Method for ascertaining single scattered radiation image for x-ray image point to be observed of x-ray image

Abstract

The invention relates to a computer-implemented method for ascertaining a single scattered radiation image for an X-ray image point to be observed of an X-ray image, comprising the following steps: obtaining an X-ray image; providing an estimation algorithm; -ascertaining a single scattered radiation image for each X-ray image point to be observed by applying an estimation algorithm to the X-ray image.

Description

Technical Field

[0001] This invention relates to a computer-implemented method for obtaining a single-scatter radiation image of an X-ray image point to be observed in an X-ray image. Furthermore, this invention relates to a computer-implemented method for training at least one trained model, a data processing apparatus, a computer program, and a data carrier. Background Technology

[0002] In the field of X-ray imaging, scattered radiation has a significant impact on the image quality achievable during imaging. For example, in the field of computed tomography, the presence of scattered radiation can cause streak artifacts, blurring, or low-frequency distortion of image contrast.

[0003] To reduce the effects of scattered radiation, an anti-scattering grid can be placed in front of the X-ray detector in the imaging mode, thereby reducing the occurrence of scattered radiation artifacts. However, a portion of the primary radiation is also intercepted by the anti-scattering grid, which may cause degraded image quality for a given X-ray dose. Conversely, this may require a higher X-ray dose to achieve the desired image quality.

[0004] Therefore, in some applications, such as neurovascular imaging, for example when examining aneurysms or embolic strokes in the brain, the use of anti-scattering grids has generally been abandoned. Scattering artifacts in these cases can be reduced through software-based scattered radiation correction.

[0005] Software-based correction of scattered radiation can, in principle, be performed by simulating X-ray imaging, for example, by solving the Boltzmann transport equation or by the Monte Carlo method. However, if imaging is to be performed with low latency, i.e., at least in near real-time, such as for accompanying medical procedures, this correction is not suitable due to the high computational cost required in simulation.

[0006] The same problem also arises in so-called empirical methods, which iteratively optimize the projected image, and are therefore very computationally expensive, making them unsuitable for real-time use.

[0007] Some approaches use radiometric correction (RTC) via models trained through machine learning for image optimization to keep computational costs low. Such models can, for example, directly derive RTC from projected images. However, the results often suffer from insufficient image quality. This is particularly true when the model is trained only for a specific imaging design, such as for a particular imaging geometry, specific imaging parameters, and / or a specific patient group. Generalization is not readily feasible because image processing in these approaches is heavily influenced by the image dataset used for training. Summary of the Invention

[0008] Therefore, the objective of this invention is to provide an improved solution for enhancing the image quality of X-ray or CT recordings. In particular, the provided solution should also enable interventional imaging or imaging with reduced latency. Furthermore, the solution should be flexible enough for various imaging tasks.

[0009] According to the present invention, the objective is achieved by the method and apparatus according to the present invention. Preferred embodiments are described below.

[0010] According to one aspect of the present invention, a computer-implemented method is provided for obtaining a single-scatter radiation image of an X-ray image point to be observed in an X-ray image. The method includes at least the following steps:

[0011] The steps include: acquiring an X-ray image, wherein the X-ray image is recorded by means of an X-ray detector and includes multiple X-ray image points. Each X-ray image point is associated with a detector region of the X-ray detector. Multiple or all of the X-ray image points are the X-ray image points to be observed.

[0012] The steps include: providing an estimation algorithm that, based on an X-ray image, calculates a corresponding single-scatter radiation image associated with each X-ray image to be observed. The corresponding single-scatter radiation image has multiple scattered radiation images, each associated with a detector region of an X-ray detector. Preferably, the image values ​​of the scattered radiation images are only related to the intensity of the scattered radiation, which originates from primary radiation incident on the detector region associated with the corresponding X-ray image to be observed, and is scattered onto the detector region associated with the corresponding scattered radiation image.

[0013] The steps include: obtaining a single-scatter radiation image for each X-ray image point to be observed in the X-ray image by applying an estimation algorithm to the X-ray image.

[0014] X-ray images can be obtained, for example, by means of a recording system. X-ray images, whether two-dimensional or three-dimensional, can be based in particular on X-ray imaging or simulations of X-ray imaging. The recording system may include an X-ray source, an X-ray detector, and / or a collimator that forms primary X-ray radiation. The imaging device may be, for example, an X-ray imaging device in which recording is obtained by means of X-ray radiation. In particular, CT (computed tomography) equipment, such as a C-arm device, may be involved.

[0015] Within the scope of this specification, an X-ray image has multiple X-ray image points. Preferably, the image values ​​of the X-ray image points describe the X-ray intensity or dose received in the detector region associated with the respective image point. In one embodiment, the X-ray image may be an X-ray attenuation image or an X-ray attenuation-based image. In this case, the image values ​​of the X-ray image points describe attenuation of the primary radiation, i.e., the X-ray beam, before it reaches the detector region associated with the respective image point.

[0016] Whether used in conjunction with the description of X-ray images or in conjunction with the description of scattered radiation images below, the detector region can be a single detector pixel or comprise multiple detector pixels. Combining multiple detector pixels, such as 2×2 pixels, into an image point is called "merging." Merging offers advantages in terms of efficiency when processing X-ray images.

[0017] In one implementation, the X-ray image is a two-dimensional X-ray image. In this case, the X-ray image points can be designed as pixels. Alternatively, the X-ray image can be a three-dimensional X-ray image, where the X-ray image points can be designed as voxels. The same applies to the scattered radiation image points described below, and is entirely applicable to the image points of tomographic images, which can be obtained, for example, by reconstruction based on the X-ray image and / or the scattered radiation image. The X-ray image, single scattered radiation image, or total scattered radiation image, or generally a tomographic image, can be a complete record of the examined object or a part of a larger image.

[0018] Some, more, or all of the X-ray points in an X-ray image are classified as X-ray points to be observed. The X-ray points to be observed can be the entire X-ray image or a sub-region of the X-ray image. In one embodiment, the sub-region defined by the observed X-ray points is a connected sub-region. Alternatively, the X-ray points to be observed can define multiple spaced-apart sub-regions of the X-ray image.

[0019] The X-ray images to be observed can be selected, for example, by pre-selection rules or algorithms. For instance, only X-ray images with intensities below and / or above a preset threshold can be considered as X-ray images to be observed. This can lead to more efficient provision of single-scatter radiation images by discarding redundant computational operations that only slightly contribute to knowledge acquisition.

[0020] The inventive concept includes: X-ray images undergoing preprocessing via an estimation algorithm before processing. X-ray points in the X-ray image can be grouped or combined, such that the preprocessed X-ray points are a combination of the intensity or dose information initially contained in the X-ray points. For example, X-ray points can be combined in groups of 2×2 X-ray points for processing using an estimation algorithm, such that the image value of each X-ray point is obtained from the image values ​​of multiple initial X-ray points after preprocessing. In other words, an additional "merging" can be performed at the X-ray point level to further improve efficiency when processing X-ray images.

[0021] The proposed implementation scheme, which combines X-ray images and X-ray image points, can also be applied to adjusted X-ray images and corresponding X-ray image points of adjusted X-ray images.

[0022] The implementation scheme can also be applied to general medical images, modified medical images, and tomographic images.

[0023] The term "medical image" should be understood broadly and includes images recorded using imaging modalities for medical purposes. Specifically, the term "medical image" includes X-ray images, i.e., projection images, and tomographic images obtained by reconstructing projection images. Such tomographic images are, for example, obtained by reconstruction from X-ray / projection images during CT recording. Tomographic images can, for example, be oriented in a transverse section.

[0024] The intensity of X-ray radiation measured at a specific area of ​​an X-ray detector has two components: First, the intensity reflects primary radiation, i.e., radiation emitted by the X-ray source that does not interact with the object being inspected and thus strikes the specific detector area without deflection. Second, the measured intensity is caused by scattered radiation, i.e., radiation initially aligned with another area of ​​the X-ray detector but deflected along its path, causing it to strike the specific area of ​​the X-ray detector.

[0025] The primary radiation can be approximated as a straight X-ray beam or as multiple X-ray beams. The X-ray beam undergoes attenuation along its path from the X-ray source to the X-ray detector, where attenuation can be described by a total attenuation characteristic. The total attenuation characteristic can be a linear integral describing the attenuation of the X-ray beam. Alternatively, it can be a total attenuation coefficient. The X-ray beam can have an initial intensity. That is, the initial number of photons, and according to Beer-Lambert's law Attenuation, in which It is the total attenuation coefficient of the object being inspected at location x, which describes the total attenuation of the X-ray beam at location x.

[0026] The scattered radiation that hits a certain area of ​​the X-ray detector originates from the surrounding X-ray beams, whose primary radiation does not hit the exact area of ​​the X-ray detector, but rather hits the surrounding area.

[0027] Therefore, an X-ray beam aligned with a specific region of the X-ray detector causes an intensity in that specific region of the X-ray detector through primary radiation. The intensity in other surrounding regions of the X-ray detector can be measured by the scattered radiation. The region of the X-ray detector is associated with certain X-ray image points, such that the X-ray beam emits primary radiation to certain X-ray image points associated with the X-ray beam or to certain X-ray image points associated with the X-ray beam, and emits scattered radiation to other X-ray image points.

[0028] By applying an estimation algorithm to an X-ray image, the distribution of scattered radiation originating from the observed X-ray beam can be calculated, whereby a single scattered radiation image is obtained for each observed X-ray image point. In other words, by providing the X-ray image as input / input / input variable to the estimation algorithm, and by implementing the estimation algorithm, a single scattered radiation image belonging to each observed X-ray image point is generated. That is, if the X-ray image includes, for example, 100 × 100 X-ray image points, and all X-ray image points are the observed X-ray image points, then a total of 100 × 100, or 10,000, single scattered radiation images are generated. Each single scattered radiation image is associated with a specific observed X-ray image point.

[0029] Correspondingly, for a given X-ray image point to be observed, the single-scatter radiation image describes the distribution of scattered radiation originating from primary radiation incident on the detector region associated with the X-ray image point to be observed. In other words, for an X-ray beam aligned with the X-ray image point to be observed, the corresponding single-scatter radiation image describes how the scattered radiation originating from the X-ray beam is distributed onto the detector.

[0030] A single-scattered radiation image comprises multiple scattered radiation points, each corresponding to a detector region of an X-ray detector. In one embodiment, the number and arrangement of the scattered radiation points correspond to the number and arrangement of the X-ray points to be observed, or to the number and arrangement of the X-ray points themselves. Preferably, the dimension of the single-scattered radiation image corresponds to the dimension of the X-ray image, such as 2D or 3D. More preferably, the size of the single-scattered radiation image corresponds to the size of the X-ray image, such as the number of pixels or voxels. This simplifies the processing of the single-scattered radiation image, for example, when the single-scattered radiation image should be combined with or compared to an X-ray image.

[0031] The image values ​​of a scattered radiation image point are related to the intensity of the scattered radiation emitted from the primary radiation associated with the corresponding X-ray image point to the detector region associated with that image point. In other words, the image values ​​shown in a single scattered radiation image are related to the intensity of the scattered radiation originating from a single straight X-ray beam, i.e., from the X-ray beam associated with the X-ray image point to be observed. Therefore, a single scattered radiation image describes the distribution of scattered radiation for a single X-ray image point associated with it.

[0032] Here, the image value of the scattered radiation image point is preferably associated only with the intensity of the scattered radiation emitted from the primary radiation associated with the corresponding X-ray image point to the detector region associated with the corresponding scattered radiation image point. This specifically means that the image value of the scattered radiation image point in a single scattered radiation image is independent of, or not related to, other primary radiation or scattered radiation. Only scattered radiation originating from the primary radiation belonging to the corresponding X-ray image point, or from the X-ray beam belonging to the X-ray image point, has an impact on the image value of the scattered radiation image point in a single scattered radiation image.

[0033] The obtained single-scatter radiation image can be further used in various ways. In one embodiment of the invention, a single-scatter radiation image is provided, displayed, output, and / or transmitted. Particularly preferably, the X-ray image can be optimized based on the single-scatter radiation image.

[0034] By obtaining a single-scatter radiation image for each X-ray image point to be observed, the determination of scattered radiation contained in the X-ray image is simplified. The computational cost required for determining scattered radiation is reduced, while simultaneously improving the accuracy of the determination. The obtained estimate of scattered radiation can be used, for example, to optimize the X-ray image, particularly to correct imaging artifacts or improve image contrast. Here, obtaining a single-scatter radiation image is associated with less computational cost compared to obtaining a total estimate of scattered radiation.

[0035] As explained in more detail below, obtaining a single-scattered radiation image enables the use of scattered radiation to adjust image values ​​in medical images for additional assessment feasibility. In particular, image values ​​of X-ray image points can be adjusted, thereby enabling additional assessment feasibility for X-ray images. Specifically, the contrast of the X-ray image can be adjusted, or the material properties of the examined object can be determined more precisely. The same advantages can be achieved when the information contained in the single-scattered radiation image is used to adjust the visualization of tomographic images.

[0036] In a preferred embodiment, the method may include an additional step. This additional step comprises: determining a total absorption characteristic value, and in particular a total absorption coefficient, for at least one X-ray image point and based on a single-scatter radiation image. Here, the total absorption characteristic value may be a linear integral that describes the total absorption of primary radiation along the entire path from the X-ray source to the detector region associated with the corresponding X-ray image point in the X-ray detector. The total absorption coefficient describes the total absorption of primary radiation at a certain location on the object being examined. In other words, the total absorption coefficient can therefore be considered as a specific form of the total absorption characteristic value.

[0037] As already stated above, the total absorption characteristic value, especially the total absorption coefficient, is affected at least by the interaction between the primary radiation and the object under inspection, and by the scattering of the primary radiation. Various scattering effects can be considered here, such as classical scattering and / or Compton scattering, in particular.

[0038] Therefore, the determination of one or more total absorption characteristic values ​​and / or total absorption coefficients is particularly based on the obtained single-scatter radiation image. For this purpose, the information contained in the single-scatter radiation image can, for example, be reclassified such that the share of the intensity of the X-ray image point caused by scattered radiation is determined. The obtained share of scattered radiation incident on the X-ray image point can then be subtracted from the total intensity, so that the intensity corresponds to the portion of the primary radiation belonging to the X-ray image point. This "reclassification" of radiation shares can be performed, particularly in the projection region, when evaluating CT records.

[0039] The obtained intensity values ​​can be converted into absorption coefficients. Within the evaluation range of CT records, particularly in tomographic regions obtained through reconstruction based on projection areas, one or more absorption coefficients can be calculated. This can be done, for example, using Beer-Lambert's law as described above. An initial intensity can be calculated. The calculated intensity of the primary radiation can be divided by the initial intensity. The logarithm can be taken from the result to subsequently calculate the location-dependent total absorption coefficient through reconstruction.

[0040] In other words, for an observed image point in an X-ray image, the following share of the radiation received can be determined, which should be attributed to scattered radiation originating from the X-ray beam associated with other images in the X-ray image. Scattered radiation originating from primary radiation different from the primary radiation belonging to the X-ray image point can be ignored when calculating the intensity of the X-ray image point. In other words, the total absorption coefficient can describe the attenuation of the X-ray beam belonging to the observed X-ray image point, i.e., the attenuation of the primary radiation belonging to the observed X-ray image point, where the attenuation is caused by interaction with the object being examined and / or by scattering of the X-ray beam.

[0041] Optionally, the adjusted medical image, particularly the adjusted X-ray image or the adjusted projection image, can be obtained based on the calculated total absorption characteristic value, particularly based on the calculated total absorption coefficient. The intensity of the X-ray image points in the adjusted X-ray image can, for example, be related to the total absorption characteristic value. Alternatively or additionally, the intensity of the image points in the tomographic image can be related to the total absorption coefficient at the corresponding location of the image point. The intensity of the image points in the adjusted medical image can, in particular, be related only to the total absorption characteristic value belonging to the image point; that is, other technical information does not affect the intensity value of the X-ray image point. Alternatively or additionally, an X-ray attenuation image based on the total absorption characteristic value can be obtained.

[0042] The resulting adjusted medical image, based on total absorption characteristics, has altered contrast compared to the initial medical image. This altered contrast is particularly effective in reducing scattered radiation artifacts and other image errors caused by scattered radiation, such as blurring or excessively low contrast.

[0043] Furthermore, in a preferred embodiment, the method includes the step of obtaining a total scattered radiation image based on individual single-scattered radiation images of corresponding X-ray image points.

[0044] A total scattered radiation image has multiple image points, where the image values ​​of the image points correspond to the intensity of the scattered radiation that is incident on the detector region of the X-ray detector associated with the image point. Here, it is not particularly important which primary radiation source the scattered radiation originates from. Therefore, the total scattered radiation image describes how the scattered radiation is distributed overall in an X-ray image.

[0045] A total scattered radiation image can be generated, for example, by overlaying, i.e., superimposing single scattered radiation images. However, other methods for obtaining a total scattered radiation image are also conceivable, such as forming a weighted sum of single scattered radiation images.

[0046] Adjusted medical images can be obtained based on total scattered radiation images. The total scattered radiation image can be used, for example, when obtaining adjusted medical images so that the adjusted images are essentially free of scattered radiation. Adjusted medical images can be obtained, in particular, by forming the difference between the image values ​​of the initial medical image and the image values ​​of the total scattered radiation image. By considering total scattered radiation when obtaining adjusted medical images, scattered radiation artifacts can be effectively reduced, resulting in better image quality.

[0047] In a preferred embodiment, the method may include the steps of determining photoelectric effect absorption characteristic values, and in particular, the photoelectric effect absorption coefficient, for at least one X-ray image point based on a single-scatter radiation image. The photoelectric effect absorption characteristic value describes the absorption of primary radiation by photoelectric absorption along the path from the X-ray source to the detector region associated with the corresponding X-ray image point. When the photoelectric effect absorption characteristic value is a linear integral, it describes the photoelectric absorption along the entire path from the X-ray source to the X-ray detector. When the photoelectric effect absorption characteristic value is a photoelectric effect absorption coefficient, it describes the absorption of primary radiation by photoelectric absorption at a certain location on the object being examined. In other words, information about the intensity attenuation of the X-ray beam due to scattering effects contained in the single-scatter radiation image is recovered, and the "lost" intensity due to scattering is added to the intensity of the primary radiation. Thus, by "recovering" and eliminating the intensity attenuation caused by scattered radiation, only the attenuation of the X-ray beam due to photoelectric absorption is considered.

[0048] In particular, information about the scattering of the X-ray beam contained in a single-scatter radiation image can be redistributed such that the energy contained in the scattered radiation of the X-ray beam is added to the primary energy of the X-ray beam. In one embodiment, the sum of the intensities of the scattered radiation image points is formed, wherein this sum is added to the intensity of the primary radiation. This scheme requires less computational cost; however, simple computational operations cannot map all radiation effects. For example, Compton radiation or the attenuation of scattered radiation in the examined object can only be mapped to a limited extent.

[0049] Therefore, in another embodiment, a weighted sum of the intensities of the scattered radiation image points is additionally or alternatively formed. Preferably, the intensity of photons that strike the detector from X-ray image points further away from the observed image points is more strongly weighted, for example because the photons are partially absorbed after being scattered as they pass through the object being inspected, or because a portion of the photon energy is absorbed through (re)Compton scattering.

[0050] Determining the photoelectric effect absorption characteristic value, especially the photoelectric effect absorption coefficient, can be performed before, after, or simultaneously with determining the total absorption characteristic value. Similar to determining the total absorption characteristic value, when evaluating CT records, the linear integral describing the photoelectric effect absorption characteristic value over the projection area can be calculated. The photoelectric effect absorption coefficient at a specific location on the examined object, especially for a specific image point in the projection image, can be obtained through reconstruction based on information within the projection area.

[0051] Therefore, optionally, the adjusted medical image can be obtained based on a single-scatter radiation image or based on photoelectric effect absorption characteristic values. The image values ​​of the pixels in the adjusted medical image can be based on the intensity of the primary radiation incident on the corresponding detector region and the intensity of the scattered radiation originating from the primary radiation. Alternatively or additionally, the image values ​​of the medical image can be based on the photoelectric effect absorption coefficient. In particular, the image values ​​can be based solely on the photoelectric effect absorption coefficient. The intensity of the X-ray beam can be expressed as... ,in It is the initial intensity of the X-ray beam, and It is the photoelectric absorption coefficient at location x.

[0052] In other words, the X-ray beam associated with an X-ray image point attenuates not only through interaction with the object being examined but also through scattered radiation on its path to the detector. This attenuation of the X-ray beam due to scattered radiation is directly related to the information mapped onto the single-scattered radiation image of the associated image point. Therefore, the primary radiation of the X-ray beam deflected by scattering, belonging to an X-ray image point, can be relinked to the X-ray image point. The image values ​​and / or intensities of the image point can be adjusted in a modified medical image such that the image values ​​and / or intensities are correlated with the primary radiation associated with the image point and with the scattered radiation originating from the primary radiation associated with the image point.

[0053] The resulting adjusted medical images, particularly based on the difference between total absorption and scattered absorption characteristics, exhibit altered contrast relative to the initial X-ray images. This altered contrast is especially effective in enabling improved identification and differentiation of the examined object or portions of the object with high CT values, such as bone or contrast agents.

[0054] As mentioned above, the photoelectric absorption coefficient is determined based on the total attenuation and the attenuation through scattered radiation. This is achieved, in particular by subtracting the attenuation of the scattered radiation from the total attenuation or by adding the corresponding intensities.

[0055] Therefore, in a preferred embodiment, the method further includes determining the scattering absorption characteristic value, especially the scattering absorption coefficient, for primary radiation based on the obtained single-scattering radiation image, the obtained total absorption coefficient, and / or the obtained photoelectric effect absorption coefficient.

[0056] If the scattering absorption characteristic is constituted as a linear integral, then the linear integral describes which fraction of the X-ray radiation emitted from the X-ray source is scattered along the path from the X-ray source to the detector region associated with the image point, such that the scattered fraction does not reach the associated detector region. The scattering absorption coefficient of the observed image point, especially the scattering absorption coefficient of an image point in a tomographic image, describes which fraction of the X-ray radiation emitted from the X-ray source is scattered at the location of the image point, such that the scattered fraction does not reach the detector region associated with the corresponding X-ray beam. Therefore, in the based X-ray image, the fraction of X-ray radiation does not affect the image value of the X-ray image point, and in particular, the intensity of the X-ray image point no longer reflects the intensity of the deflected scattered radiation in the initial X-ray image.

[0057] The scattering absorption characteristic value can be determined for all X-ray images to be examined, i.e., for all X-ray images for which a single-scatter radiation image is obtained. However, this is not mandatory, allowing the scattering absorption coefficient to be determined for additional X-ray images or even for a smaller number of X-ray images. The same applies to the photoelectric absorption coefficient and the total absorption coefficient.

[0058] In the case of CT recordings, scattering absorption characteristic values ​​can be obtained either in the projected region or in the tomographic region after reconstruction. Scattering absorption characteristic values ​​can be calculated directly based on linear integrals.

[0059] For alternative sites, reconstruction can be performed first, followed by the use of the correlation described above to obtain the location-dependent scattering and absorption coefficients. Furthermore, the scattering absorption coefficient of the X-ray beam can be based on and To obtain, for example, the method is: obtain .

[0060] Scattered radiation that is scattered along the path to an X-ray image point and is not represented in the image value of the corresponding X-ray image point can be interpreted as misdirected primary radiation, which can allow for more accurate inferences about the characteristics of the object being examined.

[0061] Therefore, in another preferred embodiment, the method further includes the step of obtaining at least one adjusted medical image based on the obtained scattering absorption coefficient.

[0062] After estimating the scattered radiation and determining its absorption characteristics, particularly the absorption coefficient, an adjusted medical image can be derived. In the adjusted medical image, especially in the adjusted tomographic image, the scattering absorption coefficient of a pixel can affect the image value of that pixel. Therefore, the image information contained in the scattered radiation can be correlated with the pixel whose primary radiation is the starting point of the scattered radiation.

[0063] The intensity of pixels in an adjusted medical image can, for example, be related to scattered radiation caused by primary radiation belonging to the X-ray pixel. Specifically, the intensity of pixels in an adjusted medical image can be related only to the scattered radiation belonging to the pixel, meaning that other technical information does not affect the pixel intensity value. The intensity of pixels in an adjusted medical image can also be additionally or alternatively associated with the corresponding scattering absorption coefficient.

[0064] The resulting adjusted medical images have altered contrast compared to traditional medical images. This altered contrast is particularly effective in improving the identification or differentiation of materials with low CT values / CT numbers / Henness values, such as soft tissue materials.

[0065] Preferably, the method further includes the step of obtaining another adjusted medical image, wherein the adjusted medical image is obtained based on different absorption characteristic values ​​or absorption coefficients, and / or based on a single-scatter radiation image, such that the adjusted medical image displays different intensity values ​​and / or absorption values.

[0066] Therefore, multiple adjusted medical images containing different image information can be provided. Adjusted medical images can be obtained, for example, based on, and especially solely on, the scattering absorption coefficient. Another adjusted medical image can be obtained based on, and especially solely on, the photoelectric effect absorption coefficient. Thus, different viewing modes can be provided to medical professionals for observing the subject without requiring multiple X-ray or CT records of the same subject with different settings. When evaluating medical records, medical professionals can compare between different viewing modes, thereby simplifying the identification and / or classification of the subject.

[0067] The estimation algorithm can be designed in different ways. In one embodiment of the invention, the estimation algorithm includes scattering kernel superposition and / or Monte Carlo simulation.

[0068] In an additional or alternative implementation, the estimation algorithm includes a trained model or a trained model. The model is trained to obtain, based on an X-ray image with multiple X-ray image points, a single scattered radiation image associated with each X-ray image point to be observed, and for each X-ray image point to be observed, a single scattered radiation image with multiple scattered radiation points. Here, the image value of the scattered radiation point is associated with, preferably only with, or only with, the scattered radiation source that is incident on the detector region associated with the corresponding X-ray image point and scattered onto the detector region associated with the corresponding scattered radiation image point.

[0069] Typically, trained models mimic cognitive functions that connect human thought processes with those of others. Through training on training data, trained models are particularly adept at adapting to new situations and recognizing and extrapolating patterns. Another term used for "trained model" is "trained function."

[0070] Although trained models can learn complex relationships, even complex trained models can be applied at least in near real-time. Therefore, trained models can, for example, learn within the training scope to provide results that substantially correspond to the results of more complex computational methods used during training, such as results corresponding to solutions to the Boltzmann transport equations or Monte Carlo methods. Thus, estimates similar in quality to those achievable based on significantly more computationally expensive methods can be obtained using trained models. Therefore, near real-time estimation of scattered radiation can be achieved, particularly for X-ray image correction based on this, despite good estimation quality. Thus, the quality of scattered radiation correction in interventional imaging can be significantly improved, for example, by the manner described.

[0071] The estimation algorithm is physically inspired by obtaining individual single-scatter radiation images for different X-ray image points to be observed. Thus, intermediate results of the estimation algorithm, such as the obtained single-scatter radiation images, can be used within the training scope and / or directly during the application of the estimation algorithm to, for example, physically identify and discard erroneous results, or suppress the generation of such results within the training scope.

[0072] The trained model can be based on a machine learning model. Typically, the parameters of a machine learning model can be tuned through training to provide the trained model. Training can be performed, in particular, prior to the method according to the invention, and thus may not be part of the method according to the invention. Alternatively, however, it is also possible to perform training as an additional pre-process step within the method according to the invention.

[0073] Supervised training, semi-supervised training, unsupervised training, reinforcement learning, and / or active learning can be used in particular. Furthermore, representation learning, also known as "feature learning," can also be used. The parameters of the machine learning model can be iteratively tuned over multiple training steps. A specific cost function can be minimized, especially within the training scope. Backpropagation algorithms are particularly useful during training, for example, when training neural networks.

[0074] Machine learning models can, in particular, include neural networks, support vector machines, decision trees, and / or Bayesian networks and / or transformers, and / or may be based on k-means clustering, Q-learning, genetic algorithms, and / or association rules. Neural networks can, in particular, be deep neural networks, convolutional neural networks, or convolutional deep neural networks. Furthermore, neural networks can be adversarial networks, deep adversarial networks, and / or generative adversarial networks. Machine learning models can have U-Net architectures and / or vision-transformer architectures.

[0075] In a preferred embodiment, the method further includes the step of compressing the obtained single-scattered radiation image by means of an approximation of the image values ​​of the scattered radiation image points using a parametric function or parametric curve.

[0076] The parametric curves can be, for example, B-splines, polynomials, or other fundamental functions. Scattered radiation mapped through a single-scattered radiation image has a low frequency, resulting in only minor information loss when approximating with parametric functions. However, this also reduces the storage space required to store single-scattered radiation images and allows for more efficient design of the processing of these images.

[0077] When obtaining a single-scattering radiation image, only first-order scattering can be considered. This results in a small number of computational operations being performed during the estimation algorithm or when creating training data for the trained model. Therefore, the method requires less computational power in its implementation.

[0078] However, scattering of orders higher than first order can also be considered when obtaining a single-scattered radiation image. Similarly, scattering of orders higher than one can be considered when training the trained model. In this case, the underlying physical relationships are mapped more accurately. This results in an improvement in the accuracy of the resulting single-scattered radiation image.

[0079] Another aspect of the present invention relates to a computer-implemented method for training at least one trained model. The trained model is, for example, intended for use in an estimation algorithm within the computer-implemented method according to any one of the described embodiments.

[0080] A method for training a trained model includes the step of obtaining training data, wherein the training data comprises multiple training datasets of source data and desired results. The source data each comprises an X-ray image having multiple X-ray image points. The corresponding desired results, for each unobserved X-ray image point in the X-ray images, comprise a single-scattered radiation image associated with that single unobserved X-ray image point. The single-scattered radiation image should be provided by the trained model when processing the source data of the corresponding training dataset. Here, the image value of the scattered radiation image point is associated with scattered radiation originating from primary radiation incident on the detector region associated with the corresponding unobserved X-ray image point and scattered onto the detector region associated with the corresponding scattered radiation image point.

[0081] Furthermore, the method includes training the trained model through supervised learning based on the training dataset.

[0082] In addition, the method includes providing a trained model.

[0083] Therefore, the trained model can be based on supervised training with a training dataset that includes source data and multiple desired results. The source data can be used directly or after preprocessing as input data for the estimation algorithm. The corresponding desired results, especially single-scatter radiation images, should be provided by the estimation algorithm when processing the corresponding source data. Here, at least the desired results can be based on simulations of X-ray imaging, especially simulations of X-ray radiation scattering.

[0084] Simulation of X-ray imaging or simulation of X-ray radiation scattering can, in particular, precede the methods used to train trained models and / or to obtain images of single-scattered radiation. That is, such simulation of X-ray imaging can form an additional preliminary methodological step to the methods described herein.

[0085] X-ray imaging and / or the scattering of X-ray radiation can be simulated, for example, by solving the Boltzmann transport equation, the Monte Carlo method, and / or ray casting. Simulations can, in particular, utilize known parameters of the simulated X-ray imaging apparatus and a three-dimensional model of the object or patient to be mapped within the simulation's scope. The three-dimensional object can, in particular, describe the spatial distribution of different materials or the spatial distribution of the scattering and interaction properties of the mapped materials. In the simplest case, such a three-dimensional model can be generated synthetically, i.e., based, for example, on anatomical atlases and known scattering and interaction properties of different body parts, or the model or a portion thereof can be generated manually.

[0086] The source data for the corresponding training dataset, i.e., X-ray images, can also be generated at least partially within the scope of this simulation. Additionally or alternatively, for at least a portion of the training dataset, or for all of the training dataset, X-ray images based on actual X-ray imaging can be used as source data or the basis for source data. In this case, a three-dimensional model of the mapped object or patient can be generated based on the same X-ray imaging image data so that, as explained above, a single-scatter radiation image associated with the corresponding X-ray image can be obtained through simulation of X-ray imaging as the desired result.

[0087] The X-ray images of the corresponding training dataset can be detected, for example, as part of a computed tomography scan, whereby the X-ray image data describes a three-dimensional mapping of the object or patient. By classifying different image regions, for example by a trained algorithm or by using an anatomical atlas to which the three-dimensional mapping is registered, the different regions of the three-dimensional mapping can then be classified, for example, as tissues, bones, etc., whereby at least approximately known interaction and scattering properties of the mapped material can be associated with the different regions. For completeness, it should be noted that, in principle, the interaction and scattering properties of the three-dimensional object can also be estimated from the individual X-ray images, where additional information about the known geometry of the mapping and the object can be used in particular.

[0088] Furthermore, in a preferred embodiment, the corresponding desired result may also include a total scattered radiation image, which is obtained based on a single scattered radiation image associated with the corresponding X-ray image point. Including the total scattered radiation image in the training data in addition to the single scattered radiation image provided as the desired result improves the training of the trained model, as more information is provided for evaluating the training results. In particular, the trained model can provide more accurate and consistent results.

[0089] Preferably, training the trained model includes optimizing at least one cost function, wherein the value of the cost function is related to the image value of the scattered radiation image point and / or to the image value of the image point in the total scattered radiation image. Particularly preferably, training the trained model includes optimizing multiple cost functions, wherein the cost function is obtained and optimized for each X-ray image point to be observed, i.e., for each single scattered radiation image to be obtained. Additionally, another cost function may be obtained for the total scattered radiation image. The total scattered radiation image may be based on the single scattered radiation image, as described above, or the total scattered radiation image may be generated by the trained model independently of the single scattered radiation image. Thus, in one example, with n×m X-ray image points to be observed, n×m+1 cost functions are optimized within the scope of training the trained model.

[0090] Alternatively, optimization can be performed under additional conditions, evaluating the image values ​​of the scattered radiation image points and / or the image values ​​of the total scattered radiation image points.

[0091] Other features set forth in the context of the computer-implemented method according to the invention for obtaining a single-scatter radiation image for an X-ray image point to be observed can also be applied, along with the advantages mentioned therein, to the computer-implemented method for training at least one trained model, and vice versa.

[0092] Furthermore, the present invention relates to a data processing apparatus configured to perform a computer-implemented method according to any one of the embodiments described above. The data processing apparatus may, for example, be configured as a suitably programmed data processing device, or the mentioned functions may be implemented at least partially in hardware. The data processing apparatus may be integrated into a medical imaging device, particularly an X-ray device, or configured separately from it. The data processing apparatus may, for example, be implemented as a workstation computer, a server, or a cloud solution.

[0093] Furthermore, the present invention relates to an X-ray apparatus having an X-ray source and an X-ray detector for imaging, wherein the X-ray apparatus includes a data processing device according to the invention. By integrating the data processing device according to the invention and thereby implementing the method according to the invention within the X-ray apparatus, the estimating or correcting of the scattered radiation described herein can be performed directly within the range of data detection or when visualization is performed for a user via the X-ray apparatus.

[0094] Furthermore, the present invention relates to a computer program having instructions configured to, when implemented on a data processing device, execute a computer-implemented method according to the present invention.

[0095] Furthermore, the present invention relates to a computer-readable data carrier having instructions configured to implement a computer-implemented method when implemented via a data processing device.

[0096] Features discussed in the context of the method or device according to the invention can also be applied, along with the advantages mentioned, to other disclosed inventive subjects. Attached Figure Description

[0097] Other advantages and details of the invention will become apparent from the following embodiments and the accompanying drawings.

[0098] This is illustrated schematically:

[0099] Figure 1 An embodiment of an imaging system is shown, which includes an embodiment of a processing device.

[0100] Figure 2a A visual display showing the scattered radiation originating from an X-ray beam.

[0101] Figure 2b A visual display showing the scattered radiation originating from another X-ray beam.

[0102] Figure 3 Flowcharts are shown for a computer-implemented method for training a trained model, and flowcharts are shown for a computer-implemented method for obtaining a single-scatter radiation image.

[0103] Figure 4a A schematic diagram of an X-ray image based on the photoelectric effect absorption coefficient adjustment is shown.

[0104] Figure 4b A schematic diagram of an X-ray image adjusted based on the scattering absorption coefficient is shown.

[0105] Figure 4c A schematic diagram of an X-ray image adjusted based on the total absorption coefficient is shown. Detailed Implementation

[0106] The embodiments described in detail below are preferred embodiments of the present invention.

[0107] exist Figure 1 The diagram schematically illustrates an exemplary embodiment of an imaging system 1 according to an improved concept, said imaging system 1 being designed, for example, as an X-ray imaging system. Figure 1 The example illustrates the structure of an X-ray imaging system or X-ray imaging device based on the principle of a C-arm device with a rotatable and movable C-arm 6, which can be correspondingly twisted and moved to image the object 4 to be imaged or the object 4 to be examined from different directions, i.e., at different recording angles.

[0108] However, the imaging device 1 based on the improved concept can also be constructed according to other structures. In particular, the improved concept is not limited in principle to X-ray-based imaging methods.

[0109] In other words, Figure 1 The imaging device 1 includes, for example, an X-ray source 2 configured to generate X-ray radiation and emit it toward the object 4 being examined. An X-ray detector 3 of the imaging device 1 is disposed on the side of the object 4 opposite to the X-ray source 2. The X-ray detector 3 includes, for example, a detector array composed of photodiodes to detect X-ray quanta passing through the object 4. The detector 3 can then transmit the corresponding detector signal to the control or computing unit 5 of the imaging system 1 for further processing. An image processing device 9 can receive the recorded image from the control or computing unit 5. The image processing device 9 can perform possible corrections on the recorded image.

[0110] The imaging system 1 can be configured, for example, to perform rotational angiography methods, such as those based on the principle of subtraction angiography. In this case, the computing unit 5 can, for example, generate multiple two-dimensional projections (also referred to herein as initial images) recorded from different angles, and the computing unit 5 can calculate a three-dimensional reconstruction from them if necessary.

[0111] In other words, by examining the material of the object 4, attenuated X-ray radiation is emitted onto the X-ray detector 3, which includes multiple detector regions 12, wherein the intensity or dose of X-rays emitted into the respective detector regions 12 is detected in order to preset the corresponding image value of the corresponding X-ray image point of the X-ray image.

[0112] However, because the interaction between X-ray radiation and the material of the object under inspection 4 is not limited to pure absorption of X-ray radiation but also involves scattering, scattered radiation also occurs on the X-ray detector 3, which may interfere with imaging. Therefore, rather than supplementally or alternatively reducing the detected scattered radiation by other means, such as by using an anti-scattering grid, it is suitable to perform an estimation of the scattered radiation distribution in the X-ray image, in particular so that the X-ray image can be subsequently corrected based on the obtained estimate, for example by subtracting the estimate or by scaling or iteratively multiplying the detected X-ray intensity according to the estimate, and an adjusted X-ray image can be provided.

[0113] Therefore, a method for obtaining a single-scatter radiation image is implemented by a processing device 9, which in this example is integrated into the imaging system 1. However, in principle, it can also be configured separately from the imaging system 1, for example, as a workstation computer, server, or cloud solution. Exemplary designs of this method will also be referred to below. Figure 3 Explanation.

[0114] However, firstly according to Figure 2a and Figure 2b Briefly explain the principles upon which this invention is based. (As already combined...) Figure 1 As described, the imaging system 1 includes an X-ray source 2, through which X-ray radiation is emitted onto the object 4 to be examined. A detector 3, having multiple detector regions 12, is positioned behind the object 4 as observed from the X-ray source 2. The intensity of the X-ray radiation measured at a specific detector region 12.x comprises two main components: an intensity 14 caused by primary radiation PR and an intensity 16 caused by scattered radiation SR. That is, on the one hand, primary radiation PR is measured at a specific detector region 12.x, which strikes the detector 3 after passing through the object 4 (see [link to documentation]). Figure 2a ).

[0115] The primary radiation PR can be approximated as a straight X-ray beam emitted at the X-ray source with a certain initial intensity and attenuating along its path to detector 3. This attenuation is due, on the one hand, to the photoelectric absorption of energy by the object being inspected 4, and on the other hand, to the scattering of the X-ray beam, where a portion of the beam's energy is deflected from its straight trajectory by the scattered radiation. Figure 2a and Figure 2b Scattered radiation is shown by dashed lines. Therefore, in a specific detector region 12.x, the scattered radiation SR projected from other X-ray beams onto that specific detector region 12.x is additionally measured for the primary radiation PR (see [reference]). Figure 2b ).

[0116] Therefore, the scattered radiation SR originating from the primary radiation PR or the X-ray beam can be considered as deflected primary radiation PR, which contains information about the structure of the object being inspected, but does not strike the detector region 12 that originally belonged to the primary radiation PR. Correspondingly, it is worthwhile to determine the scattered radiation originating from the associated X-ray beam for each detector region 12 or for the corresponding X-ray image point in the X-ray image, so that the information contained in the scattered radiation distribution can be used to optimize or adjust the X-ray image.

[0117] The method already mentioned for obtaining a single-scatter radiation image is implemented by a computer program, which is stored, for example, in the memory of the image processing device 9 and implemented by a freely programmable processor of the image processing device.

[0118] Figure 3 Flowcharts are shown of an exemplary computer-implemented method for training a trained model and an exemplary computer-implemented method for obtaining a single-scattered radiation image.

[0119] An X-ray image is obtained in step S1. As described above, the X-ray image is recorded based on X-ray radiation incident on the X-ray detector 3. The X-ray image has multiple X-ray image points, each associated with a detector region 12 of the X-ray detector.

[0120] In step S2, an estimation algorithm is provided to obtain a single-scatter radiation image based on the acquired X-ray image. A separate single-scatter radiation image is generated for each X-ray image point to be observed in the X-ray image. Therefore, n×m×z single-scatter radiation images are generated in the presence of n×m×z X-ray image points to be observed. As described above, the single-scatter radiation image for the X-ray image points to be observed describes the distribution of scattered radiation SR, which originates from the X-ray beam and is struck by the primary radiation PR of the X-ray beam onto the X-ray image point to be observed.

[0121] A single-scattered radiation image has multiple scattered radiation image points, each associated with a detector region 12. The image value of each scattered radiation image point corresponds to the intensity of the scattered radiation SR emitted from the observed X-ray beam onto the corresponding detector region 12.

[0122] In step S3, the provided estimation algorithm is applied to the provided X-ray image to obtain multiple single-scatter radiation images.

[0123] Optionally, the total absorption coefficient µ can be calculated for at least one X-ray image point in step S4. tot Total absorption coefficient µ tot For example, the total absorption coefficient µ can be obtained directly from the scattered radiation image in the case of a CT recording, wherein optionally, reconstruction can be performed subsequently to obtain the total absorption coefficient µ. tot It can be correlated with image points in a tomographic image. Total absorption coefficient µ tot Specifically, it can describe the absorption experienced by the X-ray beam at specific locations along its path from X-ray source 2 to X-ray detector 3. Here, each image point in the projected image can be associated with a specific total absorption coefficient µ. totAdditionally or alternatively, the X-ray beam can be correlated with a total absorption characteristic value, especially an average total absorption characteristic value. To determine the total absorption coefficient µ... tot Based on a single-scatter radiation image, we can determine: the proportion of the intensity PR of the primary radiation of the X-ray beam to the total intensity of the X-ray image point, and the proportion of the total intensity of the X-ray image point caused by the scattered radiation SR originating from other X-ray beams.

[0124] Optionally, the photoelectric effect absorption coefficient µ can be calculated in step S6. photo Similar to the total absorption coefficient µ tot The photoelectric absorption coefficient µ can be directly obtained from a single-scatter radiation image. photo Optionally, reconstruction can then be performed to obtain the photoelectric absorption coefficient µ. photo It can be correlated with image points in a tomographic image. The photoelectric absorption coefficient µ photo Description: What is the attenuation of an X-ray beam when only the photoelectric absorption of the beam is considered, without regard to energy loss due to scattering effects? Photoelectric absorption coefficient µ photo The value is determined based on a single-scattering radiation image, and optionally based on the total absorption coefficient µ. tot This can be done by analyzing the information about the intensity and distribution of scattered radiation contained in the single-scatter radiation image. From the single-scatter radiation image, one can determine how much energy of the X-ray beam is deflected by the scattering effect.

[0125] Optionally, the scattering absorption coefficient µ can be obtained in step S7. Sc Scattering absorption coefficient µ Sc In particular, it can be based on the previously calculated total absorption coefficient µ tot and photoelectric effect absorption coefficient µ photo To determine the scattering absorption coefficient µ. Sc The value can be obtained, in particular, from the total absorption coefficient µ in the tomographic region, i.e., after the reconstruction step. tot Subtract the previously calculated photoelectric absorption coefficient µ photo To perform. Alternative or additional sites, the scattering absorption coefficient µ Sc However, it can also be determined based on a single-scattering radiation image. This is if the total absorption coefficient µ has already been calculated previously. tot and photoelectric effect absorption coefficient µ photo Then the scattering absorption coefficient µ Sc It can be easily achieved by forming the photoelectric effect absorption coefficient µ photo and total absorption coefficient µ tot The scattering absorption coefficient µ is obtained by comparing the differences between the two values. ScThe calculation can, in principle, be performed before or after reconstructing the tomographic image composed of X-ray images, with reconstruction preferably being performed first. In other words, the scattering absorption coefficient µ... Sc It can be calculated in the projection region, where tomographic reconstruction can then be performed, or reconstruction can be performed first, followed by calculation based on the total absorption coefficient µ. tot and photoelectric effect absorption coefficient µ photo To determine the scattered radiation coefficient µ Sc .

[0126] Optionally, in step S5, an adjusted medical image, particularly an adjusted X-ray image and / or an adjusted tomographic image, may be generated based on at least one of the obtained absorption coefficients. Preferably, in step S8, another adjusted medical image, particularly an adjusted X-ray image and / or an adjusted tomographic image, may be generated based on another of the obtained absorption coefficients, such that the adjusted medical images can be compared with each other.

[0127] The core of the method lies in the use of an estimation algorithm, which preferably includes or is configured as such a trained model. The trained model can be obtained, for example, by methods for training a trained model, as described in steps ST1, ST2, and ST3. These steps can be implemented as a standalone method or as part of a method for obtaining a single-scatter radiation image. However, steps ST1, ST2, and ST3 are optional for the method used to obtain a single-scatter radiation image.

[0128] Training data is obtained in step ST1. This preferably involves generating the training data. The training data includes source data and desired results, wherein preferably, instances of the source data and multiple desired results are used to form a training dataset. The source data is used as input to the trained model. The source data particularly includes X-ray images, which can be designed as previously described. The desired results, particularly for each image point to be observed in the X-ray images, include a single-scatter radiation image associated with the X-ray image point to be observed, which illustrates the distribution of scattered radiation emanating from the X-ray beam associated with the X-ray image point to be observed. Furthermore, the desired results preferably include a total scattered radiation image, which is obtained by superimposing the single-scatter radiation images.

[0129] Single-scattered radiation images and, optionally, total scattered radiation images, can be obtained, for example, by simulating the scattering of X-ray radiation in X-ray imaging. Here, a hypothetical object under inspection, whose material properties are known, can be used to determine how X-ray radiation is scattered through the object. This simulation of scattered radiation can be performed, for example, by applying the Monte Carlo method, where first-order scattered radiation and / or higher-order scattered radiation can be considered.

[0130] In this manner, for a given X-ray image (which in one embodiment may also be a simulated X-ray image), a simulated single-scatter radiation image can be generated for each X-ray image point to be observed, the single-scatter radiation image reflecting the distribution of scattered radiation originating from the X-ray beam associated with the X-ray image point to be observed.

[0131] Alternatively, the single-scattered radiation images used in the training dataset can be generated at least in part by applying slit scanning techniques or based on slit scanning techniques. This has the advantage that the single-scattered radiation images are based on real measurements and do not require examination of the object model.

[0132] In step ST2, the trained model is trained based on the training dataset using preferred supervised learning. Here, during the iteration process, X-ray images are presented as source data to the initially untrained model, which generates single-scattered radiation images and optionally total scattered radiation images based on the source data. A cost function is provided for each single-scattered radiation image and for the optionally provided total scattered radiation image, describing how similar the generated single-scattered or total scattered radiation image is to the expected result on the training dataset. The model weights are adjusted based on the cost function, where the goal is to minimize the cost function. Therefore, a backpropagation learning algorithm is preferably applied. Training ends, for example, after a certain number of learning iterations or when the value of the cost function falls below a certain threshold.

[0133] Then, in step ST3, the trained model is provided, particularly for use in methods for obtaining single-scattered radiation images, as previously described.

[0134] Figure 4a , Figure 4b and Figure 4cThe diagram shows schematically illustrated, adjusted medical images of the same subject 4, compared with adjusted tomographic images, wherein the three illustrated adjusted tomographic images are derived based on different absorption coefficients. The subject 4, schematically shown in the adjusted tomographic images, has different materials, particularly bone 20, tissue components 22, such as the brain, and blood 24. The materials of the subject 4 interact differently with the X-ray radiation irradiating it, thus different absorption coefficients are visible in the tomographic images. Different intensities are detected in different regions 12 of the X-ray detector 3. Different materials are identified by different types of shading lines. (As shown from...) Figures 4a to 4c As seen in the overview and comparisons, and as visualized only schematically through different types of shading, medical images possess varying contrasts. Contrast depends primarily on which absorption coefficient the illustration is based on.

[0135] exist Figure 4a The adjusted tomographic images shown are based on the photoelectric effect absorption coefficient µ. photo Generation. Based on the photoelectric effect absorption coefficient µ photo The obtained, adjusted tomographic images are particularly well-suited for distinguishing materials, such as bone 20 or contrast agents, because of the good contrast present in the display area.

[0136] exist Figure 4b The adjusted tomographic images shown are based on the scattering absorption coefficient µ. Sc This is generated. As can be seen from the attached figure, based on the scattering absorption coefficient µ... Sc When the adjusted tomographic images are generated, the contrast between different soft tissue materials is particularly strong, thus allowing for good differentiation of the materials.

[0137] exist Figure 4c The adjusted tomographic images shown are based on the total absorption coefficient µ. tot Therefore, this adjusted tomographic image is particularly useful because it eliminates scattered radiation artifacts, resulting in very few image errors.

[0138] The various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the disclosed embodiments can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this hardware and software interchangeability, the various illustrative components, blocks, modules, circuits, and steps are generally described above with reference to their functionality. Whether the functionality is implemented in hardware or software depends on the respective application applied to the entire system. Those skilled in the art can implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the disclosure.

[0139] The implementation as software can be implemented in software, firmware, middleware, microcode, hardware description languages, or any combination thereof. Code segments or machine-readable instructions can be methods, functions, subroutines, programs, routines, subroutines, modules, software packages, classes, or any combination of instructions, data structures, or program instructions. Code segments can be coupled to other code segments or hardware circuitry by transmitting and / or receiving information, data, arguments, parameters, or memory contents. Information, arguments, parameters, data, etc., can be transmitted, forwarded, or transferred, for example, through memory release, message forwarding, token forwarding, network transmission, etc.

[0140] The actual software code or dedicated control hardware used to implement the systems and methods described herein is not limiting to the claimed features or the disclosure. Therefore, the operation and performance of the systems and methods are described without reference to specific software code, and it is to be understood that the software and control hardware can be developed to implement the systems and methods based on the description herein.

[0141] When implemented as software, functionality, as one or more instructions or code, can be stored in a non-volatile computer-readable or processor-readable storage medium. The steps of the methods or algorithms disclosed herein can be embodied in a processor-executable software module, which may reside on a computer-readable or processor-readable storage medium. Non-volatile computer-readable or processor-readable storage media include not only computer storage media but also accessible storage media that facilitate the transfer of a computer program from one location to another. Non-volatile processor-readable storage media can be any available computer-accessible medium. Exemplarily, but not limitingly, such non-volatile processor-readable media may include RAM, ROM, EEPROM, CD-ROM or other optical data carriers, magnetic data carriers or other magnetic data storage devices, or other accessible storage media that can store desired program code in the form of instructions or data structures, and which are accessible to a computer or processor. As used herein, disks / optical discs include compact optical discs (CDs), laser optical discs, optical discs, digital versatile optical discs (DVDs), floppy disks, and Blu-ray discs, wherein disks typically reproduce data magnetically, and optical discs reproduce data optically by means of a laser. Combinations of the above media are also included in computer-readable media. Additionally, the operation of a method or algorithm may exist as one or any combination or arrangement of code and / or instructions on a non-volatile processor-readable medium and / or computer-readable medium that may be integrated into a computer program product.

[0142] The description above should enable those skilled in the art to make and / or use the embodiments and variations thereof described herein. Different modifications to the described embodiments will be readily apparent to those skilled in the art, and the principles defined herein can be applied to other implementations without departing from the spirit or scope of the subject matter disclosed herein. Therefore, this disclosure should not be limited to the embodiments shown herein, but should be viewed with the greatest possible scope based on the principles and novel features disclosed herein.

[0143] While disclosing different aspects and embodiments, other aspects and embodiments are also contemplated. The different aspects and embodiments disclosed are for illustrative purposes only and should not be considered limiting, wherein the true scope and spirit are defined by the specification.

Claims

1. A computer-implemented method for obtaining a single-scatter radiation image for an X-ray image point to be observed in an X-ray image, the method comprising the following steps: - Obtain an (S1) X-ray image, wherein the X-ray image is recorded by means of an X-ray detector (3) and includes a plurality of X-ray image points, each of which is associated with a detector region (12) of the X-ray detector (3). The X-ray image points mentioned therein are one or more X-ray image points to be observed; - Provides an estimation algorithm (S2) that, based on the X-ray image, obtains a corresponding single-scatter radiation image associated with each X-ray image point in the observed X-ray image points. The corresponding single-scatter radiation image has multiple scattered radiation image points, which are respectively associated with the detector region (12) of the X-ray detector (3). The image value of the scattered radiation image point of the single scattered radiation image is related to the intensity of the scattered radiation (SR), which originates from the primary radiation (PR) that is incident on the detector region (12) associated with the corresponding X-ray image point to be observed and is scattered onto the detector region (12) associated with the corresponding scattered radiation image point. - By applying the estimation algorithm to the X-ray image, a (S3) single-scatter radiation image is obtained for each X-ray image point to be observed in the X-ray image points to be observed.

2. The method according to claim 1, wherein the method further comprises the following steps: - For at least one X-ray image point and based on the single-scattered radiation image, obtain (S4) at least one total absorption characteristic value, especially the total absorption coefficient (µ). tot The total absorption characteristic value describes the total absorption of the primary radiation (PR) along the path from the X-ray source (2) to the detector region of the X-ray detector (3) associated with the corresponding X-ray image point.

3. The method according to claim 1 or 2, wherein the method further comprises the following steps: - Based on the obtained single-scatter radiation image and / or based on the obtained total absorption characteristic value, especially based on the obtained total absorption coefficient (µ). tot (S5) Obtain at least one adjusted medical image.

4. The method according to any one of the preceding claims, wherein the method further comprises the following steps: - Obtain the total scattered radiation image based on the individual scattered radiation images of the corresponding X-ray image points. Optionally, an adjusted medical image can be obtained based on the total scattered radiation image.

5. The method according to any one of the preceding claims, wherein the method further comprises: - For at least one X-ray image point and based on the single-scattered radiation image, determine (S6) at least one photoelectric effect absorption characteristic value, especially the photoelectric effect absorption coefficient (µ). photo The photoelectric effect absorption characteristic value describes the absorption of the primary radiation (PR) caused by photoelectric absorption along the path from the X-ray source (2) to the X-ray detector (3) in the detector region associated with the corresponding X-ray image point, and optionally, - Based on the obtained photoelectric effect absorption characteristic value, especially based on the obtained photoelectric effect absorption coefficient (µ). photo Obtain at least one adjusted X-ray image.

6. The method according to any one of the preceding claims, wherein the method further comprises: - Based on the obtained single-scatter radiation image, based on the obtained total absorption characteristic value, especially the total absorption coefficient (µ) tot ), and / or based on the obtained photoelectric effect absorption characteristic value, especially the photoelectric effect absorption coefficient (µ), photo To obtain the (S7) scattering absorption characteristic value, especially the scattering absorption coefficient (µ) of the primary radiation (PR), we need to determine the scattering absorption characteristic value (S7). Sc ), and optionally, - Based on the obtained scattering absorption characteristic values, especially based on the scattering absorption coefficient (µ) sc Obtain at least one adjusted medical image.

7. The method according to any one of claims 3 to 5, wherein the method further comprises the following steps: - Obtain (S8) another adjusted medical image, wherein the adjusted medical image is based on different absorption characteristic values, particularly based on different absorption coefficients (µ). tot µ photo µ sc (to obtain) 8. The method according to any one of the preceding claims, wherein the estimation algorithm comprises a trained model, wherein the model is trained to obtain, based on an X-ray image having multiple X-ray image points, a single scattered radiation image associated with a corresponding X-ray image point to be observed, wherein the image value of the scattered radiation image point is associated with the scattered radiation (SR), the scattered radiation (SR) originating from primary radiation (PR) incident on a detector region (12) associated with the corresponding X-ray image point to be observed, and scattered onto the detector region (12) associated with the corresponding scattered radiation image point.

9. The method according to any one of the preceding claims, wherein the method further comprises the following steps: - The obtained single-scatter radiation image is compressed by using the approximation of the image values ​​of the scattered radiation image points by a parametric function.

10. The method according to any one of the preceding claims, wherein when obtaining the single-scattering radiation image, only first-order scattering is considered, or scattering with a higher order than first-order is considered.

11. A computer-implemented method for training at least one trained model, said trained model being used in an estimation algorithm within the computer-implemented method according to any one of the preceding claims, wherein the method for training comprises the following steps: - Obtain (ST1) training data, wherein the training data includes multiple training datasets of source data and desired results, wherein the source data includes X-ray images with multiple X-ray image points, and The corresponding expected result for each X-ray image point to be observed in the X-ray image includes a single-scatter radiation image associated with the individual X-ray image point, wherein the single-scatter radiation image should be provided by the trained model when processing the source data of the corresponding training dataset. The image value of the scattered radiation image point in the single scattered radiation image is related to the scattered radiation (SR), which originates from the primary radiation (PR) incident on the detector region (12) associated with the corresponding X-ray image point to be observed, and is scattered onto the detector region (12) associated with the corresponding scattered radiation image point. - The trained model is trained (ST2) through supervised learning based on the training dataset. - Provide the trained model described in (ST3).

12. The method of claim 11, wherein the corresponding desired result further includes a total scattered radiation image, said total scattered radiation image being obtained based on the single scattered radiation image associated with the corresponding X-ray image point.

13. A data processing apparatus, wherein the data processing apparatus includes a mechanism for performing a computer-implemented method according to any one of claims 1 to 10.

14. A computer program having instructions configured to, when implemented by a data processing device, execute a computer-implemented method according to any one of claims 1 to 10.

15. A computer-readable data carrier having instructions configured to, when implemented by a data processing device, execute a computer-implemented method according to any one of claims 1 to 10.

Images