DE102014107934A1 - A method of microscopically imaging samples on soils of fluid-filled pots of a microtiter plate - Google Patents

Abstract

A method is described for the microscopic imaging of samples (14) which adhere to trays (3) of the pots (1) in a microtiter plate (11) which has a multiplicity of pots (1) filled with fluid (4), wherein the pots (1) successively from a bottom opposite the bottoms (3) from a transmitted light source (16) along an optical axis (OA) illuminated with illumination radiation (2) and the thus top of the illuminated bottoms (3) of the pots ( 1) from the underside (15) forth individually magnifying image, wherein for improving an edge illumination of the bottom (3) at each illuminated potty (1) between the potty (1) and the transmitted light source (16) a homogenizing element (7) is arranged Illuminated by the transmitted-light illumination source (16) for which illumination radiation (2) is transparent, and a part immersed in the potty (1) and there in the fluid (4) ( 7a) which has an end face (8) lying in the fluid (4).

Description

The invention relates to a method for the microscopic imaging of samples which adhere to the bottom of the pots in a microtiter plate having a plurality of pots filled with fluid, the pots being successively separated from a bottom opposite the bottom from a transmitted light source along a Illuminated optical axis and the thus illuminated from the top soils of the pots from the bottom forth individually magnifying.

In the life sciences microscopy of living cells plays an important role. These are often cultivated in microtiter plates. The cells are located at the bottom of the vessel and are surrounded by a nutrient medium. They are usually microscoped with an inverted microscope; In this case, the lens is the bottom of the sample vessel bottom. The illumination of the sample can be done by reflected light or transmitted light. For transmitted light images, a light is placed above the sample vessel. However, since biological cells contain only low-absorbing components, bright-field transmitted light images are typically only very slightly contrasted. With the help of various Durchlichtkontrastverfahren such. B. Phase contrast, DIC u. a. The small refractive index difference of the individual cell components relative to one another and to the surrounding medium can be converted into an intensity difference, which then provides a contrasted transmitted light image.

Thus, the invention is concerned with transmitted light microscopy of samples adhering to the bottom of a microtiter plate. The floors are illuminated in transmitted light and imaged in high resolution in a microscope. This type of microscopy differs from the conventional image of an entire microtiter plate used in other applications in the prior art, as described, for example, in US Pat DE 10200541 A1 he follows. If the floors are magnified individually or in small groups in transmitted light, the quality requirements of transmitted light illumination are many times higher. This applies in particular with regard to the aforementioned transmitted-light contrast method.

The field of the invention is also delineated against so-called fluorescence readers, which check whether or how much the liquid fluoresces in a well of a microtiter plate. Again, a collection of all pots of a microtiter plate is usually done simultaneously. In addition, the quality of the floor lighting is irrelevant in these applications. The US 6074614 is concerned with such a use of microtiter plates and provides a cover plate which, matching the microtiter plate, has a plurality of cylindrical protrusions which dip into the potentially fluorescent liquid of the microtiter plate. The goal is to ensure the most uniform path length of the radiation through the pots of the microtiter plate in the fluorescence excitation and readout. The question of lighting the bottom of a microtiter plate does not matter.

A prerequisite for qualitatively good transmitted light images, irrespective of which contrast method is used, is as homogeneous illumination as possible over the entire object field captured by the objective, d. H. of the whole soil. Usually, this is realized via a Köhler illumination, which guarantees that all points of the object field are illuminated not only approximately with the same intensity, but also with an approximately equal angular spectrum, which is described by the numerical aperture (NA) of the illumination. According to classical teachings, this NA should be slightly smaller than the NA of the objective or the detection. These conditions can be easily met for samples that are located between a slide and a coverslip, as well as for samples in petri dishes.

Small sample containers, such. However, microtiter plates are a problem in this respect. In the following, microtiter plates are used as an example for small sample containers. These have z. B. 96 individual pots on. The pots are typically about 11 mm high and have a diameter of about 7 mm. Thus, the upper edge of the potty truncates the cone of light of the transmitted light illumination, which reaches a certain point of the bottom of the potty and is caught by the lens after passing through it. Depending on whether this bottom point is in the potty center or rather on the edge, the usable light cone is another, d. H. every point of the potty bottom "sees" another illumination NA area. If the lens NA is large enough to accommodate all the different illumination cones and the illumination in each of these cones provides approximately equal intensity, the illumination of the potty bottom remains homogeneous although the potty edge prunes the illumination light.

One approach to compensate for the effect of the fluid meniscus is in US 6238911 described. There, a lens array is introduced above the microtiter plate, which ensures that the image of the illumination pupil does not shift laterally to the image of the detection pupil. Under this condition, a classical phase contrast can also be realized in microtiter plates. This lens array can also be introduced so that the lens curvature points in the direction of the sample or the lens itself lies partially within the potty. However, this design enhances edge illumination only in very specific circumstances. Deviation from these conditions even results in poorer illumination than without such a lens array.

In dark field scattered light examinations, the meniscus is disturbing. This problem is in US 2007171417 solved in that the liquid containing the sample, another liquid such. As an oil is given, which reduces the meniscus in its strength. In addition, a plate may also be placed which preferably absorbs the illumination light to reduce reflections. Such plates are not useful for the area presented here because they would interfere with the transmitted light illumination, which is why no absorbing layers can be brought into the beam path.

In WO 02087763 discloses a lid for microtiter plates, which may be flat or formed with bulges and serves to create the same conditions in different pots of a microtiter plate, so that measurements such. As measuring a fluorescence spectrum of a fluorescent solution, are comparable to each other in different pots. This application has no relation to microscopy, illumination and optimization do not matter.

In US 2010197004 In the case of a liquid with cells distributed homogeneously therein, a lid for microtiter plates with bulges serves to ensure that after sedimentation of these cells all parts of the bottom of the cell have the same cell density. The lid prevents more cells from accumulating near the edge than in the center. So this lid is used for sample preparation, not observation.

None of the prior art publications describing lids for cylindrical well microtiter plates addresses the issue of transmitted light illumination, that is, a method for enlarged imaging of samples on the bottom of a transmitted light well of a microtiter plate. Since the cited documents and especially the US 6074614 As for other microscopy methods, the problems to be solved there are different from those of the present invention.

The invention is therefore based on the object, a method for the microscopic imaging of samples which adhere to a bottom of the wells in a microtiter plate having a plurality of fluid-filled pots, the pots successively from a top surface opposite the bottoms of a Illuminated transmitted-light source along an optical axis and the so-lit from the top bottoms of the pots from the bottom to zoom individually, to indicate where the lighting of the floor is even without the use of imaging optics with high numerical aperture uniform.

This object is achieved according to the invention with a method for the microscopic imaging of samples which adhere to the bottom of the pots in a microtiter plate having a multiplicity of pots filled with fluid, the pots being passed successively from a top side opposite the bottoms from a transmitted-light illumination source illuminated on an optical axis and the thus illuminated from the top bottoms of the pots are individually magnifying from the bottom, in which is arranged to improve a peripheral illumination of the bottom of each illuminated potty between the potty and the transmitted light source, a homogenizing element, from the transmitted light source is illuminated and has a in the potty and there immersed in the fluid part, which has an end face located in the fluid.

According to the invention, the edge illumination at the bottom is improved by engaging within the potty which destroys the meniscus in the optically relevant region by placing the homogenizing element between the transmitted light source and the potty having a part immersed in the fluid, one in the fluid lying end face.

As appropriate prove to z. B. transparent cylinders that dip into the pots. These can be z. B. made of glass or plastic. If they have a plane surface perpendicular to the optical axis at both the top and bottom ends, the meniscus effect is suppressed and a parallel illumination beam remains parallel after passing the potty.

In order to illuminate as many pots as possible in this way, the invention may be embodied as a simple lid for a microtiter plate, wherein at the position of each potty a cylinder is formed, which protrudes into the potty.

This lid may rest or be spaced on the well edges of the microtiter plate. These can z. B. be attached to the corners of the lid supporting elements that hold the lid relative to the upper surface of the microtiter plate or relative to their footprint. Microtiter plates are available with different potty numbers, z. B. 24, 96 or 384 pots. For plates with 96 pots, there are also two variants, the standard variant with a potty diameter of about 7 mm, and one with half volume with a potty diameter of about 5 mm. For all these variants, a lid can be constructed as described above with protruding into the liquid cylinders. It does not have to be a circular cylinder. In particular, the pots in microtiter plates with 384 wells are usually not round, but executed square with rounded corners. A corresponding shape may also have the immersing elements of the lid. Also, the lid does not have to cover the entire plate. Also, covers that cover only parts of the microtiter plate, for example, the central area or only every n-th potty are possible in the context of this invention. Instead of cylinders, other geometric shapes may be formed, for. As cones or truncated cones and the like. For the sake of simplicity, however, the following will be discussed in part only by cylinders, without this being limiting.

Since the fluids in microtiter plates are often nutrient solutions in which the sample material is located, it is advantageous if the homogenization element can be cleaned well. The protruding cylinders can create many hard to reach areas. To facilitate cleaning, corners and edges can be rounded. Usually, a cleaning or sterilization of sample vessels by autoclaving happens. Therefore, it is advantageous if the lid according to the invention consists of materials that are autoclavable, ie z. As glass or autoclavable plastic. Microtiter plates are i. d. R. designed as disposable vessels. Similarly, the execution of the cover described as a disposable or consumable material is conceivable.

If the cylinder is inserted into the potty of a microtiter plate, it may happen that an air bubble remains below the cylinder. This disturbs the homogeneity of the illumination sensitive. If only a single cylinder is used, this problem can be solved by repeated insertion. If, however, a lid with many cylinders is used, a repetition does not lead to the goal, since in each case under some cylinders air bubbles will find. To eliminate this problem, several options are available.

On the one hand, the underside of the cylinders can be slightly curved or chamfered so that air bubbles can be directed to the side and then escape between the cylinder and the wall of the potty. However, care should be taken to ensure that the curvature or slope is not too strong and again a meniscus or a breaking edge at the interface cylinder-liquid is generated, which has a negative effect on the illumination, unless the cylinder material is similar to the liquid Has refractive index.

On the other hand, the cylinder may have a smaller diameter than the potty, so that there is sufficient space between the cylinder and the potty edge, so that air bubbles can escape. However, the cylinder diameter should not be too small, because otherwise the meniscus is only destroyed in the potty center, while it forms again around the cylinder. With a potty diameter of 7 mm, cylinders with a diameter of 5 mm are suitable. When inserting almost never create air bubbles and the illumination is very good up to the potty edge. For cylinders with a diameter of 6 mm, almost always air bubbles are created which can not escape, and for cylinders with a diameter of 4 mm or smaller, the illumination deteriorates substantially due to the meniscus between the cylinder and the potty wall.

Since these cylinders, as already described, should have a smaller diameter than the pots themselves, a small meniscus forms between the cylinder and the potty edge. By clever choice of the cylinder diameter, its effect on the illumination can indeed be minimized, but it still manifests itself in a ring structure in the picture. However, the intensity modulation of these rings is significantly less than the intensity drop in edge without using the lid. If the cylinder is centered on the potty, these rings are concentric with the potty. If the cylinder is slightly eccentric, the rings are also off-center. If the cylinders are equal to the pots after each insertion, the ring structure is always the same relative to the pots. To prevent the cylinder move relative to the pots, the lid z. B. made to fit the outer dimensions of the microtiter plate. Alternatively, locking or positioning options can be provided laterally on the edge of the lid, by means of which the lid is brought into a defined position to the microtiter plate. This then makes it possible to calculate these rings out by means of suitable software. The ring structure is usually quite blurred because it comes from planes that are relatively far away from the observation plane. This makes it possible in many cases to calculate the ring structure out of the image by means of a Fourier analysis. For this purpose, a high-pass filter is applied to the image, which filters out low frame rates. In particular, if the sample structures are very small-sized, ie have high-frequency components, this approach is advantageous.

As already mentioned, it is preferable that the part immersed in the fluid has the shape of a general straight cylinder.

Air bubbles that adhere to the end face of the fluid submerged part significantly disturb the lighting. It is therefore preferred that the end face to the optical axis has an angle of inclination or tangent angle which is between 80 and 90 degrees and to prevent air bubbles from degrading the edge illumination of the bottom again. For a curved face, it is preferable to design it to have a tangent angle that is between 80 ° and 90 °, i. H. that no smaller tangent angle occurs in the optically relevant area. A chamfer or rounding at the edge does not exclude this.

The edge illumination is particularly good when the shape of the end face in plan view along the optical axis is similar to the shape of the cross section of the potty.

A gap of at least 1 mm and a maximum of 3 mm between the edge of each end face and the inner wall of the associated potty allows air bubbles to escape. At the same time the edge illumination is not disturbed.

• (a) das Töpfchen wird mit Beleuchtungsstrahlung beleuchtet und
• (b) der beleuchtete Boden des Töpfchens wird von der Unterseite der Mikrotiterplatte vergrößernd abgebildet und ein Bild des beleuchteten Bodens aufgenommen, wobei eine töpfchenverursachte Ungleichmäßigkeit der Beleuchtung des Bodens ausgeglichen wird, indem
• (c) ein probenloses Test-Töpfchen bereitgestellt wird, das bis auf die fehlende Probe dem zu mikroskopierenden Töpfchen entspricht,
• (d) an dem Test-Töpfchen wird eine Referenzmessung mittels der Schritte (a) und (b) durchgeführt, wobei ein Referenzbild aufgenommen wird, das den gesamten beleuchteten Boden zeigt,
• (e) anhand des Referenzbildes wird eine Helligkeitskorrekturangabe ermittelt, wobei die Helligkeitskorrekturangabe eine Helligkeitsschwankung als Funktion des Ortes auf dem Boden des Test-Töpfchens angibt,
• (f) das Bild des Bodens des probenenthaltenden Töpfchens wird mittels der Helligkeitskorrekturangabe korrigiert, wobei die Lage des Bildes am Boden ermittelt und der zu dieser Lage gehörende Wert der Helligkeitskorrekturangabe verwendet wird.

A further development provides for the following steps in the method for microscopically imaging a sample which adheres in a potty to the bottom of the potty:

• (a) the potty is illuminated with illuminating radiation and
• (b) the illuminated bottom of the potty is magnified from the underside of the microtiter plate and an image of the illuminated bottom is taken, compensating for potty-induced unevenness in the illumination of the soil by:
• (c) providing a sampleless test potty which, except for the missing sample, corresponds to the potty to be microscoped,
• (d) a reference measurement is made on the test potty by means of steps (a) and (b) taking a reference image showing the entire illuminated floor,
• (e) a brightness correction indication is determined from the reference image, the brightness correction indication indicating a brightness variation as a function of the location on the bottom of the test potty,
• (f) the image of the bottom of the sample-containing potty is corrected by means of the brightness correction specification, whereby the position of the image on the ground is determined and the value of the brightness correction indication belonging to this position is used.

The further development is based on the following finding: A sample vessel-related shading, which is firmly connected to the reference system of the sample vessel, is not recognized by the described methods and consequently can not be corrected.

A sample container-related shading is present when the structure of the sample vessel itself leads to shading. This occurs z. B. in small vessels (microtiter plates), as described below. If the sample is moved, this shading wanders with it. It is thus always in the same place of a sample container-related reference system, but can not be assigned to a fixed position of a beam path-related reference system. Brightness gradients occur especially in transmitted light illumination of microtiter plates. Microtiter plates are sample vessels which are used in particular in live cell observation. These are plates that are filled with a defined number of pots, z. B. 24, 96 or 384, are equipped at regular intervals. In each of these pots, a sample can be introduced, for. As cells or embryos. For microscopic observation, the pots with a transparent bottom of z. As polystyrene or glass provided. The optical properties of the pots disturb the illumination in transmitted light considerably. In the following, this effect is explained by means of a microtiter plate whose 96 pots have a height of 11 mm and a diameter of 7 mm.

The top edge of the potty trims the light cone of the transmitted light illumination, which reaches a certain point of the bottom of the potty and which is caught by the lens after passing the bottom. Depending on whether this bottom point is in the potty center or rather on the edge, the usable light cone is another, d. H. each point of the bottom of the potty is illuminated with a different numerical aperture. This can already lead to a sample vessel-dependent shading, the effect is the greater, the more the illumination light cone is trimmed by the potty and the unevenly the illumination intensity is distributed to the different illumination angles. On the other hand, if the numerical aperture of the lens is large enough to accommodate all the different illumination cones and the illumination in each of these cones provides approximately equal intensity, the illumination of the potty bottom remains homogeneous although the pottent edge trims the illumination light.

However, the aqueous medium in which the sample is present forms a meniscus on its surface. The interface between air and medium is therefore curved. The radius of the meniscus depends on the type of fluid, wall material and Coating of the pots, as well as the filling process, so whether dry or already wet pots were filled, whether they were stirred, etc. In most cases, the liquid runs along the potty wall slightly upwards, while the liquid level is lower in the potty center.

This results in a parallel incident beam diverging after passing the meniscus. This divergence is the more pronounced the smaller the radius of the meniscus is formed. For lenses with a high NA this is not a problem because the diverging beams can also be picked up. Numerous applications, however, require weakly magnifying lenses, which image a large field and thus enable an overview of the sample. For example, with only one image of a 2.5x magnifying lens, almost a complete potty of a 96-well microtiter plate can be imaged.

Typically, however, low-magnification lenses have only a low NA, z. B. 0.08 or 0.12. The NA (NA = n · sin (α)) describes the maximum angle α that a beam can make with the optical axis in order to be brought from the lens to the image. Here, n is the refractive index of the medium between the objective and the sample. In the case of weakly magnifying lenses, this is usually air, ie n = 1. Thus, all illumination beams that are refracted by the meniscus to larger angles than the limit angle defined by the objective NA do not enter the objective. Even if illumination light reaches every part of the potty bottom, the meniscus effect produces an inhomogeneously illuminated image, since the light from the sample does not get into the lens in equal proportions depending on the angle of incidence. The captured image thus has a bright potty center and dark border areas. Thus, a sample vessel-based shading arises, which can not be remedied by the known methods for shading correction.

The invention uses a brightness correction indication that is a function of the location on the bottom of the potty. The term "potty" is used to designate a sample vessel. It can be both a single vessel and a potty of a microtiter plate. As far as the singular ("potty") is referred to in the following description, it refers to both the use of a single vessel and the reference to a single well of a microtiter plate.

The test potty corresponds to the vessel containing the sample except for the difference that no sample is present in the test potty. If the sample adhering to the bottom of the potty is in a nutrient solution, i. If there is additionally a liquid in the potty, it is preferable to provide this liquid also in the test potty. Ultimately, the task of the test potty is precisely those optical conditions, i. to make the influence of the illumination radiation, which also prevails in the potty - but without a sample. Thereby, the brightness correction indication, which was determined from the test potty, is a function of the location at the bottom of the potty, can be determined from the test potty and then used to correct an uneven brightness distribution when imaging the sample in the potty.

The brightness correction specification can be provided in various ways:
In a first embodiment, the bottom of the test potty is imaged by multiple frames and a reference image is obtained. The brightness correction indication is then essentially the brightness distribution over the entire bottom of the test potty, ie over the reference image. The image of a potty with sample provides a sample image. To correct the sample image of the potty, which shows only a section of the bottom of the potty due to the enlarged image, the position of the sample image is determined on the ground. A corresponding cutout in the reference image then provides the required brightness correction data for the sample image.

In a second embodiment, a functional description of the brightness distribution takes place in the reference image. For each location at the bottom of the reference image of the test potty (eg, for each pixel) a correction factor is determined that reflects an additive or multiplicative deviation from a uniform brightness distribution. The correction of the sample image is then carried out by determining the position of the cutout on the ground, reading out the corresponding correction factors for this position from the brightness correction specification and applying the correction factors (either additive or multiplicative, depending on the design of the factors).

In a third embodiment particularly applicable to pots having a circular cross-section, the brightness correction indication includes a correction factor (again, either additive or multiplicative) which depends solely on the distance from the center of the bottom of the test potty. It is therefore sufficient for this embodiment to obtain in the reference measurement of step (d) only images covering only an area to the left of a radial coordinate from the center of the bottom of the test potty to its edge. The brightness correction of the sample image takes place in step (f) in this embodiment in that the radial coordinate of the Pixels of the image, ie the distance from the center of the bottom of the potty is determined. By reading out the corresponding correction values from the brightness correction specification by applying these correction values, the brightness correction is then achieved.

It is understood that the features mentioned above and those yet to be explained below can be used not only in the specified combinations but also in other combinations or alone, without departing from the scope of the present invention.

The invention will be explained in more detail for example with reference to the accompanying drawings, which also disclose characteristics essential to the invention. Show it:

1 two schematics illustrating the importance of the numerical aperture of a lens in transmitted light microscopy of samples arranged at the bottom of a microtiter plate,

2 five schematics similar to the 1 to clarify the importance of the interaction between the numerical aperture of the illumination and the image,

3 a schema representation similar to the 1 to illustrate a first embodiment of the invention,

4 a sectional view through a microtiter plate and a cover plate for illustrating a second embodiment of the invention,

5 a schematic representation of a microscope for imaging samples that are located on trays of pots of a microtiter plate,

6 a function for clarifying the intensity distribution at the bottom of a potty, which is caused by influences of the potty on the lighting,

7 a flow chart for explaining the general principle of the correction of uneven illumination of the bottom of a potty,

8th a schematic representation for explaining the correction according to 7 .

9 a flowchart for explaining a further embodiment of the correction and

10 a schematic representation for explaining a further embodiment of the correction.

1 schematically shows a sectional view through a potty 1 a microtiter plate, that of a light beam 2 is illuminated along an optical axis OA. On a floor 3 the potty is a sample not shown, for example, a cell culture. In the potty is a fluid 4 , For example, a culture medium for cell culture, on its surface a meniscus 5 formed.

The lighting beam 2 falls in the presentation of 1 from above. The thus illuminated floor 3 is recorded in transmitted light with a lens that has a collecting cone 6 having.

The sample in the potty 1 is usually in an aqueous environment, eg. A simple buffer or nutrient medium to allow live cell observation. This liquid 4 forms on its surface the meniscus 5 so the interface between air and medium is curved. The radius of the meniscus 5 depends on the type of liquid 4 , of wall material and coating of the potty 1 , as well as from the history of the filling, so whether dry or already wet potty 1 were filled, whether they were stirred, etc. In most cases, the liquid runs 4 along the potty wall slightly upwards, while the liquid level is lower in the middle of the potty.

This leads to a parallel incident illumination beam 2 after passing the meniscus' 5 diverges. This divergence is more pronounced the smaller the radius of the meniscus 5 is trained. For lenses with a high NA, so a large collecting cone 6 , that does not pose a problem, because the diverging rays can be absorbed. Numerous applications, however, require weakly magnifying lenses, which are intended to image a large field and thus provide an overview impression of a floor. For example, with just one image of a 2.5x magnifying lens, almost a complete floor can be created 3 the potty 1 a microtiter plate with 96 pots 1 be imaged.

Typically, however, low-magnification lenses have only a low NA, z. B. 0.08 or 0.12. The NA describes the maximum angle α of the collecting cone 6 which a beam can form with the optical axis OA so as to be still imaged by the lens, NA = n · sin (α). Here, n is the refractive index of the medium between the objective and the sample. When weak magnifying lenses, this is usually air, so n = 1. All the illumination rays passing through the meniscus 5 Broken to larger angles than the specified by the lens NA limit angle, thus not get into the lens. So even if lighting light every part of the floor 3 achieved (as in the left part of the 1 ), the meniscus effect produces an inhomogeneous image, since the transmitted light illumination does not reach the lens in equal proportions depending on the angle of incidence.

1 starts from the idealized situation of a parallel incident light beam 2 out.

However, typical transmitted light illumination provides a much wider angle spectrum, up to the angle described by the numerical aperture (NA) of the illumination. Beams of other angles of incidence have for the above-mentioned lens with NA = 0.12 in 2 illustrated illumination pattern on. In the five representations of 2 the numerical aperture of the illumination is 0.05; 0.1; 0.15; 0.2 and 0.25.

For every angle of incidence on the ground 3 the illumination pattern is different. However, with all geometries, the fringes are never or less than those of the lens. The taken picture arises from the summation of all these individual ray bundles. It will thus effectively have a bright potty center and dark edges. This consideration applies regardless of the type of transmitted light illumination, so whether it is Köhlersche, critical or a different kind of lighting. No matter what is above the potty 1 is undertaken, the effect described remains.

In a first embodiment, the in 3 is therefore a homogenizing element 7 used that at least with a part 7a into the liquid 4 dips. The homogenizing element 7 is made of a material which is suitable for the illumination beam 2 is transparent. Since the homogenizing element with the part 7a below the level of the liquid 4 standing, there is an end face 8th of the part 7a and the homogenizing element 7 in the liquid 4 , The front side 8th is planar in a first variant of the first embodiment and is perpendicular to the optical axis OA. The refraction effect through the meniscus 5 is turned off, and the illumination beam is on the ground 3 irradiated so that a uniform illumination takes place and in particular edge areas are illuminated as evenly as possible.

So that the illumination beam 2 at the front 8th so exits, that a uniform illumination of the soil 3 is reached, the homogenizing element is at its part 7b that is outside the liquid 4 is located, an upper side, which is also perpendicular to the optical axis OA. The homogenizing element is formed in the illustrated first embodiment as a generally straight cylinder.

Between an outer wall 9 the homogenizing element, which coincides in this construction with the end face 8th when viewed in a plan view along the optical axis OA, and an inner wall 10 the potty 1 There is a gap between 1 and 3 mm left to prevent air from reaching the front 8th is included when the homogenizing element 7 with his part 7a is immersed in the liquid. Such bubbles would uniformly illuminate the soil 3 severely impair.

4 shows a second embodiment in which not a single homogenizing element 7 is used, but several homogenization elements 7 be summarized. The potty 1 are in a microtiter plate 11 formed, and the homogenizing elements 7 are in the form of a cover plate 12 summarized, which is a flat top 13 having. The top 13 is the illumination beam source 16 associated with which the illumination beam 2 on the potty 1 gives up, its sample 14 currently being microscoped. This in 4 not shown microscope is designed so that each potty 1 individually microscoped. It has a drive, which the microtiter plate 11 with cover plate on top 12 appropriately shifts and aligns to the optical axis OA of lighting and imaging.

At the top 13 opposite underside of the cover plate 12 has these cylindrical projections, which are in training and function of the homogenization element 7 of the 3 correspond - but with the difference that the homogenizing elements 7 the plate 12 a common top 13 to have.

The design options for the cover plate explained in the general part of the description, which is also referred to as "cover" there, are suitable for the second embodiment 4 in question.

The embodiments may be supplemented or modified by the following features, which may be used individually or in combination, unless otherwise stated:
The homogenizing elements 7 it may take the form of a general cylinder, that is, it does not necessarily have to be one Act circular cylinder. Rectangular cross sections are for the homogenization element 7 equally possible as the realization of a general cylinder.

The front side 8th may be configured in an embodiment just and at an angle between 80 and 90 degrees to the optical axis OA and thus at an angle of 0 to 10 degrees to the flat bottom 3 lie. This facilitates the removal of air bubbles.

In another embodiment, the front side 8th convexly curved. In this case, the tangent angle is the curved end face 8th between 80 and 90 degrees to the optical axis OA and between 0 and 10 degrees to the plane of the ground 3 , This also facilitates the removal of air bubbles to the side.

The homogenizing element 7 is not formed in one embodiment as a solid body, but as a hollow body or pot-shaped. For the equalization of the edge illumination of the soil 3 it is only necessary that the end face 8th in the fluid 4 lies. The homogenizing element 7 can thus be designed either as a hollow body or cup-shaped. The homogenizing element 7 or the cover plate 12 is in one embodiment of an autoclavable material, for. As glass or an autoclavable plastic. In another embodiment, the cover plate 12 designed as disposable material.

5 schematically shows a microscope for high-resolution imaging of samples 14 floors 3 from pots of a microtiter plate 11 , The homogenizing element is not shown for the sake of simplicity. Already described elements and components are in 5 provided with the same reference numerals to avoid repetition of the description.

The microscope has an illumination beam source 16 on which the illumination beam 2 along the optical axis OA on the microtiter plate 11 emits. This is aligned to the optical axis OA, that a potty, which is a sample to be microscoped 14 contains, suitably in the illumination beam path, ie in the illumination beam 2 lies. From a bottom 15 the microtiter plate 11 the sample thus illuminated in transmitted light is produced 14 displayed. This is located under the microtiter plate 11 a corresponding imaging beam path of the microscope, of the example, a lens 17 as well as a receiver 18 are drawn. The imaging beam path lies on the optical axis OA, ie an illumination beam path, which is generated by means of the illumination beam source 16 the illumination beam 2 outputs, lies on the same optical axis OA as the imaging beam path.

A control unit 19 reads the data from the receiver 18 out. To the pots of the microtiter plate 11 individually to be able to map, this lies with its underside 15 on a sample table 20 on, via one from the control unit 19 controlled drive 21 so adjustable is that the individual pots of the microtiter plate 11 can be aligned to the optical axis OA.

In addition to further improving the illumination achieved by the homogenizing element, to further suppress the sample vessel based shading, a reference image is generated. This is done on a test potty. One of the pots of the microtiter plate 11 is considered the test potty 22 educated. It corresponds to the other pots of the microtiter plate 11 with the only difference being on the ground 3 the test potty 22 no sample 14 located. This causes the illumination of the soil 3 the test potty 22 subject to exactly the same conditions as the illumination of the soil 3 the potty, which samples 14 contain.

As already explained, has a potty 1 Effect on how the ground 3 the potty is lit up. 6 shows this by way of example in a diagram in which intensity I of the illumination, ie the brightness on the ground 3 as a function of the distance from the center of the potty 1 decreases. The center of the potty can correspond, for example, to the position of the optical axis OA. There is a maximum brightness, ie an intensity I ₁ before. The edge of the for 6 By way of example, the potty assumed to be a circular cylinder is located on a radius r _1. Due to the idealized circular cylindrical shape of the potty, only the radial coordinate r is of importance. At the edge of the potty, ie at r ₁ , the illumination intensity is minimal, z. B. zero. The intensity I remains almost constant with increasing distance r from the optical axis OA over a longer time point and drops off to the edge of the potty, ie at the radial coordinate r ₁ . This intensity profile of the illumination, which in the exemplified example of the 5 is a transmitted light, of course, affects the detection of a sample 14 out. Incidentally, a similar intensity course also occurs in the case of epi-illumination when the observation plane is not located directly on the bottom of the potty, but rather into the sample, so that the description given here is also valid for microscopes with epi-illumination.

The intensity drop to the edge is referred to in the literature as "shading". He is not through the beam path of the microscope, but generated by the sample vessel. For the correction of a ray-related shading all known in the literature methods in question, which will therefore not be described again here. However, if the sample vessel creates a sample vessel based shading, it can be removed from the sample image in the ways described below. In all these methods, the beam path-related shading is preferably removed (without this being described below) and optionally initially, so that only the sample vessel-based remains.

The sample vessel based shading is removed with a procedure shown schematically as a flowchart in FIG 7 is shown. 8th shows the individual pictures. In a step S1, the test potty becomes 22 placed in the beam path of the microscope. There is a reference recording of the soil 3 the test potty 22 ie without sample 14 , The resolution through the lens 17 is such that the object field of the lens 17 not the entire floor 3 can capture. The image therefore takes place in single images 23a by the soil z. B. is scanned in the form of a tile holder. For these shots, the test potty 22 perpendicular to the optical axis OA and parallel to the ground 3 postponed. This recording technique changes the influence of the test potty at each position 22 on the illumination of the soil 3 , This makes it clear that the shading to be corrected is a sample vessel-based shading and not a shading caused by the illumination beam path itself. Such would be completely independent of the positioning of the potty 22 ,

The single pictures 23a when scanning the soil 3 thus contain variations of the lighting of the soil 3 responsible for the location of the single image 23a on the ground 3 is characteristic. At the end of step S1 is a reference shot 24 of the soil 3 through the multiple frames 23a , for example in the form of a tile holder.

In a subsequent step S2, the brightness distribution in the reference image 24 determined. This provides a brightness correction indication because the deviation from ideal homogeneous illumination can be easily determined. The deviations depend on the location on the ground 3 the potty 22 from.

Steps S1 and S2 serve to determine a brightness correction indication. They are in the embodiment of 7 preceded by further steps S3 and S4, of which S3 is for microscopy of the sample (s) and step S4 is a brightness correction. However, steps S1 and S2 do not necessarily have to be performed before step S3. It is quite possible to execute the generation of the brightness correction specification by the steps S1 and S2 and the brightness correction in step S4 also at a later time and possibly even only when it has been found that the image of the samples 14 without such a correction is not sufficiently possible.

In step S3, a potty 1 pictured on its bottom 3 yourself sample 14 located.

In a step S4, a correction is made using the brightness correction indication; Step S4 thus assumes that steps S1 and S2 have been carried out. The correction is done by determining where on the ground 3 the sample image 23b in step S3. After determining this location information, the brightness correction indication corresponding to that location of the ground 3 is determined, and the sample image generated in step S3 23b becomes a corrected sample image 23c improved. The improved sample image 23c contains a homogenized illumination of the soil 3 , which is adjusted for influences of sample vessel based shading.

In the embodiment, the complete sample vessel - for microtiter plates z. B. a potty 1 - scanned, z. B. in the form of a tile recording with individual images 23a , Ideally, this will be a test potty 22 without a sample 14 used. If the sample vessel is a vessel in which the sample is in a liquid medium, the medium is preferably also in the test potty 22 in front. The composite reference image 24 thus contains the sample vessel based shading.

To the shading from a sample picture 23b with sample 14 It must be known at which point in the soil 3 the sample image 23b has been recorded. For this purpose, elements of the sample vessel based shading in the sample image 23b be recognized. Is in the sample image 23b z. B. a part of the edge of the potty 1 visible, the current object can be a corresponding section 25 of the reference picture 24 assigned and the shading from the sample image 23b to be counted out.

The simplest method is to get the intensity values of the pixels of the sample image 23b through the intensity values of the pixels of the matching section 25 of the reference picture 24 to divide and then renormalize the result. This calculation is also used in many methods for beam path based shading correction. The fundamental difference here is that the right reference image 24 is only determined from a larger overview image as a function of the observation position.

In this way, a corrected sample image 23c obtained, the opposite of the sample image 23b corrected for influences of the sample vessel based shading.

In this embodiment, the case may occur that a unique location of the location of the sample image 23b on the ground 3 and thus the choice of the cut-out 25 is not clearly possible. If, for example, no clear sample vessel edge is detected in the sample image 23b Because the sample field just does not capture the edge anymore, the sample vessel based shading can still be strong without it being apparent which of the matching section of the reference image is 24 is, which should be used for the shading correction.

But even in this case, a correction is possible, provided the drive 21 of the sample table 20 has a layer feedback of the sample positioning. This allows to store the xy coordinates of each point where a frame is 23a for the reference picture 24 was recorded, ie the xy positions of each point in the reference image 24 are known. Will now be the test potty 22 through the potty 1 with sample 14 replaced, the xy positions with respect to the potty change 1 Not. No matter at which point now the picture 23b can be recorded, the matching section 25 from the reference picture 24 determined and the correction can be carried out as described above.

For sample containers with several similar subunits such. B. a microtiter plate 11 with many same pots 1 , a single subunit - so z. B. a potty - as a reference vessel. The individual subunits are arranged in a fixed grid, which is either known from the manufacturer's instructions or can be easily measured by yourself. Thus, at each xy table coordinate pair on which a sample image 23b is recorded, directly on the position of the sample field within the respective potty 1 be closed and the matching section 25 in the reference picture 24 to be selected. One possible procedure is in 9 the example of a microtiter plate 11 shown.

In 9 are steps that follow the flowchart of 7 , denoted by the same reference numeral, wherein sub-steps are marked by appending a suffix. In a step S1.1 is for the test potty 22 the center point (x ₀ , y ₀ ) was searched. Subsequently, the reference image 24 of the soil 3 the test potty 22 , z. B. won by a tile recording.

In step S3, the brightness distribution of the reference image 24 converted into a brightness correction specification, for example the deviation from a mean brightness value in the form of an additive or multiplicative correction factor. This brightness correction specification depends on the coordinate (x, y) in the reference image 24 ,

Step S2 sees the recording of the sample image 23b at certain coordinates (x, y).

In a step S4 we get the distance vector r from the center of the sample image 23b calculated to the center of the corresponding potty.

In a step S4.2, a corresponding correction specification for the determined section is made for this distance vector r 25 determined in the brightness correction specification.

Finally, step S4.3 provides the correction of the sample vessel based shading by applying the brightness correction indication for the region defined by the distance vector r with respect to its center.

The brightness correction indication can be either the shape of the reference image 24 or the shape of a correction value image calculated therefrom which has a uniform brightness when linked to the reference image (either additive or multiplicative). If one sees the brightness correction indication as reference picture 24 before, the determination of the correction factors must be made from the brightness distribution of the corresponding section 25 that from the reference picture 24 has been extracted, follow in the correction step S4. By contrast, if the brightness correction specification already contains the correction factors, then the correction factors have already been determined on the reference image 24 executed, this does not have to be done again in step S4. With regard to the meaning for the correction of the sample vessel based shading thus the terms reference image 24 and brightness correction indication either identical (first option), or the reference image 24 represents a precursor of the brightness correction specification (second option), wherein the conversion is a simple mathematical operation (eg determination of the multiplicative or additive deviation from a mean value, etc.).

10 shows the procedure in the embodiment of 9 , Shown is the reference picture 24 as brightness correction specification. 10 shows. the course of the 9 once again based on the reference image 24 , In the reference picture 24 is the center 26 known. For a sample picture 23b the distance vector r is known. It denotes the center of the sample image 23b , For this distance vector r is now in the reference image 24 the cutout 25 which has the same distance vector r to the center and the same dimensions as the sample image 23b , The cutout 25 is then the excerpt from the Reference image (or the brightness correction specification) to be used for the correction.

If the pots are circular, it may be sufficient in a simplified embodiment to determine only the amount of the distance vector r. For an improvement of this embodiment, the magnitude of the distance vector is determined not only for the center but for each pixel.

Another possible embodiment is the reference image 24 (or the brightness correction specification) in absolute pixel coordinates of the reference image 24 indicate, with these pixel coordinates z. B. by evaluating the position feedback of the drive 21 of the sample table 20 be won. For a sample picture 23b to be corrected then become only the location coordinates of the pixels in the sample image 23b determined, these locations on the reference image 24 (or the brightness correction specification) are related. The location information then allow it, from the reference image 24 (or the brightness correction indication) to read out and apply the corresponding correction factors for each pixel.

The following modifications and developments for the additional correction by reference image in question:
Of course, the reference picture must 24 not present as a composite image. It is sufficient, the corresponding frames 23a to save and from them as needed, the appropriate correction image for the sample image 23b to calculate.

The process shown does not necessarily have to be processed in this order. For example, the reference image 24 be added later to the shading correction of the sample images at a later date 23b make.

So far, the case has been described in which the detector 18 only one part of the potty is imaged and therefore for complete generation of the reference image 24 several frames 23a are necessary. With sufficiently weak magnifying lenses and sufficiently small sample containers, or pots 1 However, even with a single shot, it can be the complete floor 3 be imaged; a tile recording is no longer necessary. Otherwise, the procedure described above remains valid.

If the sample vessel is a round vessel, it is not necessary to use the complete vessel as a reference image 24 when the effect of shading caused by the edge of the vessel is to be corrected. In this case it can be assumed that the shading is also radially symmetric with respect to the center of the vessel. This is z. B. in Petri dishes or circular potty 1 a microtiter plate 11 the case. It then suffices to record a line from the center of the vessel to the rim for reference, I ₀ (r). In order to be able to generate the appropriate correction image for a sample image, the distance of each image pixel r _p = (x, y) from the potty center r _{0 is still} needed, ie r = | r _p -r ₀ |. Whether from these distances r for each pixel directly the correction value I ₀ (r) on the sample image 23b applied, or first a complete reference image 24 is generated, with which the sample image is then charged, does not matter. In both variants, the sample vessel-based shading can be removed from the sample image.

Depending on the lens magnification used and the size of the sample container, the reference image 24 from many single frames 23a exist whose recording takes a lot of time and their storage or processing much storage space. In general, however, different lenses will show different degrees of sample-dependent shading. That's possible with a potty 1 a microtiter plate 11 well illustrate. As already explained above, the fluid meniscus provides 5 in the potty 1 that incident rays are refracted to higher angles, especially those near the bottom of the potty 3 happen. A lens 17 with a weak NA, this light can often no longer use, the potty edge appears much darker than the center. On the other hand, a lens collects 17 with a higher NA far more of this light, which is why the Randabschattung is lower, the sample vessel based shading is another. But the difference is only in the strength of the shading, not in its kind. With both lenses you measure this shading strength in the same object field, in principle one picture in the middle of the vessel and one on the edge suffices to use the factor calculate that the two shadings are different from each other. Then it is sufficient, with the weaker magnifying lens, the reference image 24 and to use this also for the more magnifying lens, taking into account the corresponding shading strength factor. In this way, the reference picture 24 from significantly fewer frames 23 a, which saves time and space as well as computational power. In addition, a new reference image for correcting the sample vessel based shading does not have to be taken when changing the objective.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 10200541 A1 [0003]
• US 6074614 [0004, 0011]
• US 6238911 [0007]
• US 2007171417 [0008]
• WO 02087763 [0009]
• US 2010197004 [0010]

Claims (15)

Method for the microscopic imaging of samples ( 14 ) contained in a microtiter plate ( 11 ) containing a variety of fluid ( 4 ) filled pots ( 1 ), on soils ( 3 ) the potty ( 1 ), the pots ( 1 ) successively from one of the soils ( 3 ) opposite the top from a transmitted light illumination source ( 16 ) along an optical axis (OA) with illumination radiation ( 2 ) and the so-lit from the top ( 3 ) the potty ( 1 ) from the bottom ( 15 ) are shown individually enlarging, characterized in that to improve an edge illumination of the soil ( 3 ) with every illuminated potty ( 1 ) between the potty ( 1 ) and the transmitted light illumination source ( 16 ) a homogenizing element ( 7 ) arranged by the transmitted light source ( 16 ), for the illumination radiation ( 2 ) is transparent, and one in the potty ( 1 ) and there into the fluid ( 4 ) immersing part ( 7a ), which one in the fluid ( 4 ) lying end face ( 8th ) Has. Process according to claim 1, characterized in that the fluid ( 4 ) immersing part ( 7a ) has the shape of a general, straight cylinder. A method according to claim 1 or 2, characterized in that the end face ( 8th ) has an inclination angle or tangent angle to the optical axis (OA) which is between 80 and 90 degrees. Method according to one of the preceding claims, characterized in that the shape of the end face ( 8th ) in plan view along the optical axis of the shape of the cross section of the pots ( 1 ) is similar. Method according to one of the preceding claims, characterized in that between the edge ( 9 ) of each end face ( 8th ) and an inner wall ( 10 ) of the associated potty ( 1 ) a gap of at least 1 mm and at most 3 mm is left to allow air bubbles when immersing the part ( 7a ) not under the front ( 8th ) but to let escape. Method according to one of the above claims, characterized in that a plurality of homogenizing elements ( 7 ) in the form of a cover plate ( 12 ), on whose underside the parts immersed in the fluid ( 7a ) are designed and for laying on the microtiter plate ( 11 ) is formed. Method according to one of the preceding claims, characterized in that in the enlarging image of at least one floor ( 3 ) a test image is taken and in a correction step on the basis of this test image, a value of a parameter of a suppression of a ring structure of a brightness distribution in the test image is optimized and that in subsequent images of other soils ( 3 ) the ring structure of the brightness distribution is suppressed by means of the correction step using the optimized value of the parameter. Method according to one of the preceding claims, characterized in that in the enlarging image of the soil ( 3 ) for suppressing a ring structure of a brightness distribution in images of the soils ( 3 ), which images are subjected to low-pass filtering. Method according to one of the preceding claims, characterized in that one on the test potty ( 22 ) a reference measurement is carried out, wherein a reference image ( 24 ), which covers the entire floor ( 3 ), in the reference image ( 24 ), a brightness correction indication is determined, the brightness correction indication being a brightness fluctuation as a function of the location on the ground ( 3 ) of the test potty ( 22 ) indicates the image ( 23b ) of the soil ( 3 ) of the sample-containing potty ( 1 ) is corrected by means of the brightness correction specification, the position of the image ( 23b ) on the ground ( 3 ) and the value of the brightness correction indication associated with this position is determined. Method according to claim 9, characterized in that a plurality of individual images ( 23a ) the entire floor ( 3 ) of the potty ( 1 ) and are combined to the reference image and the brightness correction indication the brightness fluctuation in the reference image ( 24 ) as a function of the place, and for the picture of the ground ( 3 ) in the assigned section ( 25 ) in the reference picture ( 24 ) and from this for this excerpt ( 25 ) valid brightness correction information is determined. A method according to claim 9 or 10, characterized in that the brightness correction indication the brightness variation as a function of a position to the center of the soil ( 3 ) indicates. A method according to claim 11, characterized in that the potty has a round cross-section and that the function is solely a radial coordinate defining the distance from the center of the bottom ( 3 ) designated. Method according to one of claims 9 to 12, characterized in that the potty ( 1 ) when imaging by means of a sample adjustment mechanism (US Pat. 20 . 21 ) is moved relative to an optical axis (OA) and that the brightness correction indication the brightness variation as a function of the adjustment of the Probenverstellmechanismus' 20 . 21 ) indicating, in particular as a function of a coordinate (x, y). Method according to one of claims 9 to 13, characterized in that the brightness correction indication the brightness variation as a function of a lens used for imaging ( 17 ) indicates. A method according to claim 14, characterized in that for the imaging in step (b) a plurality of objectives ( 17 ) with different numerical aperture, the reference measurement in step (d) with one of the objectives ( 17 ), preferably the one with the smallest numerical aperture, and that the brightness correction indication indicates the brightness fluctuation as a function of a shading strength factor, which depends on the numerical aperture of the image for the sample-containing pots (FIG. 1 ) used lens ( 17 ) depends.

Images