Microscope, method for determining field inhomogeneity in field of view of microscope, and microscopy method

Abstract

The invention relates to a microscope, a method for determining field inhomogeneity in a field of view of a microscope, and a microscopy method. A microscope includes a light source, an illumination beam path that directs excitation light to a sample, a detector that detects emitted light emitted by the sample, a detection beam path that includes a microscope objective that directs the emitted light to the detector, a mechanical driver that sets a relative lateral position between the sample and the microscope objective, and a control unit. The microscope is characterized in that the control unit is configured to perform: a setting step in which the mechanical drive is sequentially set to at least three different relative lateral positions; a collection step in which measurement data is collected from the detector for at least a subset of points in the sample in the field of view detecting the beam path, where, for each of the points in the subset, the measurement data is collected for at least two different lateral positions of the mechanical drive; and an evaluation step in which field inhomogeneity in the field of view and microscopic sample information are extracted from the measurement data.

Description

Technical Field

[0001] This invention relates to a microscope, a method for determining field inhomogeneities in the field of view of the microscope, and microscopic methods. Background Technology

[0002] A general-purpose microscope includes a light source for supplying excitation light, an illumination beam path for guiding the excitation light into a sample space, a detector for detecting emitted light emitted by a sample in the sample space, a detection beam path including a microscope objective for guiding the emitted light to the detector, a mechanical actuator for setting the relative lateral position between the sample and the microscope objective relative to the optical axis of the microscope objective, and a control unit configured to control the mechanical actuator and for collecting and evaluating measurement data from the detector.

[0003] A general method for determining field inhomogeneities in the field of view of a microscope includes the following steps: illuminating a sample with excitation light through an illumination beam path of the microscope; guiding emitted light from the sample through a detection beam path including a microscope objective to a detector; detecting the emitted light with the detector; and setting a specific lateral position between the sample and the microscope objective relative to the optical axis of the microscope objective by means of a mechanical actuator.

[0004] Modern microscopy has seen significant improvements in available spatial bandwidth products (SBPs), designed to deliver increased fields of view simultaneously with high spatial resolution, often involving megapixel to billion-pixel throughput. The need for large SBPs presents challenges to available hardware, such as larger image sensors and the necessity of aberration-corrected objectives for large fields of view. Imaging over large fields of view tends to exhibit field inhomogeneities. A first example of inhomogeneity is vignetting, caused by several factors: 1) Light emanating from off-axis points within the field of view tends to be blocked more by obstructions such as apertures and lens holders than on-axis points. 2) The illumination profile itself is already inhomogeneous, for example, because it is generated from single-mode fiber, or because condenser optics exhibit vignetting. Other effects, such as LED illumination or angle-dependent sensitivity of detector sensitivity, can also cause vignetting. A second example of field inhomogeneity is caused by field obstructions that impede any plane conjugate to the sample plane—such as field stops or dust spots on camera sensors. While vignetting can be described by a function that changes slowly or smoothly across the field of view, dust spots, field stops, and edges are described by abrupt changes and cannot be easily described by elementary functions, making it challenging to model, detect, and remove them using parametric regression.

[0005] There are key differences between prospective and retrospective approaches for correcting for irradiation inhomogeneities [4]. Prospective correction strategies use a calibration target to measure shading caused by irradiation and variable detector sensitivity [1]. This step is performed before measuring the actual sample of interest. In some cases, prospective shading methods fail to correct for shading effects caused by the sample [1], such as when using immersion objectives or when the sample is embedded in an aqueous medium around it, resulting in wavefront deformation meniscus caused by the cohesion of the liquid and the sample container (e.g., a multi-well plate). Therefore, it is not always possible to capture sample of interest data under the same conditions under which the calibration target is measured. Furthermore, calibration targets may undergo bleaching and degradation over time due to routine laboratory use (scratching and dust accumulation). Finally, calibration targets result in additional costs that must be passed on to the customer.

[0006] Several retrospective methods have been proposed for inhomogeneity correction, which involve signal post-processing strategies. For example, Preibisch et al. [2] disclosed an image stitching module as part of the open software package Fiji. The authors mitigated shading between adjacent tiles by nonlinearly weighting pixels in overlapping regions. While this technique produces perceptually attractive images, particularly mitigating hard edges between neighboring tiles, the nonlinear weights used for edge blending break the linearity between multiple fluorescence channels. This is undesirable for further post-processing that depends on the linearity of the results, such as fluorescence demixing or color deconvolution. The same drawback applies to the method proposed in [3], which also applies nonlinear blending. Furthermore, the method described in [3] is ill-posed, meaning that there is less data than the unknown pixels together with the illumination and shading profile. Therefore, the method reported in [3] may only work if certain prior assumptions about the shape and smoothness of the shading profile are satisfied.

[0007] Another routine for estimating field inhomogeneity is based on averaging over many patches. The underlying assumption in this retrospective statistical approach is that the occurrence of sample features is uniformly distributed across patches. Under this assumption, averaging over many patches should average the sample contributions (which are patch-dependent) and converge to a useful estimate of field inhomogeneity (which is patch-independent) at each sensor pixel location. However, statistical arguments show that this averaging strategy is inefficient: hundreds of samples (which in this case would be equivalent to patches) are required to converge to the mean of a uniform random variable with an accuracy better than 1%. Refinements of this averaging approach have been reported in the literature. Smith et al. [4] described a statistical setting for the retrospective approach called CIDRE (using a corrected intensity distribution that minimizes regularized energy). C orrected I ntensity Distributions using R egularized E The method of Smith et al. [4] aims to extract robust mean estimates from image sequences. Unlike the naive averaging method mentioned above, the method uses robust statistics to remove outliers.

[0008] Another recently published method is BaSiC[5], an algorithm for background and shadow correction. This algorithm assumes that the set of all patches of the sample exhibits 1) a low-rank structure and 2) that the shadow function is smooth across the field of view. Summary of the Invention

[0009] It can be considered that the object of the present invention is to specify a microscope, a method for determining field inhomogeneities in the field of view of the microscope, and a microscopic method, wherein an improvement in accuracy, at least relative to the determined field inhomogeneities, can be achieved.

[0010] This objective is achieved by the microscope, the method for determining field inhomogeneities in the field of view of the microscope, and the microscopic method according to the present invention.

[0011] According to the present invention, a universal microscope as described above is further developed, wherein a control unit is configured to perform the following steps: a setup step, wherein mechanical actuators are sequentially set to at least three different relative lateral positions; a collection step, wherein, at each of the different relative lateral positions, measurement data is collected from the detector for at least a subset of points in the sample in the field of view of the detection beam path, wherein for each point in the subset, measurement data is collected for at least two different lateral positions of the mechanical actuators; and an evaluation step, wherein, based on the measurement data collected in the collection step, the following steps are performed: extracting field inhomogeneity in the field of view from the measurement data using the assumption that field inhomogeneity is independent of the separately set relative lateral positions, and extracting microscopic sample information from the measurement data representing a portion of the measurement data caused by the sample.

[0012] According to the present invention, the general method described above is further characterized by: a setting step, wherein the mechanical actuator is sequentially set to at least three different relative lateral positions; a collection step, wherein in each of the different relative lateral positions, measurement data is collected from the detector for at least a subset of points in the sample in the field of view of the detection beam path, wherein measurement data is collected for at least two different lateral positions of the mechanical actuator for each point in the subset; and an evaluation step, wherein, based on the measurement data collected in the collection step, the following steps are performed: extracting field inhomogeneity in the field of view from the measurement data using the assumption that field inhomogeneity does not depend on the separately set relative lateral positions, and extracting microscopic sample information from the measurement data representing a portion of the measurement data caused by the sample.

[0013] The present invention further claims a microscopy method comprising the following steps: a determination step, wherein a sample is used to determine field inhomogeneity in the field of view of a microscope according to the invention using a method for determining field inhomogeneity in the field of view of a microscope according to the invention; a microscopy step comprising performing the method for determining field inhomogeneity in the field of view of a microscope according to the invention for a sample different from the sample used in the determination step, and wherein the field inhomogeneity determined in the determination step is used as an initial estimate of the field inhomogeneity.

[0014] The microscope can be any type of light microscope. For example, the microscope can be at least one of the following: wide-field microscope, scanning microscope, confocal microscope, light field microscope, light sheet microscope, TIRF microscope, SIM microscope.

[0015] In principle, the light source can be any source capable of providing excitation light with one or more desired wavelengths and suitable intensity. For example, the light source can be a laser or an LED module. The excitation light can, in principle, be coherent or at least partially coherent light.

[0016] Excitation light is electromagnetic radiation, particularly in the visible and adjacent spectral ranges. Excitation light can also be referred to as illumination light, and for most purposes, the two terms are used synonymously throughout this specification. Light emitted by the sample to be examined due to radiation from the illumination or excitation light is called emitted light and reaches the detector, such as a camera, via the detection beam path. For light to qualify as emitted light, it is only necessary that the light originates from the illuminated sample. Typically, emitted light can be fluorescence emitted by the sample, particularly dye molecules present therein, due to radiation from the excitation light. Emitted light can also be reflected, transmitted, and scattered illumination light. The only requirement set forth by the principles of the present invention for providing contrast is that the sample emits emitted light due to radiation from the excitation light. Typically, emitted light is fluorescence emitted by the sample, particularly dye molecules present therein, due to radiation from the excitation light. The sample can also be referred to as a specimen.

[0017] The term "illumination beam path" refers to all optical beam guiding and beam-modifying components, such as lenses, mirrors, prisms, gratings, filters, apertures, and beam splitters, by means of and via these components, guiding excitation light from a light source to the sample to be examined. The illumination beam path may include an illumination objective. In each case, the illumination objective and the microscope objective can be microscope objectives of a known type. In principle, the illumination objective and the microscope objective can be separate objectives. However, in a preferred embodiment, the illumination objective and the microscope objective are the same objective.

[0018] The term sample space refers to the spatial region where a sample to be studied can be arranged. In a typical embodiment, the sample can be placed or mounted on an xy-stage, for example by means of a sample holder or sample frame, which can be manipulated in the lateral direction relative to the optical axis. A z-drive may be present to change the distance between the sample stage and the illumination objective or between the sample and the microscope objective.

[0019] In principle, the sample can be of any kind. The microscope and method of this invention are particularly suitable for studying biological samples.

[0020] The light emitted by the sample to be inspected due to the radiation of the excitation light is called the emitted light, and it reaches the detector, such as a camera, via the detection beam path.

[0021] The term "detection beam path" refers to all the beam guiding and beam-modifying optical components, such as lenses, mirrors, prisms, gratings, filters, apertures, and beam splitters, by means of and via these components, guiding the emitted light from the sample to be examined to a distance as far as the camera. Conveniently, the sensor plane of the detector can be arranged in a plane optically conjugate with the focal plane of the microscope objective.

[0022] The field of view of the detection beam path represents the lateral spatial portion in the sample plane from which emitted light is collected and propagated to the detector. The size of the field of view is typically determined by the optical parameters of the microscope objective and further components in the detection beam path, such as the tube lens—e.g., numerical aperture and magnification.

[0023] The type of detector used to detect the emitted light typically depends on the type of microscope. In embodiments of the invention, the detector may include, for example, a two-dimensional spatially resolved photodetector for a camera, a one-dimensional spatially resolved photodetector for example, a linear detector array, or a single photodetector for example, a point photodetector. More specifically, the detector may include at least one of the following: a CCD element, a CMOS element, a SPAD element, or a PMT.

[0024] The mechanical actuator used to set the relative lateral position between the sample and the microscope objective can be configured to move the sample relative to the optical axis of the microscope objective and / or to move the microscope objective relative to the sample and thus move its optical axis. The mechanical actuator can be mounted to a microscope stand. The microscope stand can be an upright stand as well as a reverse stand.

[0025] The term "control unit" refers to all hardware and software components that interact with the components of the microscope according to the invention, ensuring the intended functionality of the components. Specifically, the control unit may include a computing device, such as a PC, and a camera controller capable of reading out measurement signals. Measurement data from the detector is generated by the detector when it emits emitted light.

[0026] A subset of points in the sample from which measurement data is obtained can be defined by a subset of the camera's pixels when the camera is used as a detector. In the case of a scanning microscope, the subset of points can be defined by a subset of the scanning positions of the scanning unit, and additionally by a subset of the pixels of the pixelated detector.

[0027] The term field inhomogeneity should generally be understood as explained above, that is, it should include inhomogeneities caused by vignetting and inhomogeneities resulting from field barriers, specifically those that obstruct any plane conjugate to the sample plane. The term field inhomogeneity should also cover inhomogeneities in the detection intensity distribution, i.e., the detection profile. The latter will come into play, for example, in the case of a camera where pixels have slightly different detection properties.

[0028] Vignetting can occur when light emanating from off-axis points within the field of view tends to be blocked more by obstacles such as apertures and lens holders than on-axis points. Vignetting can also be caused by inherent non-uniformities in the illumination distribution, such as vignetting originating from single-mode fiber and / or condenser optics. Finally, vignetting can be caused by other effects such as LED illumination or the angle-dependent sensitivity of the detector. Field obstacles that block any plane conjugate to the sample plane can be, for example, field barriers or dust spots on a camera sensor.

[0029] The assumption that field inhomogeneity is independent of the separately set relative lateral positions is physically perceptible because the constraints described that cause the inhomogeneity do not depend on the relative position of the sample with respect to the optical axis of the microscope objective.

[0030] The term "pattern" refers to a single image captured relative to a specific lateral position of the sample with respect to a microscope objective. A stitched image represents a combination of multiple patches collected in each case as the sample is translated laterally and / or axially. Stitching patches essentially increases the field of view. In this context, stitching means registering measurement data of individual patches in each case based on their relative positions. Typically, adjacent patches are scanned with overlapping to allow for registration and alignment of joint features.

[0031] A fluorescence channel is defined as a setting of a microscope, which includes a specific wavelength of excitation light entering the sample space and specific emission spectral filtering characteristics of the detection beam path. As is generally known in the art, these parameters can be suitably applied, for example, to the fluorophore used and the sample being examined.

[0032] The method described here is diametrically opposed to statistical methods based on averaging many patches, and is fundamentally different from the method reported by Smith et al. [4]. Instead of assuming statistical independence between patches, the microscope and method according to the invention require high overlap between individual patches, and therefore a certain degree of correlation. Instead of hoping that naive averaging or robust statistics converge to the true mean of the non-uniformity profile, the method according to the invention explicitly models the precise translation of the sample and uses the assumption that the fluorophore density should remain constant during sample translation. Thus, if signal changes and bleaching are ignored, the measured changes in intensity during sample translation provide information about the non-uniformity profile.

[0033] Widely used shadow correction methods [1] through [5] do not employ explicit modeling of signal changes during sample translation to extract field inhomogeneities. Extraction of shadow information by explicitly modeling intensity changes within highly overlapping patches is unique to the microscope and method according to the invention. The microscope and method according to the invention can particularly operate with as few as three patches and a minimum overlap of 50%.

[0034] The microscope, and in particular the control unit, can be configured to perform at least one of a method for determining field inhomogeneities in the field of view of the microscope according to the invention and a microscopic method according to the invention.

[0035] In a preferred embodiment, the method according to the invention further includes using a microscope according to the invention.

[0036] One of the key advantages of this invention is that it can capture not only the non-uniformity of illumination caused by vignetting, but also the non-uniformity caused by dust spots, field barriers and edges with high accuracy.

[0037] A further important advantage of this invention is that it can use any sample to determine irradiation non-uniformity. Unlike other methods, specific test samples are not required.

[0038] In contrast to methods reported, for example, in [3], the methods described in this application do not require satisfying prior assumptions about the shape and smoothness of the shadow distribution.

[0039] Furthermore, unlike the algorithms for background and shadow correction described in [5] where the prior smoothness assumption is necessary, the microscope and method according to the invention use an explicit physical model of signal intensity change when the sample is laterally translated relative to a non-uniform illumination distribution.

[0040] A key advantage of the microscopy method of this invention is that the determination of illumination non-uniformity can be performed temporally, i.e., with high accuracy using light-insensitive samples before or after investigating specific samples, such as light-sensitive biological samples. This is generally feasible because, typically, any fluctuations in the light source are negligible.

[0041] Typically, this invention requires setting at least three relative lateral positions during the setup step, and recording measurement data for each point in a subset of points at at least two different relative lateral positions. In a preferred embodiment of the invention, the set relative lateral positions are located on a two-dimensional regular grid. The grid can be, for example, rectangular or triangular.

[0042] The requirement to record measurement data for at least two distinct relative lateral positions for each point in a subset of points can preferably be achieved in a preferred embodiment of the invention, wherein the control unit is further configured such that, in the two-dimensional grid for setting relative lateral positions in the setting step, the overlap between adjacent tiles in the first coordinate direction is at least 50%. In a further preferred embodiment of the invention, the control unit is further configured such that, in the two-dimensional grid for setting relative lateral positions in the setting step, the overlap between adjacent tiles in the second coordinate direction, which may be particularly perpendicular to the first coordinate direction, is at least 5% and preferably at least 10%.

[0043] In a further preferred embodiment, the points in the subset are uniformly distributed across the field of view. For example, the points in the subset can lie on a regular grid. The grid of the subset can be rectangular or triangular.

[0044] In cases where a high level of accuracy is required to determine field inhomogeneity, the collection step preferably includes collecting measurement data for each point in the field of view.

[0045] In cases where field inhomogeneity needs to be determined quickly, the evaluation process may include binning of measurement data from multiple points, particularly multiple pixels.

[0046] In cases where the accuracy required to quickly determine field inhomogeneity and / or simply to determine a moderate level is sufficient, evaluating only a subset of the measurements is adequate. However, if a high level of accuracy is required to determine field inhomogeneity, the evaluation step preferably includes evaluating the measurements of each point in the field of view.

[0047] In a further preferred embodiment, the control unit is configured to perform the evaluation step as an iterative solution to a double-blind estimation problem based on initial estimates of field inhomogeneities and initial estimates of microscopic sample information.

[0048] More specifically, the control unit can be configured to perform at least some of the following steps in the evaluation step:

[0049] • Based on measurement data and using an initial estimate of field inhomogeneity, estimates of microscopic sample information are calculated, and

[0050] • Based on measurement data, and in each case using the most recent estimate of microscopic sample information and the updated estimate of field inhomogeneity, the updated estimate of field inhomogeneity is iteratively calculated.

[0051] • Assess the accuracy of the most recent estimates of field inhomogeneities and microscopic sample information in reproducing the measurement data.

[0052] • Repeat the iterative steps of calculating updated estimates of field inhomogeneity and microsample information until the measurement data can be reproduced to the specified accuracy level using the most recent estimates of field inhomogeneity and microsample information.

[0053] Additionally or alternatively, the control unit may be further configured to perform at least some of the following steps during the evaluation process:

[0054] • An initial estimate of microscopic sample information is calculated based on measurement data and using an initial estimate of field inhomogeneity;

[0055] • Calculate the first updated estimate of field inhomogeneity based on the initial estimate using measurement data and microscopic sample information;

[0056] • The first update estimate of microscopic sample information is calculated based on the measurement data and using the first update estimate of field inhomogeneity;

[0057] • The (n+1)th update estimate of field inhomogeneity is calculated based on the nth update estimate using measurement data and microscopic sample information; and

[0058] • The (n+1)th update estimate of microsample information is calculated based on the measurement data and using the (n+1)th update estimate of field inhomogeneity.

[0059] An initial estimate of field inhomogeneity can be, for example, a flat distribution (profile).

[0060] In a further embodiment, the control unit may additionally or alternatively be further configured to perform the evaluation step as minimizing the mathematical distance between the combination of measurement data and field inhomogeneity and microscopic sample information, wherein the mathematical distance is based on an arbitrary mathematical norm.

[0061] In a further embodiment, the control unit can be additionally or alternatively configured to perform the following steps in the evaluation step: if the value of a first scalar cost function for the new estimate of field inhomogeneity and the new estimate of micro-sample information is less than the value of a first scalar cost function for the most recently updated estimate of field inhomogeneity and the most recently updated estimate of micro-sample information, and if the value of a second scalar cost function for the new estimate of field inhomogeneity and the new estimate of micro-sample information is less than the value of a second scalar cost function for the most recently updated estimate of field inhomogeneity and the most recently updated estimate of micro-sample information, then the new estimate of field inhomogeneity is used as the new updated estimate of field inhomogeneity, and the new estimate of micro-sample information is used as the new updated estimate of micro-sample information, wherein in each case the first scalar cost function and the second scalar cost function contain a mathematical distance between the measurement data and the combination of field inhomogeneity and micro-sample information, and wherein the first scalar cost function contains the mathematical norm of field inhomogeneity and the second scalar cost function contains a mathematical norm of field inhomogeneity. A mathematical norm that contains information about microscopic samples.

[0062] The combination of field inhomogeneity and microscopic sample information can be, for example, the product of field inhomogeneity and microscopic sample information.

[0063] A mathematical norm can be one of the following norms: Lp norm, L2 norm (Euclidian norm), Manhattan norm (L1 norm), infinity norm (L∞ norm), Shannon-Entropy norm, Jensen-Shannon Divergence norm, Renyi-Entropy norm, or Tsallis-Entropy norm.

[0064] In a particularly preferred embodiment, the first cost function Second cost function The following are given respectively:

[0065] ,

[0066] .

[0067] in,

[0068] It is measurement data.

[0069] It is a field inhomogeneity.

[0070] It is microscopic sample information.

[0071] and It is a non-zero scalar.

[0072] yes Two-dimensional vectors in a plane

[0073] yes Two-dimensional vectors in a plane

[0074] is a positive integer.

[0075] In a preferred embodiment, p = 2, that is, the norm used is the Euclidean norm. Therefore, the minimization algorithm minimizes the quadratic distance between the combination of field inhomogeneity and microscopic sample information—e.g., the product of field inhomogeneity and microscopic sample information—and the measurement data.

[0076] This variant and / or underlying algorithm of the method for determining field inhomogeneities in the field of view of a microscope according to the present invention may be referred to as the Alternating Least Squares Shading Correction Algorithm or the ALS-SC Algorithm.

[0077] For this embodiment, field inhomogeneity The (n+1)th update estimate can be calculated as follows:

[0078] ,

[0079] Microscopic sample information can be calculated as follows: The (n+1)th update estimate:

[0080] .

[0081] in

[0082] It is measurement information.

[0083] It is a field inhomogeneity The (n+1)th update estimate,

[0084] It is a field inhomogeneity The nth update estimate,

[0085] Microscopic sample information The nth update estimate,

[0086] Microscopic sample information The (n+1)th update estimate,

[0087] and It is a non-zero scalar.

[0088] yes Two-dimensional vectors in a plane

[0089] yes Two-dimensional vectors in a plane

[0090] is a positive integer. Attached Figure Description

[0091] Further properties and advantages of the invention are described below with reference to the accompanying drawings. In the drawings,

[0092] Figure 1 A schematic diagram of a microscope according to the present invention is shown;

[0093] Figure 2 A schematic representation of the field inhomogeneity of the detection beam path in a microscope is shown;

[0094] Figure 3 A schematic representation of a stitched image of measurement data consisting of multiple image patches is shown, and in each case, field inhomogeneity and, in some cases, portions generated by the sample are shown.

[0095] Figure 4 A schematic representation of six overlapping image patches is shown;

[0096] Figure 5 A schematic representation using coordinates is shown;

[0097] Figure 6 A schematic representation of two overlapping image patches is shown;

[0098] Figure 7 A schematic representation of three overlapping image patches is shown; and

[0099] Figure 8 A further schematic representation of the three overlapping image patches is shown. Detailed Implementation

[0100] Reference Figures 1 to 7 An embodiment of the microscope 100 according to the present invention is described. Identical components and equivalent components are generally indicated by the same reference numerals.

[0101] First, the microscope 100 includes a light source 10, such as a laser, for supplying excitation light 12, and an illumination beam path for guiding the excitation light 12 to the sample space 1. In the illustrated example, the illumination beam path includes a tube lens 20, a main beam splitter 23, and a microscope objective 40. The tube lens 20 produces an intermediate image plane 18, which is optically conjugate to a plane 11 in the sample 2 within the sample space 1. In the illumination beam path, the excitation light 12 travels through the intermediate image plane 18 and reaches the main beam splitter 23 via the tube lens 20, where it is reflected in the direction of the microscope objective 40. The excitation light 12 then passes through the back focal plane 42 of the microscope objective 40 and is subsequently guided to the sample space 1. The sample 2 may be a biological sample and may be prepared using a fluorophore that can be excited by the excitation light 12. The wavelength and intensity of the excitation light 12 can be appropriately selected relative to the sample 2 and the fluorophore used. The light source may include multiple different lasers. The wavelength and / or intensity may be adjustable.

[0102] Furthermore, the microscope 100 includes a detector 50 for detecting emitted light 16 emitted by sample 2 in sample space 1 and a detection beam path including a microscope objective 40 for guiding the emitted light 16 to the detector 50. In the example shown, the microscope 100 is a wide-field microscope, and the detector 50 is a camera, i.e., the field of view 30 for detecting the beam path (see [reference]). Figure 2 The image is formed on the sensor plane 51 of the camera 50. The sensor plane 51 is optically conjugate with the plane 11 in the sample space.

[0103] The emitted light 16 from sample 2 can typically be redshifted fluorescence emitted by fluorophores in sample 2. The main beam splitter 23 is configured to transmit the redshifted emitted light 16 and to reflect the excitation light 12, thus preventing most of the excitation light 12 scattered back from sample space 1 from traveling in the direction of camera 50.

[0104] In the detection beam path, the emitted light 16 emitted by sample 2 is collected by microscope objective 40, passes through main beam splitter 23, and is then imaged onto sensor plane 51 of camera 50 by lens 22.

[0105] As is generally known in the art, an excitation filter may be present in the excitation beam path, for example, between the lens 20 and the main beam splitter 23, and / or an emission filter may be present in the detection beam path, for example, between the lens 22 and the main beam splitter 23. Multiple different main beam splitters 23, suitable for various fluorophores, may also be present and can be switched in the beam path.

[0106] Furthermore, the microscope 100 includes a tool for setting the relative lateral position between the sample 2 and the microscope objective 40 relative to the optical axis 41 of the microscope objective 40. The mechanical actuator 44 and the control unit 90, such as a PC, configured to control the mechanical actuator 44 and to collect and evaluate measurement data from the detector 50. In the example shown, the optical axis 41 of the microscope objective 40 is at... Extending in the direction of the axis. The mechanical drive 44 can be, for example, part of a motorized sample stage, and in the example shown, is used to set the sample 2 to a specified position—that is, a specified position of the sample relative to the optical axis 41a. The purpose of coordinates. A right-handed and orthogonal coordinate system is shown below the mechanical actuator 44. In the example shown, microscope 100 further includes an axial driver 46 for setting a specified axial distance—that is, the distance between the sample and the microscope objective 40. The purpose is the distance in the direction.

[0107] According to the present invention, the control unit 90 is configured to perform a setting step, wherein the mechanical drive is sequentially set to... At least three distinct relative lateral positions in the plane ,in Representing a two-dimensional vector Each specific relative lateral position Corresponding to different sections or patches of sample 2. The measurement data collected by camera 50 is represented as follows: ,in It is a two-dimensional vector representing the detector coordinates in a suitable reference frame. For example, the origin is located at the center of the camera chip. Figure 5 The diagram shows block 67. Exemplary vectors in a plane The patch 67 corresponds to the area within the image 60 defined by the patch 67. and corresponding coordinates The field of view 30 is defined for a specific relative lateral positioning position.

[0108] According to the invention, the control unit 90 is further configured to perform a collection step, wherein at different relative lateral positions Measurement data are collected from detector 50 at least for a subset of points in sample 2 within the field of view 30 of the detection beam path at each relative lateral position. For each point in the subset, for at least two distinct lateral positions of the mechanical actuator 44. Collect measurement data. The latter feature can be, for example, at the set relative lateral position. Located in a configuration that causes overlap between adjacent tiles in the first coordinate direction The above is implemented on at least 50% of the two-dimensional regular meshes.

[0109] This is Figure 6 and Figure 7 It is depicted schematically in the middle. Figure 6 The first block 71 is shown and in the first coordinate direction A second patch 72 of the same size overlaps with the first patch 71. The amount of overlap 'a' exceeds that in the first coordinate direction. The width of blocks 71 and 72 in the above figure is 50%. Figure 7 Patches 73, 74, and 75 are shown as having the same size in each case. There, exactly half of patch 73 overlaps with patch 74, and the other half of patch 74 overlaps with patch 75; that is, in each case, the overlap 'a' is in the first coordinate direction. The upper part is exactly 50%. The relative horizontal position is set. A two-dimensional grid can be configured such that the overlap b between adjacent tiles is particularly in the first coordinate direction. Vertical second coordinate direction The percentage is at least 5%, and preferably at least 10%. This is in... Figure 4 The image is depicted schematically. Figure 4 Six identical tiles 61, ..., 66 are shown. In each case, in the first coordinate direction... The overlap with the corresponding neighbor is 50% (arrow a), and in the second coordinate direction The overlap with the corresponding neighbor is 10% (arrow b). Points in a subset of points can be evenly distributed across the field of view, for example, on a rectangular grid.

[0110] Figure 2 The field of view 30 as seen by camera 50 is schematically depicted without any contribution from the sample. This can be considered an image of a perfectly uniform sample. The image illustrates field inhomogeneity. In the example shown, the field inhomogeneity is essentially composed of vignetting 19, i.e., reduced intensity or shadows, in the corner regions. Furthermore, dust spots 17 are also shown, which may be, for example, in the plane 51 of the camera 50. While the vignetting has a smooth progression, the dust spots 17 have relatively sharp contours. By definition, field inhomogeneity... This includes the detected radiation profile, without any contribution from sample 2. The vignetting 19 at the corner is caused by the optical properties of the light source 10, the illumination beam path, the detection beam path, and the camera. In the example shown, dust spot 17 is located within camera 50. Therefore, field inhomogeneity... The structure does not depend on the relative lateral position of a specific setting. And so it is shown in the tile of each record.

[0111] This will be relative to Figure 3 Further explanation. Figure 3 A stitched image 80, consisting of 80 individual tiles, is schematically shown, depicting the measurement data captured from detector 50 in each case. The relative horizontal position is set. Located on a two-dimensional regular grid, specifically a rectangular grid. It can be seen that each tile and each element within a tile exhibits the same field inhomogeneity. Additionally, at least some of the tiles in the map show the contribution of the samples. It is important to note that... Figure 3 For illustration only, to show field inhomogeneities The purpose of the facts existing in each tile of the map. Figure 3 No overlapping tiles are shown. Thus, image 80 can be considered as an image showing only half of the tiles, for example, the total measurements.

[0112] In fact, an image can consist of hundreds of image patches that are fused together to form a sample from a whole slice of image. It is worth noting that... Figure 3 The image exhibits shadow artifacts across each tile, resulting in a checkerboard-like artifact across the merged entire slide image. This invention describes a novel nonparametric field inhomogeneity removal technique that produces… Figure 2 The estimation of the shadow distribution depicted in the image and the corrected photomicrograph of the sample—that is, both microscopic sample information.

[0113] In this respect, according to the invention, the control unit 90 is further configured to perform an evaluation step, wherein the evaluation is based on the measurement data collected in the collection step. Perform the following steps:

[0114] • Field inhomogeneity Independent of the separately set relative lateral positions The assumptions are based on the measurement data. Extracting field inhomogeneities in the field of view ,as well as

[0115] • From measurement data Extract the measurement data caused by sample 2. Partial microscopic sample information .

[0116] The control unit 90 of the microscope 100 can be specifically configured to perform the method according to the invention. The method according to the invention may include using the microscope 100.

[0117] Removing field inhomogeneity artifacts from microscopic data can be considered an inverse problem. Ignoring multiple scattering of light within thick samples, measurements such as camera signals can be processed as follows: Modeling as illumination distribution, i.e., field inhomogeneity With sample profile, i.e., microscopic sample information The product of:

[0118] .

[0119] here, Next is a two-dimensional vector representing the detector coordinates in a suitable reference frame. Typically, the origin is at the center of the camera chip, and It is a two-dimensional translation vector describing the lateral displacement experienced by sample 2 by translation via mechanical actuator 44. The goal is to obtain, for example, a variable translation vector. Estimating field inhomogeneity from a sequence of camera image sequences and measurement data sequences. and microscopic sample information .

[0120] Control unit 90 can be configured to perform the evaluation step as a field inhomogeneity-based process. Initial estimates and micro-sample information The initial estimate is performed using an iterative solution scheme for the double-blind estimation problem. More specifically, the control unit 90 can be further configured to perform the evaluation step as a measurement data... With field inhomogeneity and microscopic sample information Minimize the mathematical distance between combinations of , where the mathematical distance is based on an arbitrary mathematical norm.

[0121] More specifically, the control unit 90 can be configured to perform at least some of the following steps in the evaluation step:

[0122] •Based on measurement data And use the initial estimate of field inhomogeneity To calculate microscopic sample information The estimate, and

[0123] • Iterative calculation

[0124] ■Based on measurement data And in each case, the most recent estimate using micro-sample information is used. To iteratively calculate the update estimate of field inhomogeneity ,as well as

[0125] ■Based on measurement data And in each case, the most recent estimate of field inhomogeneity is used. To iteratively calculate the updated estimate of microscopic sample information ,

[0126] • Recent estimates assessing field inhomogeneity Recent estimate of micro-sample information Reproducing measurement data The level of accuracy, and

[0127] • Iteratively calculate the updated estimate of field inhomogeneities and the update estimation of micro sample information The steps continue until the most recent estimate of the field inhomogeneity is obtained. Recent estimate of micro-sample information Measurement data Reproduce to the specified accuracy level.

[0128] More specifically, the control unit 90 may be further configured to perform some or all of the following steps in the evaluation step:

[0129] •Based on measurement data And use the initial estimate of field inhomogeneity To calculate the initial estimate of microscopic sample information ;

[0130] •Based on measurement data And the initial estimate using micro-sample information To calculate the first update estimate of field inhomogeneity ;

[0131] •Based on measurement data And the first update estimate using field inhomogeneity To calculate the first update estimate of microscopic sample information ;

[0132] •Based on measurement data And the nth update estimate using micro-sample information To compute the (n+1)th update estimate of field inhomogeneity ;as well as

[0133] •Based on measurement data And the (n+1)th update estimate is used to utilize field inhomogeneity. To compute the (n+1)th update estimate of microscopic sample information .

[0134] Initial estimation of field inhomogeneity It can be, for example, a flat distribution.

[0135] More specifically, the control unit 90 can be further configured to be based on a scalar cost function. , Minimize to find field inhomogeneity and microscopic sample information The estimation. In this regard, the control unit 90 can be further configured to perform the following steps in the evaluation step: if used for field inhomogeneity New estimates and micro-sample information The new estimate of the first scalar cost function The value is less than the most recent update estimate used for field inhomogeneity. And the latest update estimate of micro sample information First scalar cost function The value of , and if the second scalar cost function is used for the new estimate of field inhomogeneity and the new estimate of micro-sample information. The value is less than the most recent update estimate used for field inhomogeneity. And the latest update estimate of micro sample information The second scalar cost function If the value is given, then the new estimate of the field inhomogeneity will be used as the new updated estimate of the field inhomogeneity. Furthermore, the new estimate of the micro-sample information is used as the new updated estimate of the micro-sample information. .

[0136] First scalar cost function in each case Second scalar cost function Includes measurement data With field inhomogeneity and microscopic sample information Combinations, especially the mathematical distance between products. First scalar cost function. Includes field inhomogeneity The mathematical norm and the second scalar cost function Includes microscopic sample information The mathematical norm of .

[0137] Mathematical measures can typically be based on any mathematical norm. Preferably, the Lp norm is used. In this case, the following first cost function can be used. Second cost function :

[0138] ,

[0139] .

[0140] Among them, for all points in the corresponding tile Summation over a specific relative lateral position exist Summing the above, and among them, It is a positive integer that defines the norm, and and It is a non-zero scalar.

[0141] Preferably, That is, the norm used is the Euclidean norm. Then, it is necessary to solve the problem relative to... and Minimize cost function and This is an alternating least squares problem. The (n+1)th update estimate of the field inhomogeneity can then be calculated as follows:

[0142] ,

[0143] The (n+1)th update estimate of the micro-sample information can be calculated as follows:

[0144] .

[0145] in, and It is a non-zero scalar to prevent division by zero.

[0146] The following sections will discuss the minimum number of required tiles, overlap considerations, non-measuring points, and further algorithmic extensions.

[0147] Overlap parameter It can be defined as ,in It is the size of the scan step and It is the horizontal size of the tile.

[0148] refer to Figure 6 Considering overlapping parameters in the interval 0 to 1 of Scanning. In this configuration, the amount of known data is given by multiplying the number of scan locations by the number of pixels measured. It is assumed that a single tile consists of a total of... It consists of pixels. Then in Existing during scanning There are known data. However, the number of unknown data... Always greater than The number of known data to be scanned, because it is necessary to estimate the shadow ( (Unknown data) and samples (left panel) 1 pixel plus the right boundary 1 pixel, of which (For overlap parameters). Then... Strictly applicable The scanning status.

[0149] Using the same logic as above, Scan by Composed of known data, the opposite of the following:

[0150] ,

[0151] The number of known data points must be at least as large as the number of unknown data points, i.e. This generates general conditions.

[0152] .

[0153] in, Indicates overlap and This refers to the number of scan positions. The above inequality must be satisfied. This allows us to have more data than unknown data in the current inverse problem.

[0154] Consider specific circumstances For an overlap of 50%, the above condition satisfies the equation. This situation occurs in... Figure 7 It is depicted in the text.

[0155] In conclusion, we can conclude that:

[0156] 1) Has A scan with only one scan point and no further prior knowledge is theoretically impossible;

[0157] 2) For those with A typical scan of a map patch needs to meet certain conditions. ;

[0158] 3) The minimum number of scan points is ;

[0159] 4) The minimum required overlap is 50%, that is... .

[0160] The argument given above can be easily extended to two dimensions. The constraint of recounting the number of scan points required... .exist In a special case, the example would be an equilateral triangle scan grid.

[0161] This is symbolic, as Figure 8 The image depicts three identical and partially overlapping tiles 76, 77, and 78. Tiles 76, 77, and 78 are shifted relative to each other by a shift vector c, thus forming an equilateral triangle. Figure 8 The situation can show the scanning steps. and tile size The ratio between them must satisfy:

[0162] .

[0163] In reality, the observed data is contaminated with noise. This can be mitigated by using data with high overlap. Record more The scanning position is used to improve the signal-to-noise ratio. In addition, the sample itself may not be densely filled with fluorophores. This provides the possibility that there may be non-measured points in the illumination distribution. In other words, the density of the structure on the sample can be regarded as the sampling density of the shadow, and vice versa. In the absence of structure in the illumination, the measurement of each sample looks the same except for the introduction of displacement (the sample brightness does not change when laterally translated). Conversely, in the absence of structure in the sample, no information can be obtained about the illumination when laterally translated. A uniform sheet of full fluorescence will look the same when laterally translated. Similarly, in a bright-field transmission system, a completely flat glass slide will remain transparent when laterally translated. Although not all of them are lost in the previous examples, and an empty and completely flat glass block (e.g., an empty cover glass without dust and scratches) will allow so-called flat-field measurements (as used in [0] for shadow correction), this technique cannot be applied to fluorescence microscopy by designing to block the light incident on the sample after it is reflected into the detection path. Another challenge in fluorescence microscopy is that a single fluorescent bead is insufficient to sample a large field-of-view illumination distribution when only a few images are captured. In this case, two measures can be taken to offset non-measured points:

[0164] 1) Increase the number and overlap of scan points to ensure the irradiation distribution is fully sampled by the sample, or

[0165] 2) Regress or interpolate from the measured point to non-measured points.

[0166] The method for determining field inhomogeneity in the field of view of a microscope is essentially a microscopic method, since a sample is necessary, and microscopic sample information is incidentally determined for said sample. However, for the microscopic method according to the invention, a determined field inhomogeneity can be used. That is, the field inhomogeneity in the field of view of the microscope is determined using the method for determining field inhomogeneity according to the invention and in the determination step using a first sample. Then, in the microscopy step, the method for determining field inhomogeneity according to the invention is performed on a second sample different from the first sample, and wherein the field inhomogeneity determined in the determination step is... Used as field inhomogeneity The initial estimate.

[0167] This invention provides a method for detecting and removing field inhomogeneities from various types of captured microscopic data—e.g., wide-field fluorescence, transmission bright-field, and laser scanning confocal microscopy. In addition to correcting inhomogeneities in images, the detection of field inhomogeneities can also be used to monitor their changes over time. Based on such changes, instrument alignment maintenance can be triggered. The method involves laterally translating the sample and recording spatial overlap data. The spatial overlap between the captured patches enables computational detection and removal of field inhomogeneities.

[0168] The method according to the invention can be applied to a variety of imaging systems to remove vignetting and field inhomogeneities. Such systems include wide-field systems (e.g., systems from Carl Zeiss such as Axio Scan, Observer, DaVinci, Elyra, Apotome, AxioZoom, and AxioVert), scanning systems (e.g., systems from Carl Zeiss such as LSM and Light Sheet), and light-field systems.

[0169] Unlike parametric shading correction methods, the method according to the invention includes, for example, estimating dust spots on the detector that are removed alongside the vignetting effect of the illumination distribution. The resulting image is free from vignetting artifacts and obstructions, such as those caused by dust spots on the detector.

[0170] Note that the method according to the invention, and especially the ALS-SC algorithm described herein, works well for many different microscope modalities—e.g., fluorescence wide-field data or transmission bright-field data.

[0171] In the testing of the method for transmitting bright-field data according to the invention, the illumination is spatially partially coherent. Specifically, the numerical aperture of the illumination, defined by the condenser, is as large as the numerical aperture of the objective lens. Therefore, the method according to the invention, and specifically the ALS-SC algorithm, can be applied to spatially partially coherent illumination settings.

[0172] Further testing was conducted using the method according to the invention with measurement data from a laser scanning microscope (LSM). In this regard, it is important to emphasize that the method according to the invention cannot be confused with effects such as those caused by sample tilting that would result in a reduction in signal through the optical section. This is because, during lateral translation of the out-of-focus portion of the sample, the dark area must move with the sample, as the sample remains out of focus during pure lateral translation. However, when the dark area on the sample remains constant after sample translation, the method according to the invention, and particularly the ALS-SC algorithm, uniquely distinguishes shadows from sections in the LSM system.

[0173] List of reference numerals

[0174] 1 Sample Space

[0175] 2 samples

[0176] 10. Light sources, such as lasers

[0177] 11 Sample Plane

[0178] 12 Excitation Light

[0179] 16. Emit light

[0180] 17 dust spots, for example, on detector 50

[0181] 18. The intermediate image plane, which is optically conjugate to the sample plane 11.

[0182] 19. Fainting

[0183] 20. Lens tube in the excitation beam path

[0184] 22. Detecting the lens barrel in the beam path

[0185] 23 Main beam splitter

[0186] 30. Field of view for detecting beam path

[0187] 40 Microscope Objectives

[0188] 41. Optical axis of microscope objective lens: 40°

[0189] 42 Back focal plane, pupil plane

[0190] 44 Mechanical actuators for lateral manipulation

[0191] 46 Mechanical drives for axial manipulation

[0192] 50 detectors, such as cameras

[0193] 51. Detection plane, optically conjugate with sample plane 11.

[0194] 60. A stitched image consisting of 6 image tiles.

[0195] Image tiles 61-66

[0196] 67 Image tiles

[0197] Image tiles 71-78

[0198] 80 is a stitched image consisting of 80 image tiles.

[0199] 81 triangular mesh elements

[0200] 90 Control units, such as PCs

[0201] 100 Microscope according to the present invention

[0202] a Overlap in direction

[0203] b Overlapping in direction

[0204] scan step size

[0205] Measurement data = Measurement information

[0206] Measurement data

[0207] Field inhomogeneity

[0208] Field inhomogeneity

[0209] Field inhomogeneity initial estimate

[0210] Field inhomogeneity First Update Estimate

[0211] Field inhomogeneity The nth update estimate

[0212] Field inhomogeneity The (n+1)th update estimate

[0213] First scalar cost function

[0214] Second scalar cost function

[0215] Two-dimensional vectors in a plane

[0216] exist Components in direction

[0217] exist Components in direction

[0218] positive integer

[0219] Microscopic sample information

[0220] Microscopic sample information

[0221] Microscopic sample information initial estimate

[0222] Microscopic sample information First Update Estimate

[0223] Microscopic sample information The nth update estimate

[0224] Microscopic sample information The (n+1)th update estimate

[0225] horizontal size of the tile

[0226] Two-dimensional vectors in a plane

[0227] Right-handed orthogonal coordinate system

[0228] exist Components in direction

[0229] exist Components in direction

[0230] Optical axis direction

[0231] nonzero scalar

[0232] nonzero scalar

[0233] References

[0234] [0]US8253789B2.

[0235] [1]Young, Ian T. "Shading correction: compensation for illumination and sensor inhomogeneities." Current Protocols in Cytometry 14, no. 1 (2000): 2-11.

[0236] [2]Preibisch, Stephan, Stephan Saalfeld, and Pavel Tomancak. "Globally optimal stitching of tiled 3D microscopic image acquisitions." Bioinformatics 25.11 (2009): 1463-1465.

[0237] [3]Legesse, Fisseha Bekele, Olga Chernavskaia, Sandro Heuke, Thomas Bock-litz, Tobias Meyer, Jürgen Popp, and Rainer Heintzmann. "Seamless stitching of tile scan microscope images." Journal of microscopy 258, no. 3 (2015): 223-232.

[0238] [4]Smith, Kevin, Yunpeng Li, Filippo Piccinini, Gabor Csucs, CsabaBalazs, Alessandro Bevilacqua, and Peter Horvath. "CIDRE: an illumination-correction method for optical microscopy." Nature methods 12, no. 5 (2015): 404-406.

[0239] [5] Peng, Tingying, Kurt Thorn, Timm Schroeder, Lichao Wang, Fabian J.Theis, Carsten Marr, and Nassir Navab. "A BaSiC tool for background and shading correction of optical microscopy images." Nature communications 8, no. 1 (2017): 14836.

Claims

1. A microscope, comprising: Light source (10), which is used to supply excitation light (12). Illumination beam path, the illumination beam path being used to guide the excitation light (12) to the sample space (1). A detector (50) is used to detect emitted light (16) emitted by a sample (2) in the sample space (1). The detection beam path includes a microscope objective (40) for guiding the emitted light (16) to the detector (50). Mechanical actuator (44), the mechanical actuator (44) being used to set the relative lateral position between the sample (2) and the microscope objective (40) relative to the optical axis (41) of the microscope objective (40). ), A control unit (90) is configured to control the mechanical actuator (44) and to collect and evaluate measurement data from the detector (50). Its features are, The control unit (90) is configured to perform the following steps: The setup step involves sequentially setting the mechanical actuator (44) to at least three different relative lateral positions. ), Collection steps, wherein, at different relative lateral positions ( Each relative lateral position in ) In the process, measurement data is collected from the detector (50) for at least a subset of points in the sample (2) within the field of view (30) of the detection beam path. ), wherein, for each point in the subset, for at least two different lateral positions of the mechanical actuator (44) Collect measurement data, and The evaluation step, wherein, based on the measurement data collected in the collection step ( Perform the following steps: •Utilize field inhomogeneity ( Independent of the separately set relative lateral positions ( The assumption is based on the measurement data ( Extracting the field inhomogeneity in the field of view ( ),as well as • From the measurement data ( Extract the measurement data representing the result of the sample (2). Microscopic sample information () ).

2. The microscope according to claim 1, Its features are, The relative lateral position set ( It is located on a two-dimensional regular grid.

3. The microscope according to claim 2, Its features are, The grid is rectangular or triangular.

4. The microscope according to claim 2 or 3, Its features are, The control unit (90) is further configured to the set relative lateral position ( The two-dimensional grid is configured such that the overlap (a) between adjacent tiles (61, 62) is in the first coordinate direction ( The percentage is at least 50%.

5. The microscope according to any one of claims 2 to 4, Its features are, The control unit (90) is further configured to the set relative lateral position ( The two-dimensional grid is configured such that the overlap (b) between adjacent tiles (61, 64) is particularly perpendicular to the first coordinate direction. The second coordinate direction () The content is at least 5%, and preferably at least 10%.

6. The microscope according to any one of claims 1 to 5, Its features are, The points in the subset of points are uniformly distributed across the field of view.

7. The microscope according to claim 6, Its features are, Points in the subset of points lie on a regular grid.

8. The microscope according to claim 7, Its features are, The grid of the subset of points is a rectangle or a triangle.

9. The microscope according to any one of claims 1 to 8, Its features are, The collection step includes collecting measurement data for each point in the field of view. ).

10. The microscope according to any one of claims 1 to 9, Its features are, The evaluation steps include measurement data at multiple points, particularly multiple pixels. The merger of ).

11. The microscope according to any one of claims 1 to 10, Its features are, The evaluation step includes evaluating the measurement data of each point in the field of view. ).

12. The microscope according to any one of claims 1 to 11, Its features are, The control unit (90) is configured to take the evaluation step as a process based on the field inhomogeneity ( The initial estimate of ) and the microscopic sample information ( The iterative solution scheme for the double-blind estimation problem is executed by taking the initial estimate of the first estimate.

13. The microscope according to any one of claims 1 to 12, Its features are, The control unit (90) is configured to perform at least some of the following steps in the evaluation step: • Based on the measurement data ( And use the initial estimate of field inhomogeneity ( To calculate the microscopic sample information ( The estimate of ) and • Iterative calculation ■Based on the measurement data ( And in each case, the most recent estimate of the microscopic sample information is used. ), to iteratively compute the updated estimate of the field inhomogeneity ( ),as well as ■Based on the measurement data ( And in each case, the most recent estimate of the field inhomogeneity is used. ), to iteratively calculate the update estimate of the microscopic sample information ( ), • The most recent estimate assessing the field inhomogeneity ( ) and the most recent estimate of the microscopic sample information ( Reproduce the measurement data ( The accuracy level of ) and • Repeatedly iterate to calculate the updated estimate of the field inhomogeneity ( ) and the update estimate of the microscopic sample information ( The steps continue until the most recent estimate of the field inhomogeneity is obtained. ) and the most recent estimate of the microscopic sample information ( The measurement data ( Reproduce to the specified accuracy level.

14. The microscope according to any one of claims 1 to 13, Its features are, The control unit (90) is further configured to perform some of the following steps in the evaluation step: • Based on the measurement data ( And using the initial estimate of the field inhomogeneity ( To calculate the initial estimate of the microscopic sample information ( ); • Based on the measurement data ( And using the initial estimate (based on the microscopic sample information) ) to calculate the first update estimate of the field inhomogeneity ( ); • Based on the measurement data ( And using the first update estimate of the field inhomogeneity ( To calculate the first update estimate of the microscopic sample information () ); • Based on the measurement data ( And use the nth update estimate of the microscopic sample information ( To calculate the (n+1)th update estimate of the field inhomogeneity ( );as well as • Based on the measurement data ( And use the (n+1)th update estimate of the field inhomogeneity ( To calculate the (n+1)th update estimate of the microscopic sample information ( ).

15. The microscope according to claim 13 or 14, Its features are, The initial estimate of the field inhomogeneity ( () is a flat distribution.

16. The microscope according to any one of claims 1 to 15, Its features are, The control unit (90) is further configured to perform the evaluation step as the measurement data ( ) and the field inhomogeneity ( ) and the microscopic sample information ( Minimize the mathematical distance between combinations of ), where the mathematical distance is based on an arbitrary mathematical norm.

17. The microscope according to claim 16, Its features are, The control unit (90) is further configured to perform the following steps in the evaluation step: • If used for the field inhomogeneity ( The new estimate and the microscopic sample information () The new estimate of the first scalar cost function () The value of ) is less than the most recent update estimate used for the field inhomogeneity ( ) and the most recent update estimate of the microscopic sample information ( The first scalar cost function () The value of ) and • If the second scalar cost function is used for the new estimate of the field inhomogeneity and the new estimate of the microscopic sample information ( The value of ) is less than the most recent update estimate used for the field inhomogeneity ( ) and the most recent update estimate of the microscopic sample information ( The second scalar cost function () The value of ) • Wherein, in each case, the first scalar cost function ( ) and the second scalar cost function ( ) contains the measurement data ( ) and the field inhomogeneity ( ) and the microscopic sample information ( The mathematical distance between combinations of ) and • Wherein, the first scalar cost function ( ) includes the field inhomogeneity ( The norm of ) and the second scalar cost function ( ) contains the microscopic sample information ( The mathematical norm of ) Then the new estimate of the field inhomogeneity is used as the new updated estimate of the field inhomogeneity. ), and the new estimate of the microscopic sample information is used as the new updated estimate of the microscopic sample information ( ).

18. The microscope according to claim 16 or 17, Its features are, The field inhomogeneity ( ) and the microscopic sample information ( The combination of ) is the field inhomogeneity ( ) and the microscopic sample information ( The product of ).

19. The microscope according to any one of claims 16 to 18, Its features are, The mathematical norm is one of the following: Lp norm, L2 norm (Euclidean norm) Manhattan norm (L1 norm) Infinite norm (L∞ norm) Shannon entropy norm, Jensen-Shannon divergence norm Reni entropy norm, Tsallis entropy norm.

20. The microscope according to any one of claims 17 to 19, Its features are, The first cost function ( ) and the second cost function ( The following are given respectively: ， ， in, The measurement data is as described. This refers to the field inhomogeneity. This refers to the microscopic sample information. and It is a non-zero scalar. yes Two-dimensional vectors in a plane yes Two-dimensional vectors in a plane Corrective integer.

21. The microscope according to any one of claims 19 or 20, Its features are, Where p = 2.

22. The microscope according to any one of claims 20 or 21, Its features are, The (n+1)th update estimate of the field inhomogeneity is calculated as follows: ， And the (n+1)th update estimate of the microscopic sample information is calculated as follows: ， in, The measurement data is as described. It is the field inhomogeneity mentioned above The (n+1)th update estimate, It is the field inhomogeneity mentioned above The nth update estimate, The microscopic sample information The nth update estimate, The microscopic sample information The (n+1)th update estimate, and It is a non-zero scalar. yes Two-dimensional vectors in a plane yes Two-dimensional vectors in a plane Corrective integer.

23. The microscope according to any one of claims 1 to 22, Its features are, The detector includes: Two-dimensional spatially resolved photodetector (50), such as a camera, One-dimensional spatially resolved detectors, such as linear detector arrays, or A single light detector, such as a dot light detector.

24. The microscope according to claim 23, Its features are, The detector (50) includes at least one of the following: CCD components, CMOS components, SPAD components, PMT.

25. The microscope according to any one of claims 1 to 24, It is at least one of the following: Wide-field microscope, Scanning microscope, Confocal microscope, Light field microscope, Light sheet microscope, TIRF microscope, SIM microscope.

26. The microscope according to any one of claims 1 to 25, Its features are, The control unit (90) is configured to perform the method according to any one of claims 28 or 29.

27. A method for determining field inhomogeneities in the field of view of a microscope, comprising the following steps: The sample (2) is illuminated by excitation light (12) passing through the illumination beam path of the microscope (100). The emitted light (16) emitted by the sample (2) is guided through a detection beam path including the microscope objective (40) to the detector (50). The emitted light (16) is detected by the detector (50). A specific lateral position between the sample (2) and the microscope objective (40) is set by means of a mechanical actuator (44) relative to the optical axis (41) of the microscope objective (40). ), Its features are, The setup step involves sequentially setting the mechanical actuator (44) to at least three different relative lateral positions. ), The collection step, wherein, at the different relative lateral positions ( Each relative lateral position in ) In the process, measurement data is collected from the detector (50) for at least a subset of points in the sample (2) within the field of view (30) of the detection beam path. ), wherein, for each point in the subset, for at least two different lateral positions of the mechanical actuator (44) Collect measurement data, and The evaluation step, wherein, based on the measurement data collected in the collection step ( Perform the following steps: Using the aforementioned field inhomogeneity ( It does not depend on the separately set relative lateral position. The assumption is based on the measurement data ( Extracting field inhomogeneities in the field of view (30) ),as well as From the measurement data ( Extract the measurement data representing the result of the sample (2). Microscopic sample information () ).

28. A microscopic method, Includes the following steps: The determination step, wherein the field inhomogeneity in the field of view of the microscope is determined using a sample according to the method of claim 27. ), A microscopic step, the microscopic step comprising performing the steps of the method according to claim 27 on a sample different from the sample used in the determining step, and wherein the field inhomogeneity determined in the determining step ( ) is used as the field inhomogeneity ( The initial estimate of ).

29. The method according to any one of claims 27 or 28, Further includes the use of a microscope according to any one of claims 1 to 26.

Images