Computer-implemented method, evaluation device, X-ray machine, program and data carrier

Abstract

The invention relates to a computer-implemented method for evaluating a two-dimensional X-ray image (1) showing a contrast agent in a hollow organ structure (2).

Description

Technical Field

[0001] The invention relates to a computer-implemented method for evaluating two-dimensional X-ray images recorded by an X-ray machine using X-ray radiation of the X-spectrum, which X-ray images show contrast agents in hollow organ structures, in particular in the vascular tree, in the recording region of a patient. The invention also relates to an evaluation device, an X-ray machine, a computer program and an electronically readable data carrier. Background Art

[0002] The use of contrast agents in X-ray imaging is widely known in the prior art. Here, a particularly commonly used component in contrast agents is iodine. Compared with other substances normally present in the human body, iodine absorbs X-ray radiation significantly more violently, so that hollow organ structures, especially blood vessels filled with iodine show strong contrast. In particular, for the case of iodine intolerance, other contrast agents, such as carbon dioxide or contrast agents containing carbon dioxide, are also used. There are different application fields in this regard, such as digital subtraction angiography (DSA), in which contrast agents are placed as so-called "contrast agent bolus" in the patient's vascular tree to be examined, wherein, in the DAS process, X-ray images are taken when contrast agents are used and X-ray images are taken when contrast agents are not used, and they can be subtracted, so that only the blood vessels filled with contrast agents are retained. However, contrast agents are also used in other areas to highlight specific hollow organ structures, in particular blood vessels in the vascular tree, for example in fluorescence microscopic monitoring, in particular in minimally invasive surgery, or in radiological and / or fluorescence microscopic examinations of other vital signs, such as radiological and / or fluorescence microscopic examinations of the blood flow in specific tissue areas. In the case of the vascular tree, the injected contrast agent flows with the blood in the vessels. The flow of the contrast agent is usually measured with the aid of an image set of X-ray images at a specific image acquisition rate (frame rate). This is usually a two-dimensional X-ray image, wherein the use of three-dimensional contrast agent images has also been proposed, which is obviously also based on the acquisition of two-dimensional X-ray images, for example projection images from different projection directions.

[0003] For the best possible visibility of the contrast medium and thus of the hollow organ structure, it is helpful to achieve a high filling degree of the hollow organ with contrast medium, since this results in the highest contrast in the X-ray image. The amount of contrast medium and thus the filling degree can be controlled during the injection of the contrast medium by means of the filling pressure, which is usually done manually or with an automatic contrast medium dosing device, such as a contrast medium injector. However, if the filling pressure is too high, there is a risk of damage to the hollow organ, such as rupture of a blood vessel, since the hollow organ can only accommodate what it normally transports, such as the blood transported in a normal blood vessel.

[0004] In the case of manual administration of contrast medium, the filling pressure may be controlled based on a corresponding mechanical feedback, wherein as another possibility, the contrast of the hollow organ in an X-ray image already taken during the injection can also be used. However, the method described in the former method presupposes a great deal of experience, and the appearance of the X-ray image is also closely related to other parameters that must be taken into account. Summary of the invention

[0005] Therefore, the object of the present invention is to provide an improved, in particular real-time, feedback on the contrast agent feed.

[0006] To solve this technical problem, the invention provides a computer-implemented method, an evaluation device, an X-ray machine, a computer program and an electronically readable data carrier according to the independent claims. Advantageous embodiments are given by the dependent claims.

[0007] According to the invention, a method of the above type comprises the following steps:

[0008] - Partitioning hollow organ structures in X-ray images,

[0009] - determining the midline and the boundaries of the partitioned hollow organ structure and, in the partitioned hollow organ structure, determining the respective diameter of each hollow organ along its midline,

[0010] along the center line of each hollow organ, in particular for each position along the center line for which the diameter has been determined, determining an image absorption value which describes the absorption enhancement due to the contrast agent in the hollow organ in comparison with the surroundings of the hollow organ in the x-ray image,

[0011] - determining for each image absorption value a theoretical calculated absorption value that would be expected if the hollow organ were completely filled with contrast medium, using at least the diameter and contrast medium information describing the contrast medium composition and / or absorption properties,

[0012] - determining the filling degree along the center line of each hollow organ of the hollow organ structure by comparing the image absorption values ​​with the calculated absorption values, and

[0013] -Shows fill level information including fill level.

[0014] In addition, the present invention also proposes an automatic estimation of the filling degree of the hollow organ of the hollow organ structure, in particular the blood vessels of the vascular tree, with contrast agent at the time of recording the X-ray image, which automatic estimation can be implemented simply, in particular can be realized in real time, and for example provide feedback to the user. As a result, the user who has injected contrast agent before the X-ray image itself is recorded can obtain clear feedback: how well the hollow organ of the hollow organ structure is filled and how much its capacity in terms of contrast agent dosage is. The user can especially estimate: how close it is to the ideal use of contrast agent. In this way, the user no longer has to make his own estimation, which is increasingly difficult in view of the X-ray images shown, because a large number of recording techniques and image post-processing steps are used. In particular, image processing is currently less and less concerned with the display of the real absorption map, but rather with the interpretation of the X-ray image, that is, all relevant structures can already be shown very well in the first recording without and with post-processing steps.

[0015] In principle, it is also conceivable within the scope of the present invention that, as an alternative or in addition, the filling level information is also transmitted to other evaluation units and / or control units in order to derive other, for example statistical, information therefrom, to improve the control program of the automatic contrast agent dosing device and / or even to adjust the control of the automatic contrast agent dosing device in real time, as will be explained in further detail.

[0016] In summary, the present invention facilitates an optimized use of contrast agents and in doing so also reduces the risk of hollow organ damage.

[0017] The invention makes use of the fact that there are a large number of possible ways of partitioning hollow organ structures, such as vascular trees (also for determining centerlines, diameters, etc.), which can be carried out in part in real time and can also be used advantageously within the scope of the invention. Known partitioning rules that can also be used within the scope of the invention include, for example, threshold-based rules, "region growing" rules, methods using classification and / or support by anatomical atlases. In particular, for the steps of partitioning and determining centerlines and boundaries and diameters, it is also possible to at least partially use the functions of trained artificial intelligence, such as neural networks, in particular convolutional neural networks (CNN). This allows particularly rapid image processing and analysis.

[0018] Still it should be noted that, in addition to the X-ray images, additional information can also be introduced into the determination of the partition and, if necessary, the center line, the boundary and the diameter. For example, it is conceivable in principle that not only a single two-dimensional X-ray image is processed, but also other X-ray images of the X-ray image group are used auxiliaryly. In the design mode of this type, it is also feasible in principle that when there are X-ray images with different projection directions, the reconstructed, three-dimensional X-ray data group is processed, for example, a three-dimensional partition is performed, and its effect on the two-dimensional X-ray image can be easily inferred based on the known projection geometry. This means that the boundary, the center line and / or the diameter can be recalculated from the three-dimensional space of the X-ray image back to the two-dimensional space. In the case where hollow organs, especially blood vessels may overlap, other rules can also be considered, wherein, for example, a dual-plane X-ray machine simultaneously shoots X-ray images by different, especially mutually 90° shooting geometries, and thus the blood vessel direction in the vascular tree can be better derived. However, it is obviously also conceivable in principle and reliably implemented using modern methods, that is, partitioning is performed directly in the two-dimensional X-ray image. However, it is of course also possible to take into account additional information, for example segmentation results for previously recorded X-ray images, other previous recordings of the patient and / or anatomical atlases.

[0019] In general, it is expedient within the scope of the invention to assume that the hollow organ is circular. In particular, it can be provided that the diameter and / or the calculated absorption value are determined in the partitioned hollow organ structure assuming that the cross section of the corresponding hollow organ is circular. In this way, not only a simple calculation is possible, but also only one characteristic value, namely the diameter, needs to be maintained at each corresponding position along the center line.

[0020] In addition to determining the center line and the diameter, the results of the zoning are used in order to measure the absorption actually visible in the x-ray image by the contrast agent. For this purpose, an image absorption value is determined, which ultimately expresses how much stronger the x-ray absorption in the hollow organ is due to the contrast agent than in the adjacent, in particular directly adjacent, i.e., surroundings. A particularly easy-to-implement design is provided in this regard, namely that for determining the image absorption value, the ratio of the amount of the difference between the image value of the hollow organ (in the centerline region around the centerline containing at least one image point) and the image value of the surroundings (in a reference region outside the partitioned hollow organ structure containing at least one image point) is determined relative to the image value of the surroundings of the reference region. The zoning results are useful for this step, in particular information about the boundaries, since the information about the boundaries also contains information whether a certain image point is located inside or outside the hollow organ structure. The surrounding image value expresses the degree of absorption in the patient without the use of contrast agent, whereby the surrounding image value is used as a reference, but also as a basis, since, since the hollow organ does not fill the entire patient, this absorption degree also applies to the hollow organ and the x-ray radiation that thereby crosses the contrast agent. Accordingly, it is proposed that, in order to measure the proportion of additional absorption by the contrast agent in the hollow organ structure, a difference is formed between the image value in the center of the hollow organ, i.e. the hollow organ image value, and the image value outside the hollow organ, i.e. the ambient image value, which describes the additional absorption. A ratio of this difference to the ambient image value is determined, so that the image absorption value ultimately describes what proportion of the ambient absorption is additionally carried out in the hollow organ by the contrast agent.

[0021] The image absorption values ​​include image values, in particular grayscale values, outside and inside the hollow organ structure. That is, the image absorption value is determined based on at least one image value inside and outside the partitioned hollow organ structure. It is expedient to select a position along the center line with a known diameter so that the determination of the calculated absorption value can be carried out more simply at this position. The image value of the interior of the hollow organ is then particularly preferably acquired on the center line at this position, while the exterior of the hollow organ structure can be measured at different positions in principle, but preferably at least one image value is selected at the center line at this position in the immediate vicinity of the hollow organ.

[0022] In principle, it can be considered that as the centerline area, only the image points that include the centerline at the corresponding position are used, so as to directly obtain the corresponding image value, especially the grayscale value. Correspondingly, a single image point can also be used as a reference area outside the hollow organ, especially as close as possible to the position and thus forming the image point of the centerline area. However, it is also possible to use a centerline area and / or a reference area including multiple image points, especially if a statistical summary should be implemented. Therefore, it can be provided that the image value of the hollow organ when the centerline area includes multiple image points and / or the image value of the environment when the reference area includes multiple image points is determined by the statistics of the image value of the X-ray picture on the image points, especially the weighted average. Here, the suitability of using multiple image points is also particularly related to the spatial resolution of the X-ray picture. If multiple image points should be used, for example, nine new image points or sixteen image points can also be used. If necessary, abnormal value detection can also be considered in this case.

[0023] The reference region can be particularly advantageously located in the immediate vicinity of the hollow organ and / or in a section extending through the midline region, in particular in a section perpendicular to the midline. In a two-dimensional X-ray image, for example, a straight line perpendicular to the midline can be defined (as part of such a section), wherein from the midline along this straight line, in particular, a movement forwards by at least half a diameter is made until the surroundings of the hollow organ and hollow organ structure are reached according to the partitioning result. Subsequently, for example, at least the first and / or second image point located outside the hollow organ structure is used as part of the reference region, which can in particular also be done on both sides.

[0024] In view of the structure adopted for determining the image absorption value in this specific embodiment, the image absorption value at least substantially corresponds to the result of applying the Lambert-Beer law to the contrast amount actually transmitted. In other words, the ambient image value describes the basic radiation intensity that is further attenuated by the contrast agent and is specifically attenuated by the difference between the ambient image value and the hollow organ image value. Thus, the quotient at least substantially reflects what is expressed by the Lambert-Beer law, which constructs the quotient of the attenuated intensity and the incident intensity as an exponential function that is dependent on the amount of material transmitted.

[0025] Accordingly, a generally useful development of the invention provides that, for determining the calculated absorption value, the Lambert-Beer law is applied to the X-ray radiation of the X-ray spectrum, which passes through a diameter-corresponding distance of the contrast medium of the contrast medium composition. It is conceivable that the hollow organs, in particular blood vessels, are completely filled with contrast medium and that a 100% theoretical absorption is calculated for this filling degree, wherein the contrast medium used and the diameter of the hollow organ are calculated, in particular assuming a perfect cylindrical shape. This results in a simple calculation rule that allows the calculated absorption value to be determined very quickly.

[0026] If the image absorption value is determined as the quotient of the difference and the surrounding image value, as specifically proposed, and the calculated absorption value is determined by means of the Lambert-Beer law, then the two absorption values ​​can be directly compared with one another. In particular, the fill factor is then obtained, for example, by dividing the image absorption value by the calculated absorption value, thus by dividing the measured absorption by the expected absorption.

[0027] In general, by describing the absorbance in the contrast agent with an absorbance value, an X-ray spectrum that can be described by spectral information can also be considered. It can be provided that the filtering in the optical path is already taken into account in the spectral information. In principle, it can be considered within the scope of the Lambert-Beer law that the absorbance values ​​for different spectral information (for example, including the voltage of the X-ray tube (using the X-ray tube to generate the X-ray spectrum)) are provided as a functional relationship and / or a look-up table, and simplification can also be considered within the scope of the present invention, especially when the absorbance value changes only slightly or only slowly with the spectral information. Then, for example, it can be provided that, in order to consider the X-ray spectrum, absorbance values ​​for high-energy and low-energy X-ray spectra are provided respectively. For example, it can be provided that a limit value for the electron tube voltage is defined, above which a high-energy X-ray spectrum is generated, but below which a low-energy X-ray spectrum is generated. When two predefined spectra, such as a high-energy spectrum and a low-energy spectrum are used, this type of design is particularly suitable for dual-energy imaging.

[0028] Furthermore, in this case it can also be provided that the spectral information is modified to take into account the radiation hardening of the X-ray radiation by the patient, in particular depending on the relative position of the corresponding hollow organ in the patient. This takes into account the relationship with which the transmission spectrum of the patient is continuously shifted to higher energies, which is usually described as a radiation hardening effect. This radiation hardening effect can also be taken into account when the relative position of the hollow organ in the patient can be estimated and thus the extent to which the radiation hardening has changed the X-ray spectrum can also be taken into account. This makes it possible to further increase the accuracy of the filling degree information. This consideration of radiation hardening is a purposeful consideration of other imaging effects when estimating the filling degree. This can also be generally advantageous.

[0029] Therefore, within the scope of the present invention, it can be provided that at least another imaging effect, in particular beam hardening and / or scattered radiation effects, is taken into account in the determination of the calculated absorption value and / or the image absorption value. In principle, it is conceivable to apply known calibration measures to the X-ray images, but this is usually not implemented for two-dimensional X-ray images, in which absolute values ​​are usually not important with regard to attenuation. The present invention can then provide in this case that, at least for determining the fill level, a calibration, in particular a radiation hardening calibration and / or a scattering calibration, is still applied to the X-ray images. However, at least one corresponding calibration can also preferably be used in the determination of the calculated absorption value. For radiation hardening, corresponding rules have already been described with respect to the spectral information.

[0030] For the scattered radiation calibration, in a specific embodiment it can be particularly advantageous to provide that a calibration value in the form of a percentage is determined, which in particular describes the ratio of the scattered radiation to the total signal for the reference region, and that the portion of the calculated absorption value determined in particular using the Lambert-Beer law corresponding to the calibration value is subtracted from the calibration value. In this case, it can be utilized that the variables known in principle in the prior art (which describe the percentage of the scattered radiation to the total signal at the detector and are generally referred to as "S / T") can be used in a simple manner for the scattered radiation calibration. This is because when determining the image absorption value, the scattered radiation portion (for which a constant offset can be set) can fall into the difference but not into the ambient image value, so that a deviation occurs when the ratio is formed with the calculated absorption value (without taking into account the scattered radiation). However, it can be shown that this deviation is precisely eliminated when the calculated absorption value is reduced by its portion which corresponds to the ratio of the scattered radiation signal to the total signal (scattered radiation signal plus ambient image value). It is of course also conceivable in principle to apply this type of calibration in another way, for example in the form of a product with respect to the image absorption value.

[0031] It can be particularly advantageous if the filling level information is at least partially displayed on a display device, in particular a display device of an X-ray machine, when a manual injection of contrast medium is performed. In this way, the user gains knowledge of the filling level and can thereby learn about future contrast medium administrations, i.e., gain feedback on how well he is using contrast medium.

[0032] For example, it can be provided that the filling degree is displayed in an X-ray image and / or a hollow organ structure map derived from the partitioned hollow organ structure, in particular in a color-coded, position-resolved, superimposed manner. This type of hollow organ structure map that can be derived from the partitioning result is known in principle, for example as a vascular tree map or "vessel map". In particular, it can be provided that the filling degree is superimposed on this hollow organ structure map or the X-ray image itself, in particular in a color-coded manner. Here, in this case, it is particularly preferred that in the color-coded display, a warning color, in particular a red tone, is used when the filling degree is above a threshold value, in particular 95-100%. In this way, the user's line of sight can be directed directly to the area where there is a prominent or too intense contrast agent filling. This is particularly suitable for warning colors for filling degrees exceeding 100% (for example, when hollow organs, in particular blood vessels, even expand).

[0033] As an alternative or in addition, it can also be provided that the maximum value of the filling degree is displayed as a number. This number can also be displayed, for example, together with the display of the X-ray picture and / or the structure of the hollow organ, in particular at the edges or corners of the picture or the diagram and / or the corresponding display area. The maximum filling degree is usually particularly important information, which can be provided in addition particularly advantageously.

[0034] As mentioned above, however, the filling level information can also be used in other ways, for evaluation at an evaluation unit, for later visualization, for documentation, etc. This is particularly useful when using an automatic contrast medium dosing device, such as an automatic contrast medium injector. For example, an optimization of its control can then be achieved based on this filling level information.

[0035] However, it is also conceivable that the filling degree information is taken into account directly in the control and / or regulation. It can be provided that the filling degree information is sent as feedback information to the control unit of the automatic contrast agent dosing device for injecting contrast agent, where the feedback information is taken into account when controlling the contrast agent dosing device. If the contrast agent dosing that has been carried out so far is, for example, part of a contrast agent dosing process that is still ongoing, the filling pressure for the subsequent process of the contrast agent dosing process can be automatically increased in order to achieve a better filling degree if the filling degree is too low so far. At the same time, it is also possible to react quickly in the case of excessive filling pressure, for example, a filling degree of more than 100%, and reduce the filling pressure by controlling the contrast agent dosing device. In this case, it is particularly expedient to be able to provide the filling degree information in real time, which can be easily achieved in particular when using artificial intelligence and the simple calculation for absorption value and filling degree described here.

[0036] In addition to the method, the invention also relates to an evaluation device for evaluating two-dimensional X-ray images recorded by an X-ray machine using X-ray radiation from the X-ray spectrum, which X-ray images show contrast agents in hollow organ structures, in particular in vascular trees, in the recording region of a patient, the evaluation device comprising:

[0037] - Partitioning unit, used to partition the hollow organ structure in the X-ray image,

[0038] a measuring unit for determining the centerline and the boundaries of the partitioned hollow organ structure and detecting the respective diameter of each hollow organ of the partitioned hollow organ structure along the centerline,

[0039] a first detection unit for determining, along the center line of each hollow organ, in particular for each position along the center line for which a diameter has been determined, an image absorption value which describes the absorption enhancement due to the contrast agent in the hollow organ in comparison with the surroundings of the hollow organ in the X-ray image,

[0040] a second detection unit for detecting, for each image absorption value, a theoretical calculated absorption value that would be expected if the hollow organ were completely filled with contrast medium, using at least the diameter and contrast medium information describing the composition and / or absorption properties of the contrast medium,

[0041] a determination unit for determining the filling degree along the center line of each hollow organ of the hollow organ structure by comparison of the image absorption values ​​with the calculated absorption values, and

[0042] - An interface for outputting filling level information including the filling level.

[0043] In other words, the evaluation device is designed to implement the method according to the invention. The evaluation device can, for example, include at least one processor and at least one memory, wherein the functional units can be provided at least partially by the processor, in particular by hardware and / or software. Likewise, further functional units for implementing further steps of the method according to the invention can be provided accordingly. For example, a calibration unit can also be present in order to incorporate calibration of further imaging effects. The embodiments of the method according to the invention can be transferred analogously to the evaluation device according to the invention, with which the above-mentioned advantages can also be achieved.

[0044] The X-ray machine according to the invention has a control device with an evaluation device according to the invention. All the above descriptions also apply to the X-ray machine. The X-ray machine comprises at least one recording component with an X-ray emitter and an X-ray detector, wherein an X-ray image can be recorded by means of the recording component. In this respect, the control device can also have a recording unit in order to control the recording operation of the X-ray machine. For example, the X-ray image can be recorded as part of a set of X-ray images during the flow of a contrast agent in a hollow organ structure, wherein a specific frame rate can be set. This can be a fluorescence microscopic or radiographic examination.

[0045] The X-ray machine can be, in particular, an X-ray machine with a C-arm commonly used in angiography laboratories, on which an X-ray emitter and an X-ray detector are fixed opposite each other. The C-arm can be arranged at different positions in order to record a plurality of X-ray images with different projection geometries or recording geometries. The X-ray machine can also be a biplane X-ray machine with two C-arms, each of which has a recording component.

[0046] In a suitable development of the X-ray machine, it can be provided that the X-ray machine also has an automatic contrast agent dosing device together with a control unit for controlling it, wherein the control unit has an interface for receiving filling degree information and is designed to use this filling degree information when controlling the contrast agent dosing device, in particular for adjusting the injection pressure during the contrast agent dosing process, the filling degree information being linked to the injection pressure. The control unit can be part of the control device. As mentioned above, during the contrast agent dosing process, the filling pressure can be adjusted by correspondingly controlling the contrast agent dosing device with the help of the control unit on the basis of the filling degree information that is available quickly, in particular in real time. For example, such a contrast agent dosing process can last for several seconds.

[0047] The method can also be used for very different contrast media, but particularly preferably contrast media containing iodine and / or carbon dioxide come into consideration.

[0048] The computer program according to the invention can be directly loaded into the memory of the evaluation device and has a program medium so that when the computer program is executed on the evaluation device, the steps of the method according to the invention are implemented. The computer program can be stored on an electronically readable data carrier, which data carrier thus includes control information stored on the data carrier, which control information includes at least one computer program according to the invention and is designed to cause the evaluation device to perform the steps of the method according to the invention when the computer program is executed on the evaluation device when the data carrier is used in the evaluation device. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Further advantages and details of the invention are apparent from the exemplary embodiments described below and with reference to the accompanying drawings. In the drawings:

[0050] Figure 1 The flow chart of an embodiment of the method according to the present invention is shown.

[0051] Figure 2 schematically showing an X-ray image including annotations related to the method,

[0052] Figure 3 shows possible illustrations on a display device,

[0053] Figure 4 shows the functional structure of the evaluation device according to the present invention, and

[0054] Figure 5 A schematic diagram of the principle of an X-ray machine according to the present invention is shown. DETAILED DESCRIPTION

[0055] Figure 1A flow chart of an embodiment of the method according to the invention is shown, wherein a vascular system, in particular a vascular tree, is used as an example of a hollow organ structure, the hollow organ of the hollow organ structure being a blood vessel. Here, a contrast agent has been previously injected into the patient manually or using an automatic contrast agent dosing device, so that in step S1, an X-ray image of the imaging region of the patient including the hollow organ structure can be taken with an X-ray machine, wherein the hollow organ structure can be better displayed due to the contrast agent. The X-ray image can, for example, be part of a time series of X-ray images, which can, for example, be used for fluorescence microscopic monitoring of a process, for radiographic examinations and / or for reconstruction of a three-dimensional image data set under different recording geometries. The X-ray image is taken using an X-ray spectrum, which can, for example, be adjusted according to the electron tube voltage. Multi-energy applications are also possible, in particular dual-energy imaging, wherein, for example, high-energy and low-energy X-ray spectra are used alternately, or high-energy and low-energy X-ray spectra are used simultaneously in a dual-plane X-ray machine.

[0056] For the x-ray images recorded in step S1 , fill level information is to be determined by comparison with theoretically expected values ​​at a fill level of 100%. To this end, the absorption achieved by the contrast agent is first determined in a spatially resolved manner from the image values ​​of the x-ray images.

[0057] First, in step S2, a partitioning of the hollow organ structure in the X-ray image is performed. In this case, a partitioning algorithm known in principle from the prior art is used, in particular also such measurement information that is derived to a certain extent from the partitioning result, which is done in step S3 of the method. In this case, the center line of the hollow organ (here a blood vessel), its boundaries and diameter should be derived, and within the scope of the present invention, circular blood vessels, i.e. blood vessels with a circular cross section, are generally assumed. For the partitioning and / or for determining the center line, boundaries and diameter along the center line, artificial intelligence can be used at least partially, for example in the form of at least one trained function, such as a CNN.

[0058] It should be noted here that, in addition to the X-ray images, other information can obviously be used for the segmentation and / or detection of center lines, borders and diameters, such as segmentation results or other information from parts of other X-ray images that form a time series and / or from previous recordings and / or from reconstruction results using X-ray images, if necessary. Anatomical atlases can also be used.

[0059] For explanation, Figure 2 A purely schematic detail of an X-ray image 1 is shown, wherein a hollow organ structure 2, which is kept simple for the sake of illustration, is shown together with three hollow organs 3, 4, 5. In this case, during the segmentation, center lines 6 and boundaries 7 are first determined. From this, the corresponding diameter 8 can then be derived perpendicular to the center line 6.

[0060] Return to Figure 1 In step S4, the absorption is measured at different positions along the center line 6 in the X-ray image 1 in such a way that image absorption values ​​are determined. To determine the corresponding image absorption values, all points of known diameter 8 along the center line 6 are considered. Figure 2 At the position shown for the hollow organ 4, a straight line 9 is determined that is perpendicular to the center line 6. In addition, at least one pixel forms a center line region 10, in which the center line 6 is contained. In this case, the number of pixels (pixels) that form the center line region 10 is selected according to the position resolution of the X-ray image 1; for example, one pixel is used here. In addition, along the straight line 9 extending outwards to the intersection with the boundary 7, a reference region 11 is defined outside the hollow organ structure 2, that is, the hollow organs 3, 4, 5, especially closely adjacent to the corresponding hollow organ 4, which reference region also includes at least one pixel. For example, at least the first or second pixel along the straight line 9 can be selected, which is no longer located within the partitioned hollow organ structure 2, that is, the boundary 7.

[0061] When both the centerline region 10 and the reference region 11 consist of a single pixel, the image values ​​of the centerline region 10 directly form the corresponding hollow organ image values, and the image values ​​of the reference region 11 form the corresponding environment image values. When the regions 10 and 11 consist of multiple pixels, statistical averaging can be performed.

[0062] The image absorption value is determined by dividing the difference between the surrounding image value and the hollow organ image value by the surrounding image value. This describes very precisely the portion of the "base intensity" (surrounding image value) that is additionally removed due to the passage of contrast agent.

[0063] Subsequently, in step S5, a calculated absorption value is determined for each image absorption value. Here, the Lambert-Beer law is used for this purpose, wherein the absorption along a section with a length of diameter 8 (due to the assumption of circular hollow organs 3, 4, 5) that is completely filled with contrast agent is determined in a computational manner. Here, the contrast agent composition or the properties of the contrast agent determine the absorbance value to be used. Since the absorbance value can also depend on the energy of the X-ray radiation, spectral information describing the X-ray spectrum can also be used in addition. A distinction is made here between high-energy spectra and low-energy spectra, but a more precise breakdown or even the use of functional relationships can also be considered. If radiation hardening is to be taken into account and the relative position of the hollow organs 3, 4, 5 of particular interest in the patient is known, the spectral information can also be modified accordingly due to the radiation hardening expected from the transmission to the patient in order to take this imaging effect into account.

[0064] The calculated absorption value is likewise determined as an intensity ratio, which likewise indicates in percentage form how much the intensity is attenuated by the contrast medium when it is present in all hollow organs 3, 4, 5. That is to say, the calculated absorption value can be directly compared with the image absorption value, but before the comparison carried out in step S6, other important imaging effects, namely scattered radiation, are also taken into account in step S5 for determining the fill level.

[0065] In principle, it is conceivable to apply a scattered radiation calibration to the X-ray image 1 and thus not to modify the determination of the calculated absorption value, but preferably and in this exemplary embodiment, to subtract from it a proportion of the preceding value obtained by the Lambert-Beer law corresponding to the calibration value in percentage in order to finally determine the calculated absorption value. The calibration value corresponds here to the proportion of the scattered radiation in the overall signal for the ambient image value.

[0066] If this calibration is carried out, it is possible to obtain the fill level in step S6 by dividing the image absorption value by the calculated absorption value, since the quotient of the absorption values ​​describes how much contrast agent is actually present.

[0067] This will be explained with some specific examples. In the first embodiment, an X-ray image is taken using a 0.3mm copper filter when the tube voltage is 89kV. The patient has a thickness to be transmitted of 300mm and an SID of 1110mm (SID=source image distance). As hollow organs 3, 4, 5, the vessel diameter at the position of the observed blood vessel is 4.45mm. For the case of iodine-containing contrast agent and a filling degree of 100%, the expected absorption, that is, the previously calculated absorption value of 60.5%, is formed, but it is also necessary to consider that the radiation is calibrated with a calibration value of 26%, thereby obtaining the calculated absorption value of 44.5% to be used. However, an image absorption value of 42.5% is measured in the picture. The filling degree is thus 94%.

[0068] In a second exemplary embodiment, a 420 ml thick patient imaging region or focal region is recorded as an X-ray image at a tube voltage of 123 kV using a 0.2 mm copper filter at an SID of 1200 mm. The vessel diameter is 3 mm. Thus, at a fill level of 100% of contrast medium, an absorption of 32% is first calculated by the Lambert-Beer law. Given the calibration value for the scattered radiation of 45%, a calculated absorption value of 18% is obtained by subtracting approximately 14% as the calibration value. If the measured image absorption value is 14%, the result is a fill level of approximately 77%.

[0069] The filling degree is calculated in step S6 along all center lines like the diameter. The filling degree information thus formed is always sent to the interface in step S7. In this case, different options are conceivable in principle. It is then possible, for example, to display the filling degree information at least partially on a display device, in particular a display device of an X-ray machine, in step S8.

[0070] In this regard, Figure 3 By way of example, a display diagram 33 is shown in which the filling degree information of a hollow organ structure diagram 34 is superimposed and displayed in a color-coded manner. In this case, different shades in the area 12 represent different colors for different efficiency areas. It is expedient for the color coding to use a warning color, in particular red, in the case of a filling degree above a threshold value, the threshold value being, for example, 100%.

[0071] Furthermore, the maximum achievable filling level, which is 95% in this case, is indicated by the number 13 in the display image 10 .

[0072] The filling degree information can also be stored in a storage device for subsequent evaluation. However, when using an automatic contrast medium dosing device, it is also possible to perform the following steps according to step S9 (see Figure 1 ) transmits the filling level information to a control device of a contrast medium dosing device, for example a contrast medium injector, where it is taken into account in the control. In particular, it is also possible to infer the injected contrast medium from the X-ray images 1 during the contrast medium dosing process by determining the filling level information at least substantially in real time, in order to adjust the filling pressure as required and thus optimize the contrast medium usage during the contrast medium dosing process.

[0073] Figure 4 The functional structure of the evaluation device 14 according to the invention is shown. In addition to the memory 15, the evaluation device also mainly includes a partitioning unit 16 for implementing step S2. A measuring unit 17 is provided for determining the center line, the boundary and the diameter, which is performed according to step S3. The first detection unit 18 for detecting the image absorption value is constructed for implementing step S4, while the second detection unit 19 is designed for implementing step S5, in which the calculated absorption value is determined. According to step S6, the filling degree is determined along the center line 6 in the determination unit 20, and the filling degree information is determined therefrom, which can be provided via the interface 21 according to step S7. Other functional units can also be provided, such as a calibration unit 22 for performing radiation hardening calibration (correction of spectral information) and scattered radiation calibration (determination and use of signal components).

[0074] The evaluation device can be Figure 52 is a part of a control device 23 of an X-ray machine 24 schematically shown in FIG. The X-ray machine 24 comprises an X-ray radiator 25 and an X-ray detector 26 as recording components, which are arranged opposite each other on a C-arm 27. The C-arm can be placed in different positions relative to a patient placed on a bed 28. In order to control the recording work of the X-ray machine, in particular to record an X-ray picture in step S1, the control device also has a recording unit 29. A display device 30 is configured for the control device 23, for example, for displaying fill level information according to step S8.

[0075] Here, the X-ray machine 24 also includes an automatic contrast medium dosing device 31, here a contrast medium injector, which is controlled by a control unit 32, which is also located in the control device 23. The control unit is designed to carry out step S9 and can also take into account filling level information when controlling the contrast medium dosing device 31, in particular during a continuous contrast medium dosing process, in particular with regard to filling level optimization.

[0076] Although the details of the present invention have been illustrated and described in detail through preferred embodiments, the present invention is not limited to the disclosed examples, and those skilled in the art may derive other different variations therefrom without departing from the scope of protection of the present invention.

Claims

1. A computer-implemented method for evaluating a two-dimensional X-ray image (1) recorded by an X-ray machine (24) using X-ray radiation of the X-spectrum, the X-ray image showing a contrast agent in a hollow organ structure in the recording region of a patient, the method comprising the following steps: - partitioning of the hollow organ structure (2) in the X-ray image (1), - determining a centerline (6) and a boundary (7) of the partitioned hollow organ structure, and determining the respective diameter (8) of each hollow organ (3, 4, 5) of the partitioned hollow organ structure (2) along the centerline (6), - along the center line (6) of each hollow organ (3, 4, 5), determining an image absorption value which describes the absorption enhancement by comparison of the contrast agent in the hollow organ with the surroundings of the hollow organ in the X-ray image, - using at least the diameter (8) and contrast medium information describing the composition and / or absorption properties of the contrast medium, for each image absorption value a theoretical calculated absorption value that would be expected if the hollow organ (3, 4, 5) were completely filled with contrast medium, - determining the filling degree along the center line (6) of each hollow organ (3, 4, 5) of the hollow organ structure (2) by comparing the image absorption values ​​with the calculated absorption values, and - Outputs fill level information including fill level.

2. The method according to claim 1, characterized in that: The hollow organ structure is a vascular tree.

3. The method according to claim 1, characterized in that An image absorption value is determined along the center line (6) of each hollow organ (3, 4, 5) and for each position along the center line (6) of the detected diameter (8), which image absorption value describes the absorption enhancement by comparing the contrast agent in the hollow organ with the environment of the hollow organ in the X-ray image.

4. The method according to claim 1, characterized in that: The detection of the diameter (8) and / or the calculation of the absorption values ​​in the segmented hollow organ structures (2) is carried out assuming that the cross section of the corresponding hollow organ (3, 4, 5) is circular.

5. The method according to claim 1 or 2, characterized in that: To detect the image absorption value, the ratio of the difference between the hollow organ image value in a midline area (10) including at least one image point around the midline (6) and the ambient image value in a reference area (12) including at least one image point outside the partitioned hollow organ structure (2) to the ambient image value of the reference area (12) is detected.

6. The method according to claim 5, characterized in that If the centerline region (10) and / or the reference region (12) comprises a plurality of image points, the corresponding image value used in the ratio is determined by a statistical average of the image values ​​of the x-ray image (1) at said image points.

7. The method according to claim 6, characterized in that If the centerline region (10) and / or the reference region (12) comprises a plurality of image points, the corresponding image values ​​used in the ratio are determined by a statistically weighted average of the image values ​​of the x-ray image (1) at the image points.

8. The method according to claim 5, characterized in that The reference region (12) is located in the immediate vicinity of the hollow organ (3, 4, 5) and / or in a cross-section extending through the midline region (10).

9. The method according to claim 1, characterized in that: To determine the calculated absorption values, the Lambert-Beer law is applied to the X-ray radiation of the X-ray spectrum which has traveled a distance corresponding to the diameter (8) through the contrast medium of the contrast medium composition.

10. The method according to claim 1, characterized in that ,In the calculation of the absorption values ​​and / or the detection of the image absorption values, at least one further imaging effect is taken into account.

11. The method according to claim 10, characterized in that The at least one further imaging effect is a beam hardening and / or a scattered radiation effect.

12. The method according to claim 10, characterized in that For scattered radiation calibration, a calibration value is determined as a percentage and the part of the calculated absorption value determined using the Lambert-Beer law that corresponds to this calibration value is subtracted from the calibration value.

13. The method according to claim 12, characterized in that The calibration value describes the ratio of the scattered radiation to the overall signal for the reference region.

14. The method according to claim 1, characterized in that The filling level information is at least partially displayed on a display device (30).

15. The method according to claim 14, characterized in that The display device (30) is a display device of an X-ray machine.

16. The method according to claim 14, characterized in that The fill level is displayed in a color-coded, spatially resolved, superimposed manner on the X-ray image (1) and / or on the hollow organ structure image (11) determined from the segmented hollow organ structures (2).

17. An evaluation device (14) for carrying out the steps of the method according to any one of claims 1 to 16, the evaluation device for evaluating a two-dimensional X-ray image (1) recorded by an X-ray machine (24) using X-ray radiation of the X-ray spectrum, the X-ray image showing a contrast agent in a hollow organ structure in the recording area of ​​a patient, the evaluation device comprising: - a partitioning unit (16), used for partitioning the hollow organ structure (2) in the X-ray image (1), a measuring unit (17) for determining a center line (6) and a boundary (7) of the partitioned hollow organ structure and for determining a respective diameter (8) of each hollow organ (3, 4, 5) of the partitioned hollow organ structure (2) along the center line (6), a first detection unit (18) for determining, along the center line (6) of each hollow organ (3, 4, 5), an image absorption value which describes the absorption enhancement by the contrast agent in the hollow organ (3, 4, 5) compared to the surroundings of the hollow organ (3, 4, 5) in the X-ray image (1), a second detection unit (19) for determining, for each image absorption value, a theoretical calculated absorption value that would be expected if the hollow organ were completely filled with contrast medium, using at least the diameter (8) and contrast medium information describing the composition and / or absorption properties of the contrast medium, a determination unit (20) for determining the filling degree along the center line (6) of each hollow organ (3, 4, 5) of the hollow organ structure (2) by comparing the image absorption values ​​with the calculated absorption values, and - an interface (21) for outputting filling level information including the filling level.

18. Evaluation device (14) according to claim 17, characterized in that The hollow organ structure is a vascular tree.

19. Evaluation device (14) according to claim 17, characterized in that The first detection unit (18) is used to determine an image absorption value along the center line (6) of each hollow organ (3, 4, 5) and for each position along the center line (6) of the detected diameter (8), wherein the image absorption value indicates absorption enhancement by comparing the contrast agent in the hollow organ with the environment of the hollow organ in the X-ray image.

20. An X-ray machine (24) comprising a control device (23) having an evaluation device (14) according to claim 17.

21. The X-ray machine (24) according to claim 20, characterized in that The X-ray machine also has an automatic contrast medium dosing device (31) with a control unit (32) for controlling it, wherein the control unit (32) has an interface for receiving fill level information and is designed to use this fill level information when controlling the contrast medium dosing device (31).

22. The X-ray machine (24) according to claim 21, characterized in that The control unit (32) uses the filling degree information when controlling the contrast agent feeding device (31) to adjust the filling pressure during the contrast agent feeding process, and the filling degree information is associated with the filling pressure.

23. A computer program product comprising a computer program which, when executed on an evaluation device (14) according to claim 17, implements the steps of the method according to any one of claims 1 to 16.

24. An electronically readable data carrier having stored thereon a computer program according to a computer program product as claimed in claim 23.

Images