8 main color dividers 9 optical lenses 10 units for beam deflection 10.1 First quasi-static scanner 10.2 Second quasi-static scanner 10.3 Resonant Scanner 11 second excitation radiation 11T partial beam 12 Lens 13 Object level 14 Sample 15 first detection radiation 16 second detection radiation 17 polarizing beam splitter 18 Image detector 19 Wavefront detector Area 20, Patch 21 Spot generator 22 Spot 22-5 Spot of the first excitation radiation 5 23 variable beam splitter 24 control unit 24.1 Evaluation unit / WF-Control 24.2 Control unit / Control PC 25 Relay optics 26 aperture 27 wedge

Claims

[1] Method for imaging a sample (14) wherein - a first excitation radiation (5) is focused into a volume of the sample (14), - a first detection radiation (15) caused by the first excitation radiation (5) is detected and evaluated with regard to the formation of its wavefront, - based on the evaluation results, a second excitation radiation (11), which is focused into a volume of the sample (14), is manipulated by adjusting a spatial phase distribution of the second excitation radiation (11) in a pupil plane in order to correct determined deviations of the wavefront, - an area (20) of the sample (14) to be imaged is scanned by means of the second excitation radiation (11), and - a second detection radiation (16) caused by the second excitation radiation (11) is captured as image data, characterized by , that - the second excitation radiation (11) is directed into the sample volume in at least two partial beams (11T) into each of a spot (22) illuminated by the respective partial beam (11T) in a focal position, wherein the spots (22) are separated from each other in an object plane (13) determined by the focal position and to be detected - the second detection radiations (16) caused by the partial beams (11T) are detected separately. [2] Method according to claim 1, characterized by , that the partial beams (11T) are superimposed in a plane optically conjugated to the objective pupil, imaged onto a wavefront manipulator (7) and manipulated together by it. [3] Method according to claim 1 or 2, characterized by, that the object plane (13) of the sample (14) to be detected is divided into a number of areas (20); the wavefront is evaluated for each area (20); the second excitation radiation (11) is manipulated based on the result of the evaluation and the area is scanned using the partial beams (11T) of the manipulated second excitation radiation (11). [4] Method according to claim 3 characterized by , that the spots (22) of the partial beams (11T) in the object plane (13) are effected in the form of an array with rows and columns and that the sampling of each area (20) is carried out using the array. [5] Method according to any one of the preceding claims, characterized by , that at least one partial beam (11T) has an intensity that differs from the intensities of the other partial beams (11T). [6] Method according to any one of the preceding claims, characterized by, that the first excitation radiation (5) is directed into a future object plane (13) or onto a future area (20) to be detected, which is only occupied by the second excitation radiation (11) in a future detection step, which is carried out after a current detection step, so that the evaluation results of the wavefront of the future object plane (13) or of the future area to be detected (20) are already available at the end of the current detection step. [7] Method according to claim 6, characterized by that the future data collection step will be executed immediately following the current data collection step. [8] Method according to any one of the preceding claims, characterized by , that the second detection radiation (16) is manipulated by means of a further wavefront manipulator (7.1) before it is detected by means of an image detector (18). [9] Device for imaging a sample (14), comprising in a lighting beam path - at least one light source (4) to provide a first excitation radiation (5) and a second excitation radiation (11), - a lens (12) for focusing the first and second excitation radiation (5, 11) into an object plane (13) of the sample (14) to be detected, - a beam deflection unit (10) by means of which the object plane (13) to be detected is or can be scanned with the first and the second excitation beam (5, 11), - a wavefront manipulator (7) for manipulating the second excitation radiation (11), and in a detection beam path - a beam splitter for separating first and second excitation radiation (5, 11) on the one hand from a first detection radiation (15) caused by the first excitation radiation (5) and a second detection radiation (16) caused by the second excitation radiation (11) on the other hand, - an image detector (18) for capturing the second detection radiation (16) as image data, and - a wavefront detector (19) for detecting wavefronts of the first detection radiation (15), as well as - an evaluation and control unit (24) for evaluating recorded data of the wavefronts of the first detection radiation (15) and for controlling the wavefront manipulator (7) based on the evaluation results, characterized by , that optical means for generating at least two partial beams (11T) of the second excitation beam (11) are present in the illumination beam path, and the image detector (18) is designed for the simultaneous, separate acquisition of the image data of the detection radiation (16) of the partial beams (11T) of the second excitation radiation (11). [10] Device according to claim 9, characterized by , that an optical unit is present, by whose action the partial beams (11T) are superimposed on each other in a plane optically conjugated to the pupil of the objective (12) and are jointly imaged onto the wavefront manipulator (7). [11] Device according to claim 10, characterized by , that the optical means for generating at least two partial beams (11T) can be switched on and off and in the switched-off state only one beam enters the illumination beam path, which is provided as the beam of the first excitation radiation (5). [12] Device according to any one of claims 9 to 11, comprising either a first quasi-static scanner (10.1) for deflecting the first and second excitation radiation (5, 11) in a first direction and a second quasi-static scanner (10.2) for deflecting the first and second excitation radiation (5, 11) in a second direction orthogonal to the first direction, or the first quasi-static scanner (10.1) and the second quasi-static scanner (10.2) as well as a resonant scanner (10.3) for deflecting the first and second excitation radiation (5, 11) in the first or in the second direction. [13] Device according to any one of claims 9 to 12, characterized by , that optical means are arranged in the detection beam path, by whose action a portion of the first detection radiation (15) is directed onto the wavefront detector (19). [14] Device according to any one of claims 9 to 13, characterized by, that in the detection beam path in front of the image detector (18) at the location of an intermediate image an aperture (26) is arranged to suppress extrafocal components of the detection radiation (15, 16). [15] Device according to any one of claims 9 to 14, characterized by , that in the detection beam path, a further wavefront manipulator (7.1) is located upstream of the image detector (18), which is designed to manipulate the second detection beam (16) based on the evaluation results of the first detection beam (15).

Images