WO2009080790A1

WO2009080790A1 - Method for determining the radius and/or the position of characteristic eye elements

Info

Publication number: WO2009080790A1
Application number: PCT/EP2008/068103
Authority: WO
Inventors: Thomas Schuhrke; Günter Meckes
Original assignee: Carl Zeiss Surgical Gmbh
Priority date: 2007-12-21
Filing date: 2008-12-19
Publication date: 2009-07-02
Also published as: DE102007055924A1; DE102007055924B4

Abstract

The invention relates to a method for determining the radius and/or the position of characteristic eye elements during the examination or treatment of an eye. According to this method, a digital image of at least one section of an eye is recorded by a camera. The method according to the invention is characterized by correlating the image with annular comparative objects of different sizes in such a manner that the agreement between the image and the reference object is highest when an annular reference object coincides with an annular density jump of the same radius in the image. The reference objects having a high local agreement with the digital image are then determined and the radius and/or the position of the characteristic eye element is/are derived from the reference objects of high agreement.

Description

Method for determining the radius and / or the position of characteristic eye components

The invention relates to a method for determining the radius and / or the position of characteristic eye components according to the preamble of claim 1 and a corresponding device according to the preamble of claim 22.

Such methods are known, for example, in the field of ophthalmic surgery.

For example, in corneal surgery for the removal of defective vision of the human eye (LASIK), in which a part of the cornea is ablated by means of a laser, it is of interest to the surgeon at which point the visual axis of the patient pierces the cornea. By accurately determining this point during surgery, laser ablation can be more precise from this point than when choosing a theoretically assumed or estimated midpoint of the cornea.

Another example of this is a cataract surgery in which a natural lens of the human eye, which has become clouded, is replaced by an artificial lens. The surgeon performs such an intervention under a surgical microscope. After a circular opening of the front capsule leaf usually the lens is smashed and sucked. Subsequently, an artificial lens is inserted into the empty capsular bag.

From DE 10 2004 055683 A1 a surgical microscope for eye surgery is known, which superimposes a pattern on the eye to be operated on. The pattern may aid in setting the cutting position, but it may also serve as a guide in the placement of toric intraocular lenses, or may assist in the insertion of a suture in a corneal transplant. To position the pattern in the right place, it is necessary to determine the position of the pupil or iris on the eye to be treated. Ideally, the position will also be during The operation again and again determined or tracked, as it may come during the procedure to movements of the entire eye or the pupil.

It is also of fundamental importance for other applications in the field of ophthalmic surgery to determine the position or diameter of the iris of the eye to be treated. For example, the diameter of the iris is necessary to calculate the strength of an intraocular lens to be implanted after cataract surgery. This and other possible applications, as well as a method for determining positions and orders of magnitude within an eye section are discussed in more detail in DE 101 08 797 A1.

Some methods are known in which the position of the pupil is determined on the basis of the current image of the eye section to be operated, which is obtained with the camera on the surgical microscope. Both DE 10 2004 055683 A1 and DE 101 08 797 A1 propose methods in which a binary image is first generated by means of thresholding in order to determine the dark regions in the image. Thereafter, the largest contiguous region in the dark regions is identified, which is identified as a pupil. In order to determine the edge of the pupil or iris in more detail, an edge detection is usually performed in this method. These methods have some disadvantages. On the one hand, the pupil is not always the largest contiguous dark area; on the contrary, the pupil may be disturbed by a reflex and have a completely different appearance. On the other hand, the edge detection during the introduction of microsurgical instruments during surgery can be severely impaired. Basically, it is difficult to define a meaningful threshold value for all methods which work with thresholding, which on the one hand does not leave too much information in the picture, but on the other hand does not remove important details from the picture.

The invention has for its object to develop a method for determining the radius and / or the position of characteristic ocular components, which is robust against interference and works independently of the individual design of the eye reliably. The object is achieved according to the invention by a method for determining the radius and / or the position of characteristic eye components with the features of claim 1 and a corresponding device having the features of claim 22.

According to the invention, based on the correlation of the digital image acquisition to be analyzed with an annular comparison object, varying its radius or size, the comparison object is pushed over the image, its match with a recorded in the image characteristic eye component, preferably the limbus or the pupil determined. As soon as the comparison object and the characteristic eye component, which is preferably likewise annular or circular and a density transition, coincide in position and size, the greatest value of the match results. If the comparison object and the image are in great agreement, it can be assumed that an object corresponding to the comparison object, that is to say a characteristic eye component, has been determined in the image. Great agreement means that the match is in the range of a local maximum match and that its value does not fall below the value of the largest match by more than 20%, preferably not more than 5%. Thus, the radius and / or the position of the comparison object in which great agreement has been established can be selected as the radius and / or position of a characteristic eye component. Since there may be several such objects in the image, there may be multiple comparison objects and multiple positions of those where there is a large match. In particular, it is advantageous to select the position and / or radius of the comparison object with maximum agreement, since it can be assumed that this matches best with the desired characteristic eye component, but it is also possible to have a region of locally large match, preferably one Deviation of less than 20%, preferably less than 10%, ideally less than 5% of the maximum of the match, to select a comparison object. For this purpose, it may be advantageous to consult further criteria such as, for example, the color of the characteristic ocular component, the gradient of the increase in density or the like, in order to thereby prefer a determination of the characteristic ocular component based on additional features. The invention is based on the recognition that in the case of the images of the eye taken during an eye examination or treatment, the edges of the characteristic ocular components, in particular iris and pumice pill may be seen as essentially circular or annular density transitions, the limbus / pupil radius respectively representing the largest / second largest annular density transition object within the eyelid, and this being based on the comparison with a corresponding comparison object, eg via the convolution with a corresponding one annular filter is preferably particularly easy to find under difference formation and reliable. However, the absolute size of the limbus / pupil radius is unknown at the beginning of the procedure. Therefore, the size of the comparison object is varied so that the density transition object best matching the limbus / pupil radius can be determined. This best match, in which, in the case of a filter, the result of the filtering, ie the correlation function, shows a maximum, arises precisely if the comparison object has the same radius as the characteristic eye component, which is the density transition object in the image see, has and the centers of the two lie on each other. The search for the largest / second largest ring-shaped density transition object in the image is extremely robust against impairments that image-dominating features during the operation can distort the recording, for example, if instruments are visible in the recording or the entire eyeball in the course of the Operation is deformed or compressed. All this does not change the fact that the limbus / pupil radius continues to exist in the image, albeit as an exposed or slightly deformed annular element. It should also be stressed at this point that an absolute thresholding, which is necessary in the conventional edge detection method, and thus the problem of choosing a suitable threshold by this method can be completely avoided.

The limbus radius is preferably determined in the method. It has been shown that this is always reliably regardless of the recording quality, the properties of the eye and the course of surgery, the largest annular structure in the recorded section. While the pupil can be severely impaired by the operation and can hardly be distinguished from the iris even in the case of very dark eyes, the transition from the limbus to the whites of the eyeball can always be reliably determined. In particular, it is always the largest light-dark transition within the eyelid and thus clearly and reliably to identify. Also different with it This method relies heavily on the methods presented in the prior art, which generally aim to locate the pupil.

In an advantageous embodiment, an annular comparison object is selected for locating the limbus / pupil radius, which contains at least two concentric annular constituents. The fact that the comparison object has at least two components results in the possibility, in each case one component to the eye area outside the density transition, e.g. the sclera, and the second component to the eye area within the density transition, e.g. the iris, adapt. By means of these two components, the density transition can thus be strengthened to a certain extent via a correlation with the comparison object. The optimum match with the comparison object results when the inner ring of the comparison object is e.g. on the iris, the outer e.g. on the sclera and thus the limbus edge is enclosed by the two annular components. At the same time, the center of the eye coincides with the center of the object of comparison. In this design of the comparison object not only the shape feature, so the ring or circular appearance of Limbus- / Pupillenradius, but also the surface feature, the density transition at Limbus- / pupil radius is used to search for this. Ideally, the two components of the comparison object are two narrow annular components. These are advantageously spaced apart so far that one component is as completely as possible in the region of lesser, the other in the region of higher density, but neither of the two is in the range of density increase. This allows a clear identification of the object. The rings are as narrow as possible, since this ensures that as few influences of other eye areas as possible are detected so that a clear correlation function is to be expected. In principle, however, the method also works if, for example, the inner annular component is designed as a disk. However, the reliability of the process would suffer somewhat.

In a further advantageous embodiment of the invention takes place within the context of the correlation of the comparison object with the recording, a difference of the annular components of the comparison object or the correlated with these areas within the eye detail. Thus, in the correlation, preferably one annular component of the comparison object may have a positive, the other a negative sign. will see. Ideally, the comparison object or its annular constituents are designed so that when correlated with a gray area, ie a uniform area without density transition, it yields a neutral result such as zero, whereas when it is correlated with an area in the area of a density overlap. With increasing strength of the transition, the result is ever greater values.

In a preferred embodiment, the comparison object is realized by a filter with which the image is folded. The annular filter is chosen so that there is a maximum filter response whenever the annular filter comes to rest on an annular density transition such as the limbus or pupil radius. Since the radius of the limbus and pupil in the image is not known in advance, the image is folded with annular filters of different radius and the maximum filter response per filter radius is determined in each case. The better the radius of the filter matches the radius of the searched object, ie the limbus or the pupil, the larger this maximum filter response will fail, it will reach a maximum if the radii agree. The radii which belong to the locally maximum values of the maximum filter response determined per image in each case thus correspond to the radii of annular density transfer objects in the image. It has been recognized that the largest of these objects found in the picture is always the limbus radius for these special shots and the second largest, usually the pupil radius. Thus, the limbus radius or, correspondingly, the pupil radius can be determined by searching for the locally maximum values of the maximum filter responses which match the corresponding filter radii. The location of the maximum filter response for this radius then corresponds to the center of the iris / pupil. The great advantage of this filtering method is that it can be processed very easily and quickly relatively large amounts of data, whereby the detection of the eye components and thus the provision of assistance to the surgeon or optician can be done so quickly that it does not affect its operation is.

Such a filtering method can work so fast, in particular, that any recording recorded with the camera can be processed immediately and, to a certain extent, assisted in real time if, in a particularly advantageous embodiment, the radius of the filter or of the comparison object is only once in a localization step is determined, but this radius thereafter in the further course of Eye exam or treatment is recorded. Due to this two-stage structure of the method, in which all variables are determined in detail in a first step, but in all further steps all variables which change little are recorded, now only the best correlation with the then defined comparison object, or the maximum filter response for a fixed filter are determined in order to always follow the changing location of the eye center exactly.

Advantageously, the radius of the comparison object is automatically adapted to changes in the recording conditions. By taking into account changes in the device settings that affect the limbus size, such as a change in the zoom factor on the microscope, automatically in the correlation of the comparison object with the recording, it can be ensured that the once adjusted radius over the entire Process remains consistent and can be used consistently. This avoids the need to re-determine the radius whenever a changed setting has been made on a device related to the recording. As a result, an interruption of the display of assistance for the surgeon or optician, which would be unavoidable in the case of a new determination of the radius, be avoided. In order to be able to adapt the comparison object, it is necessary for there to be an interface between the device which alters the device parameter, for example the microscope and the device at which the analysis of the image takes place.

Although the annular configuration of the comparison object is important, it would not change anything essential to the method if a polygon or something similar were used. It is also not necessary that a closed ring is used. The comparison object may as well be composed of annular segments. Essential for the process is only that the overall annular character of the object to be compared is preserved. In particular, in the edge region of the image, it is even more reliable to use only ring segments. In the case of these ring segments, it is preferable to expose the region which lies at the edge to which the comparison object approaches in the correlation and thus also the limbus in the image. In the correlation, the comparison object thus better corresponds to the object to be found, which, as soon as it reaches the edge of the image, is partially cut off. In a further preferred embodiment, the red-extraction of the image is always used for the correlation with the comparison object. Surprisingly, it has been shown that this is the least affected by disturbances during eye treatment, since in this color separation, the red of the hemorrhages and veins with the white of the sclera forms a homogeneous surface. This makes it possible to achieve a more reliable result in this color channel than in other color separations.

Further details and advantages of the invention will become apparent from the dependent claims in conjunction with the description of an embodiment which is explained in detail with reference to the drawings.

Show it:

Fig. 1 shows schematically an apparatus for carrying out the inventive

process

FIG. 2 shows an example of an advantageous ring filter superimposed on a receptacle of an eye cutout, FIG.

3 shows an example of a filter response, and FIGS. 4 and 5 show examples of filter responses over the radius.

Fig. 1 shows a schematic representation of the basic structure, as it is typical for an eye treatment in which the method according to the invention can be used particularly advantageously. The patient's eye 1 to be treated, which is illuminated by a light source (not shown), is observed firstly by means of an eyepiece 2 and secondly by means of a video camera 3, wherein the observation beam path is split by a beam splitter 4 into two observation beam paths for the observing instruments. The data recorded on the video camera 3 are transferred to a computing unit 5, where the data is stored and analyzed. Based on the data, an auxiliary pattern formed by a pattern generating unit 6 and superimposed on the image visible in the eyepiece 2 is calculated, so that the surgeon 7 can view the eye 1 to be treated together with the superimposed pattern formed on the pattern generating unit 6 , The Mus- Tererzeugungseinheit 6, for example, as a projector with an annular LED display, which overlays a pattern on the beam splitter 4 in the eye, be executed.

In a cataract operation, the eye 1 is continuously recorded digitally in very short time sequences with the camera 3 or analog recorded data is converted to digital and the digital data of the recording of the eye detail, as shown for example in Fig. 2 (to illustrate the comparison object with superimposed ring filter ) is transmitted to the arithmetic unit 5. There, according to the method according to the invention, the eye center and the limbus radius are determined, so that the optimum cutting position for the cut, for the removal of the clouded and for insertion of the artificial lens, can be determined. Once this cutting position is determined, the image seen by the eye of the surgeon 7 through the eyepiece 2 is superimposed on a pattern generated on the pattern generating unit 6 which indicates this cutting position. As a result, the eye of the surgeon 7 always sees during the treatment the optimal cutting position for preparing the cut.

The course of the method according to the invention for determining the limbus radius and thus localization of the eye center will be explained below with reference to FIGS. 2-5. The image taken on the camera 3 of a section of an eye to be treated, as can be seen in FIG. 2, is folded on the arithmetic unit 5 with the ring filter 8 shown schematically in the receptacle. The ring filter 8 includes two concentric rings 9 and 10, which are placed in Fig. 2 symmetrically around the examined Limbusradius 1 1. The ring filter 8 is normalized so that the outer ring 9 provides positive contributions to the filter response while the inner ring 10 gives negative contributions. In addition, the ring filter 8 is normalized so that the filter response in the convolution with a gray area is zero. This means that the two rings 9 and 10 are weighted according to their area proportions in the picture. This filter is now folded with the image section, that is, the filter response is determined at each point of the image.

The result of the convolution with a ring filter 8 such as that shown in FIG. 2, ie the filter response when the filter center and limbus center 11 are approximately identical, is shown in FIG. 3 as an example. At the location where the filter radius and limbus radius 11 exceed lie to each other, the maximum filter response, which is shown here bright. The center of the brightest area corresponds to the center of the eye and as such is transferred to the pattern generation unit 6. However, the exact determination of this center is only possible if the Limbusradius 1 1 or the radius of the best matching ring filter 8 has been determined. At the beginning of the process this is still unknown. In order to find it, therefore, a folding of the image section is performed with filters for a radius range to be examined. The image is folded with a filter of a different radius and the respective maximum filter response is determined. The resulting maximum filter responses are each plotted over the associated radius. As a result of this investigation, a curve is obtained, as shown for example in FIG. At the best fitted radius, the curve shows a distinct peak. Depending on the brightness of the pupil and the characteristic coloring of the iris, however, at least a second maximum may occur at the pupil radius, often still another between them. An example of this is shown in FIG. However, it could be proved that the first pronounced maximum starting from the largest radius always represents the filter response for a filter whose radius corresponds to the limbus radius 1 1. This defines the radius at which the first pronounced maximum can be seen as the limbus radius 1 1. The limbus center corresponds to the location of the maximum filter response which was determined during the convolution with the ring filter whose radius corresponds to the limbus radius 1 1.

LIST OF REFERENCE NUMBERS

1 eye

2 eyepiece

3 video camera

4 beam splitters

5 arithmetic unit

6 pattern generation unit

7 Eye of the surgeon

8 ring filters

9 Outer filter ring

10 inner filter ring

1 1 limbus radius

Claims

claims

1 . A method for determining the radius and / or the position of characteristic ocular components, during an eye examination or treatment, wherein a digital image of at least a section of an eye is taken with a camera, characterized in that

the image is correlated with ring-shaped comparison objects of different sizes,

- That the greatest match between image and comparison object results in the meeting of an annular comparison object with an annular density jump of the same radius in the image, and

- The comparison objects are determined with locally large agreement with the digital image and

- From these comparison objects of great agreement, the radius and / or the position of the characteristic eye component is derived.

2. Method for determining the radius and / or the position of characteristic ocular components according to claim 1, characterized in that the radius and / or the position of the characteristic ocular component is derived from comparison objects whose agreement with the digital image is less than 20% of that of FIG deviates locally largest match.

3. Method for determining the radius and / or the position of characteristic ocular components according to claim 2, characterized in that the radius and / or the position of the characteristic ocular component is derived from comparison objects whose agreement with the digital image is less than 10% of that of FIG deviates locally largest match.

4. Method for determining the radius and / or the position of characteristic eye components according to claim 3, characterized in that the radius and / or the position of the characteristic eye component is derived from comparison objects whose conformity to the digital image is less than 5%. deviates from the locally largest match.

5. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that the radius of the comparison object with locally greatest agreement as the radius of the characteristic

Ophthalmological component and / or the position of the comparison object is selected at the local largest match as a position of the characteristic eye component

6. A method for determining the radius and / or the position of characteristic eye components according to claim 1, 2 or 5, characterized in that the radius and / or the position of the largest comparison object locally large match selected as the radius and / or position of the eyelid becomes.

7. A method for determining the radius and / or the position of characteristic eye components according to claim 1 or 6, characterized in that the radius and / or the position of the largest comparison object locally large agreement within the eyelid as limbus radius and / or position selected becomes.

8. A method for determining the radius and / or the position of characteristic eye components according to claim 7, characterized in that the radius and / or the position of the largest comparison object locally large match within the limbus is selected as the pupil radius and / or position.

9. Method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that the annular the same object has two concentric annular components.

10. A method for determining the radius and / or the position of characteristic eye components according to claim 9, characterized in that a difference of the annular components takes place.

1 1. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that

- folded the digital image with annular filters of different radius, - each determined the maximum filter response and

- The largest, a local maximum value of a maximum filter response resulting radius is determined as the radius and / or its location of the maximum filter response as the position of the limbus.

12. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that

the digital image is folded with annular filters of different radius,

- In each case determines the maximum filter response and

- The second largest, a local maximum value of maximum filter response resulting radius is determined as the radius and / or its location of the maximum filter response as the position of the pupil.

13. A method for determining the radius and / or the position of characteristic eye components according to claim 1 or 9, characterized in that - the digital image is folded with annular filters of different radius,

in each case the difference of the filter responses of two different, adjacent radii is formed,

- local maximum differences and the radii associated with these are determined and - the smaller of the largest radii determined as radius and / or des- sen location of the maximum filter response is determined as the position of the limbus.

14. A method for determining the radius and / or the position of characteristic eye components according to claim 13, characterized in that the smaller of the second largest to a maximum difference radii associated radius and / or its location of the maximum filter response is determined as the position of the pupil ,

15. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that the determined radius is recorded in the analysis of further recordings during the same eye treatment as the radius for the annular comparison object in order to determine the position of the characteristic Eye component in each recording to determine.

16. A method for determining the radius and / or the position of characteristic eye components according to claim 15, characterized in that the determined radius is automatically adjusted when changing the shooting mode of the camera.

17. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that a ring is used as a ring-shaped comparison object

18. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that a polygon is used as a ring-shaped comparison object

19. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that a disc is used as the annular comparison object

20. A method for determining the radius and / or the position of characteristic eye components according to claim 1, characterized in that a suspended ring is used as a ring-shaped comparison object.

21. A method for determining the radius and / or the position of characteristic eye components according to claim 20, characterized in that the exposed ring is used in the edge region of the digital image.

22. Device for determining the radius and / or the position of characteristic ocular components, during an eye examination or treatment, with a camera for taking a digital image of at least a section of an eye and an image processing device for evaluating the image, characterized in that - the image processing device the image is correlated with ring-shaped comparison objects of different sizes,

that the greatest correspondence between the image and the comparison object is obtained when an annular comparison object coincides with an annular density jump of the same radius in the image, and the image processing device determines the comparison objects with locally large conformity to the digital image and

- Derives from these comparison objects of great agreement the radius and / or the position of the characteristic eye component.

23. A computer program for carrying out the method for determining the radius and / or the position of characteristic ocular components, during an eye examination or treatment, wherein a digital image of at least a section of an eye recorded with a camera is analyzed, characterized in that - the image with annular comparison objects of different sizes so is correlated,

that the largest match between the image and the comparison object results from the coincidence of an annular comparison object with an annular density jump of the same radius in the image, and the comparison objects with locally large correspondence with the digital image

Image to be detected and