Determining soft tissue-related properties from X-ray imaging data

Abstract

An apparatus 100 for determining soft tissue-related properties of a bone constellation from X-ray imaging data is provided. The apparatus 100 has a data processor 110 configured to acquire X-ray imaging data of a region of interest including a bone constellation, and to determine a pose estimate of the bone constellation and at least one geometric parameter d, t, r defined between bones of the bone constellation by utilizing a 3D pose estimation computer model CM into which the X-ray imaging data is fed, to determine the soft tissue-related properties of the bone constellation.

Description

[Technical Field]

[0001] The present invention relates to medical imaging, and in particular to an apparatus and computer-implemented method, and computer program elements, for determining soft tissue-related properties of a bone constellation from X-ray imaging data. [Background technology]

[0002] In medical imaging, automatic, i.e., computer-aided, fracture detection is widely used because bones are clearly visible in, for example, musculoskeletal X-ray images. Generally, X-ray technology only allows for clear visualization of high-density objects such as bones. However, when lower-density objects are examined by imaging, the data required for them are not explicitly visible in the X-ray imaging data, and therefore, non-bone-related or soft tissue injuries or abnormalities cannot be directly observed or diagnosed therefrom. Rather, they can only be observed or diagnosed indirectly, and therefore only by experienced physicians, which does not allow for computer automation.

[0003] Therefore, more time-consuming or costly image acquisition techniques, such as magnetic resonance imaging (MRT), are generally indicated for the purpose of observing or diagnosing soft tissue-related injuries or abnormalities, which, however, require appropriate equipment, skilled personnel, and patient patience. Summary of the Invention [Problem to be solved by the invention]

[0004] Therefore, a means for automated analysis of less dense objects in X-ray images may be needed. [Means for solving the problem]

[0005] The object of the present invention is solved by the subject matter of the independent claims, further embodiments are incorporated in the dependent claims.

[0006] According to a first aspect, an apparatus for determining soft tissue-related properties of a bone constellation from X-ray imaging data is provided, the apparatus having at least one data processor configured to: (i) acquire X-ray imaging data of a region of interest including a bone constellation; and (ii) determine a pose estimate of the bone constellation and at least one geometric parameter defined between bones of the bone constellation by utilizing a 3D pose estimation computer model fed with the X-ray imaging data, thereby determining the soft tissue-related properties of the bone constellation.

[0007] This allows for automated analysis of less dense objects in the X-ray imaging data, i.e., based on the X-ray imaging data alone. In particular, this allows the X-ray imaging data to be analyzed using a processor, i.e., a computer, for bone constellation-related properties, particularly at least one soft tissue-related property, that are not directly visible in the X-ray image. This automated analysis provides significant cost savings, since an MRI is not required. Furthermore, the automated analysis requires less well-trained personnel, and the results of the analysis can be further processed computationally, automatically documented, etc.

[0008] The inventors have discovered that by considering or encoding, in a non-trivial way, at least one additional geometric parameter defined between the bones of the bone constellation, a significant source of information is available for performing soft tissue related analysis and / or diagnosis that goes beyond purely bone related diagnosis, such as fracture detection, from X-ray imaging data alone.

[0009] For example, the 3D pose estimation computer model may be referred to as an augmented and / or parameterized 3D joint model, where the 3D pose estimation computer model is configured to estimate pose parameters of a bone constellation, which may be a joint, e.g., an ankle joint. The 3D pose estimation computer model is parameterized using additional geometric parameters described herein. The term pose, as used herein, may refer to projection geometry (i.e., pose in the narrow sense, including gaze direction and detector position), joint pose (i.e., flexion parameters), etc. A basic 3D pose estimation computer model may have two modules or stages: in the first stage, pose-discriminating features (bone silhouettes plus some inner contours) are detected by a convolutional neural network (CNN) as binary segmentation masks, and in the second stage, another CNN regresses all pose parameters from these feature masks. An advantage of this approach is that both components can be trained independently of each other, so that pose ground truth labels are not required for X-ray images. An exemplary 3D pose estimation model, without limiting the described aspects, that can be extended with additional geometric parameters as described herein, is described in Kronke et al. “CNN-based pose estimation for assessing quality of ankle-joint X-ray images”, SPIE Medical Imaging 2022, which is incorporated herein by reference in its entirety.

[0010] For example, feature detection for a 3D pose estimation model can be implemented as follows: As a preprocessing step for the feature detection network, a region of interest (ROI) can be automatically extracted from a given input X-ray image. This ROI cropping approach can be based on a foveated fully convolutional network (FNet) that predicts bone contours on downsampled X-ray images and a probabilistic anatomical atlas of these bone contours. The detected bone contours can be registered to the atlas using a similarity transform with additional flipping degrees of freedom to compensate for left / right lateralization. The resulting transformation can be used to map anatomically defined ROIs of joints from the atlas to the input image and thereby resample the determined ROIs. A feature detection network can be trained on this resampled ROI with high spatial resolution to sufficiently resolve the spatial constellation of contours that predicts 3D pose. For example, for a lateral ankle (LAT) X-ray image, an FNet can be trained to segment the lateral silhouette of the fibula and selected salient contours of the tibia, talus, and calcaneus. A contour-wise softmax layer may be attached to the FNet architecture, whereby the network is configured to predict heatmaps for each target contour and its complement, which are condensed into a binary prediction by applying a pixel-wise argmax operation to each of these heatmap pairs, allowing for the spatial coexistence of different contour classes. Training data may be generated by manually or automatically annotating structures in multiple X-ray images, e.g., multiple, tens, hundreds, etc., and discretized by extending the contour annotations on the pixels of the ROI. The feature detection task may thereby be formulated as a standard segmentation problem.

[0011] For example, pose estimation for a 3D pose estimation model can be implemented as follows: To train a pose estimation network without requiring X-ray images with pose ground truth, an articulated 3D model of a bone constellation, e.g., an ankle joint, can be constructed from a series of MR images acquired at various flexion angles. Semi-manual segmentation of the tibia, fibula, talus, and calcaneus in one MR image allowed for the creation of a surface model represented using a triangular mesh. For each triangle of the mesh, features for model adaptation can be trained using image intensities, and the resulting model was used to delineate the corresponding bone surface in other MR images in the sequence. Because all images were created using the same volunteer, a rigid transformation for each bone can be applied for model-based segmentation. A flexion angle δ can be assigned to each MR image by measuring the angle between the major axis of the tibia and the major axis of the calcaneus. Smooth interpolation of the surface model with respect to δ allows the mesh constellation to be estimated at any flexion angle between approximately 80° and approximately 135°. To generate synthetic projection data from this articulated 3D model, several pose parameters are defined. The central beam (or gaze direction) is determined by the angle θ relative to the head-feet axis in the transverse plane and the angle φ relative to the medial-lateral axis. To simplify the sampling of pose parameters for training the pose estimation network, the detector and source can be assumed to rotate in a fixed C-arm-like configuration around the joint center, meaning that φ and θ uniquely determine the source and detector positions. Here, the source-detector distance and joint-detector distance can be selected to be within a realistic regime for clinical lateral ankle examination and to roughly match the ROI in which the feature detection network operates. The pose parameters may include detector in-plane rotation by angle γ, translation by 2D vector t, scaling by factor s, and flexion angle δ.This basic 3D pose estimation computer model is configured to determine acquisition parameters, such as the geometric relationships between the source, joint model, and detector, as well as some free parameters of the joint itself, such as scaling parameters and joint flexion angles. This can be used, for example, to analyze and / or evaluate whether projection images of the joint have been acquired in a desired pose and, therefore, whether the image quality is sufficient to enable a desired diagnosis. This basic 3D pose estimation computer model is not currently used for medical diagnosis of images.

[0012] It should be noted that in the basic 3D pose estimation computer model described above, bone constellations, e.g., bones of a joint, are constrained to have fixed constellations relative to each other, which limits the ability of the 3D pose estimation computer model to be fitted to a given x-ray image, e.g., due to pathologically induced variations in bone constellations.

[0013] Here, we propose extending the above (basic) 3D pose estimation computer model with additional free geometric parameters between bones and matching them to determine soft-tissue-related or related geometric parameters from radiographic data alone, i.e., from X-ray images. In addition to interbone angles, which may reflect typical joint movements such as flexion, abduction, and rotation found in the acquired pose, this disclosure focuses on at least one interbone geometric parameter in 3D described herein. The additional geometric parameters are a valuable source of information in various diagnostic tasks, not only for image quality assessment but also for medical diagnosis of images, particularly. In contrast to attempting to perform dislocation detection solely on 2D images, this method is limited to deviations (dislocations) in the image plane alone and is affected by alignment errors whenever they exist. The devices, systems, and methods described herein, based on 3D pose estimation, can avoid both limitations.

[0014] In addition to at least one geometric parameter, other relevant parameters may also be combined, for example in a linear fashion or in a logistic regression.

[0015] The apparatus may be implemented in software executed by a data processor, in hardware, or a combination of hardware and software. The 3D pose estimation model may be stored as a program element in a computer memory and executed by a data processor, or may be implemented in a chip.

[0016] As used herein, a bone constellation may be any arrangement of bones in a human or animal body, such as the torso, skull, spine, arm, hand, leg, foot, etc. In particular, the bone constellation may form a body seam or joint, such as a knee, elbow, hand, shoulder, etc., that is mapped in a 3D pose estimation model.

[0017] Furthermore, as used herein, soft tissue may be understood as distinct from bone constellations or bones, as bodily objects that have a lower density than bones due to the property that soft tissues are invisible, poorly visible, or barely visible in X-ray imaging data. In relation to bone constellations, soft tissues may include, for example, cartilage, ligaments, tendons, muscles, fascia, etc. Soft tissues may extend or be located between bones, adjacent to bones, or at a (small) distance around bones.

[0018] It should be noted that the device may further comprise a suitable data interface configured to receive X-ray imaging data. The imaging data may be provided by a Picture Archiving and Communication System (PACS), an X-ray detector, etc. Furthermore, the device may be configured to generate a report regarding the soft tissue-related properties upon detection.

[0019] Optionally, the device may further be configured to trigger an output of at least the soft tissue-related properties. Output may be understood as any type of making available for further processing, such as data output for further processing by a data processor or computer, output to a user or simply output of properties for information, etc. This may allow the determined soft tissue-related properties to be used in, preferably automatic, medical diagnosis.

[0020] Optionally, the apparatus may be further configured to describe and / or visualize the soft-tissue-related properties in or based on the 3D pose estimation computer model. For example, the determined soft-tissue-related properties may be graphically illustrated and / or described in a graphical representation of the 3D pose estimation computer model, may be graphically highlighted therein, may be labeled in text form, etc.

[0021] Optionally, the device may be further configured to determine, particularly automatically, soft-tissue-related pathologies and / or abnormalities from at least one soft-tissue-related characteristic. This may include any type of dislocation within or of a bone constellation that may involve damaged soft tissue. For example, a dislocated (too far, too narrow) bone may indicate, for example, a torn ligament from trauma, weakened ligaments (osteoarthritis risk), degenerated cartilage (osteoarthritis), other bones fractured (trauma, osteoporosis), other muscles or tendons ruptured (trauma, overload), congenital anomalies, or postural effects (errors). Furthermore, injuries may be more difficult to detect from, for example, an X-ray image than a dislocated foot. In trauma cases, such as motorbike or motorcycle accidents, hand injuries are often overlooked as the focus shifts to other life-threatening injuries to the abdomen and head. The devices, systems, and methods described herein may be used to flag suspicious cases for early treatment once the patient has otherwise been stabilized.

[0022] Optionally, the device may be further configured to compare the geometric parameters and / or posture estimates with corresponding thresholds to determine soft-tissue-related pathologies and / or abnormalities. For example, the geometric parameters may be distances, including the relative distance between two bones, whose normal values ​​may lie between a lower threshold and an upper threshold. Consequently, trauma, abnormalities, etc., may be determined by comparing the geometric parameters with this range of normal values. Also, by comparing with the corresponding thresholds, it may be determined whether posture was correct during image acquisition, thus avoiding misdiagnosis. For example, deviations from normal parameters or thresholds may serve as additional diagnostic information in routine radiology services, even in asymptomatic cases. They may indicate, for example, congenital anomalies or provide prognostic parameters, such as a risk score for developing osteoarthritis. A database of typical or normal bone distances may be provided and used for diagnosis, with each individual patient assessed for statistical values.

[0023] Optionally, the device may be further configured to adapt the respective thresholds in the case of implants present within or associated with the bone constellation. For example, following trauma, bones may be surgically fixed using implants and other devices, which strongly constrains the possible parameter variations and may introduce parameter settings that are not typical for normal anatomical structures (e.g., outside the learned range). Such situations may require additional constraints and / or conditions to be introduced in an ad hoc / situation-specific manner. In this regard, the device may be configured to obtain information about implants or other bone fixation devices, for example, from data entered by a user, from patient data, by feature detection in X-ray imaging data, etc., and may take into account the effect of the implant on soft tissue. For example, to eliminate misdiagnosis, relatively small distances between bones caused by implants or bone fusion may be given correction values.

[0024] Optionally, the device may be further configured to combine the at least one geometric parameter, the at least one geometric parameter, and / or the soft tissue-related characteristics with patient-related data to determine soft tissue-related lesions and / or abnormalities. The patient-related data may, for example, include the patient's age, since some findings may be more or less likely depending on age, and may include the patient's body mass index (BMI), since this may affect the distance between bones, etc. The patient's pre-existing medical conditions, etc. may also be taken into account as additional parameters. This may make the automatic diagnosis more accurate.

[0025] Optionally, the device may be further configured to trigger an alert to be output based on the determined soft-tissue-related lesions and / or abnormalities. For example, an image reader with limited expertise in MSK applications may be alerted to an "incidental finding." Often, the reader's or diagnostic attention is focused on a major structural lesion, such as a large fracture or other bony trauma, which can lead to more subtle (e.g., soft tissue) lesions or abnormalities being overlooked. Here, automated alerts may be provided, and additional (secondary) findings may be automatically fed into the reporting scheme.

[0026] Optionally, the 3D pose estimation computer model may be parameterized by parameters including one or more of the flexion angle of the bone constellation, the line of sight direction of the bone constellation, the X-ray detector position relative to the bone constellation, the translation in the X-ray detector plane, the rotation in the X-ray detector plane, the scaling in the X-ray detector plane, and geometric parameters, and the 3D pose estimation computer model is trained to regress the one or more parameters.

[0027] Optionally, the at least one geometric parameter may include one or more of interbone distance, orthogonal translation, and rotation. For example, interbone distances, such as the talus-tibia distance, encoded as additional free parameters in the 3D pose estimation computer model are an important source of information for patient diagnosis. The interbone distances may reflect the size of a joint space, such as the superior ankle joint space. Low distance values ​​may thereby indicate cartilage degeneration, for example, due to age or osteoarthritis, while abnormally high numbers may indicate ligament tears and foot dislocations.

[0028] Optionally, the device may be further configured to determine an axis between two adjacent bones of the bone constellation and measure the distance between the two adjacent bones along the axis. For example, inter-bone distances, such as the talus-tibia distance, encoded as additional free parameters in the 3D pose estimation computer model are an important source of information for patient diagnosis.

[0029] Optionally, the at least one geometric parameter may include distances between bones, and a tree or graph is used to encode several distances between bones in the 3D pose estimation computer model, where the bones are vertices and the distances are modeled by edges.

[0030] Optionally, at least one geometric parameter may be a variable parameter of the 3D pose estimation computer model. As described above, in the basic 3D pose estimation computer model, the bones of the bone constellation are constrained to have a fixed constellation relative to each other. According to this example, the geometric parameter is a freely configurable parameter that can be varied. This may indicate soft tissue-related properties.

[0031] According to a second aspect, there is provided a computer-implemented method for determining soft tissue-related properties of a bone constellation from X-ray imaging data, the method comprising the steps of obtaining X-ray imaging data of a region of interest including a bone constellation, and determining a pose estimate of the bone constellation and relative distances between bones of the bone constellation by utilizing a 3D pose estimation computer model fed with the X-ray imaging data, the relative distances being assigned to at least one soft tissue-related property of the bone constellation.

[0032] The method may preferably be carried out by use of an apparatus according to the first aspect.

[0033] According to a third aspect, there is provided a computer program element configured, when executed by a processor, to perform the method of the second aspect and / or to control an apparatus according to the first aspect.

[0034] According to a fourth aspect, there is provided a computer readable storage or transmission medium storing or carrying a computer program element according to the fourth aspect.

[0035] It should be noted that the above examples and embodiments may be combined with each other regardless of the aspect involved. Thus, methods may be combined with structural features of devices of other aspects, and similarly, devices may be combined with features of each other and with features described above with respect to methods. These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0036] Embodiments of the present invention are illustrated in the following drawings. [Brief explanation of the drawings]

[0037] [Figure 1] 1 illustrates, in block diagram form, an exemplary apparatus for determining soft tissue-related properties of a bone constellation from X-ray imaging data. [Figure 2A] 1 illustrates model parameters for an exemplary 3D pose estimation computer model. [Figure 2B] 1 illustrates model parameters for an exemplary 3D pose estimation computer model. [Figure 2C] 1 illustrates model parameters for an exemplary 3D pose estimation computer model. [Figure 2D] 1 illustrates model parameters for an exemplary 3D pose estimation computer model. [Figure 2E] 1 illustrates model parameters for an exemplary 3D pose estimation computer model. [Figure 2F] 1 illustrates model parameters for an exemplary 3D pose estimation computer model. [Figure 3] 1 illustrates an exemplary graph encoding adjacent bones in an exemplary bone constellation. [Figure 4A] Geometric parameters for determining soft tissue related properties of bone constellations from radiographic data are shown in each radiographic image. [Figure 4B] Geometric parameters for determining soft tissue related properties of bone constellations from radiographic data are shown in each radiographic image. [Figure 5] 1 illustrates a flow chart of a method for determining soft tissue-related properties of a bone constellation from X-ray imaging data. DETAILED DESCRIPTION OF THE INVENTION

[0038] 1 shows in a schematic block diagram an exemplary apparatus 100 for determining soft-tissue-related properties of a bone constellation from X-ray imaging data. The apparatus 100 has a data processor 110, which may be any suitable computer, operably connected to a suitable data interface, and which may be operably connected to further computer peripherals. The X-ray imaging data may be provided as input data to the apparatus 100, and in particular to the data processor 110, by a data source 200, such as an X-ray detector, a picture archiving and communication system (PACS), etc. Output data OUT of the data processor 110 may be provided for further processing to another computer device, a software module, an output device, e.g., a display, a reporting system, etc., a PACS, etc.

[0039] The data processor 110 is configured to acquire X-ray imaging data of a region of interest (ROI) including a bone constellation, which may be acquired from a data source 200. The data processor 110 is further configured to utilize a 3D pose estimation computer model CM, into which the acquired X-ray imaging data is fed or input, to determine a pose estimate of the bone constellation and at least one geometric parameter defined between the bones of the bone constellation, and to determine soft tissue-related properties of the bone constellation.

[0040] The soft tissue related properties may include any soft tissue within, at and / or surrounding the bone constellation included in the ROI, i.e., the X-ray imaging data (input data). With respect to the bone constellation, the soft tissue may include, for example, cartilage, ligaments, tendons, muscles, fascia, etc. The soft tissue may extend or be located between bones, adjacent to bones, or at a (small) distance around bones.

[0041] The at least one geometric parameter may include one or more of a distance between bones, an orthogonal translation, and a rotation. The at least one geometric parameter is a variable parameter of the 3D pose estimation computer model CM.

[0042] The 3D pose estimation computer model CM is implemented in the device 100 and may be a computer program element stored in a computer memory or the like and may be executed by the data processor 110 .

[0043] Further, the device 100, e.g., the data processor 110, may be configured to determine a soft-tissue-related lesion and / or abnormality from the at least one soft-tissue-related characteristic. To determine the soft-tissue-related characteristic and / or to determine the soft-tissue-related injury, lesion, and / or abnormality, the device 100 may be further configured to compare the geometric parameters and / or pose estimates with corresponding thresholds. The device 100, e.g., the data processor 110, may be further configured to output the determined at least one soft-tissue-related characteristic and / or the determined injury, lesion, and / or abnormality. The device 100 may be further configured to trigger an alert to be output based on the determined soft-tissue-related lesion and / or abnormality.

[0044] Optionally, the apparatus 100 may be further configured to describe and / or visualize soft-tissue-related characteristics within or based on the 3D pose estimation computer model CM. Furthermore, to determine the at least one soft-tissue relationship, the apparatus 100 may be further configured to determine an axis between two adjacent bones of the bone constellation and measure the distance between the two adjacent bones along the axis, as an example of the at least one geometric parameter. Furthermore, to determine the at least one soft-tissue relationship, the apparatus 100 may be further configured to adapt a respective threshold value for an implant present within or associated with the bone constellation. Optionally, the apparatus 100 may be further configured to combine the at least one geometric parameter, the at least one geometric parameter, and / or the soft-tissue-related characteristics with patient-related data to determine soft-tissue-related lesions and / or abnormalities.

[0045] Figure 2 illustrates, in sub-figures 2A-2F, exemplary model parameters of an exemplary 3D pose estimation computer model CM utilized by data processor 110. This is merely an exemplary selection of possible model parameters, and not all model parameters necessarily need to be considered or coded. Note that the at least one geometric parameter discussed above is an additional free parameter to the model parameters shown in Figures 2A-2F. Each of Figures 2A-2F is an example of an ROI that includes a bone constellation, here a joint, specifically an exemplary ankle joint. The 3D pose estimation computer model CM may be referred to as a 3D joint kinematic model.

[0046] The model parameter in Figure 2A is the flexion angle. Figure 2B shows the model parameters for the line of sight, which indicate (1) azimuthal rotation and (2) polar rotation. The model parameter shown in Figure 2C is the detector position. Figure 2D shows the model parameters for the translator at the detector plane. The model parameter in Figure 2E is the rotation at the detector plane. Figure 2F shows the model parameters for scaling at the detector plane, which may include, for example, changes in joint size or image sampling.

[0047] The 3D pose estimation model CM is trained to regress the model parameters considered or coded therein. For example, such training of the 3D pose estimation model CM can be performed by using images artificially generated by using the 3D articulated shape model and its possible poses. This is not a limitation herein, as the training data can also be generated in different ways.

[0048] 3 shows an exemplary graph encoding adjacent bones in an exemplary bone constellation. In the example shown in FIG. 3, which serves only for better illustration, the bone constellation included in the ROI is again the ankle joint, with the fibula (a), tibia (b), talus (c), and calcaneus (d), and can be shown as a graph encoding these bones. For example, the at least one geometric parameter includes one or more of the distance d between the bones, the orthogonal translation t, and the rotation r. Note again that each of the geometric parameters of distance d, orthogonal translation, and rotation r are model parameters that can be encoded in the 3D pose estimation model in addition to the model parameters shown in FIGS. 2A-2F.

[0049] For the exemplary ankle joint as an example of a bone constellation, introducing the talus-tibia distance d as an additional free parameter allows for more accurate adaptation of the 3D pose estimation computer model to cases with different or abnormal distances. Once this model is parameterized and fitted to an image, the talus-tibia distance d encoded in the model parameter(s) provides diagnostically useful information. For example, the distance d may reflect the size of the upper ankle gap; low values ​​may indicate cartilage degeneration due to age or osteoarthritis, while high values ​​may indicate ligament tears and / or foot dislocation.

[0050] To implement a 3D pose estimation computer model CM with additional geometric parameters, for example, a tree or graph may be used to encode some distances between bones in the 3D pose estimation computer model, where the bones are vertices and the distances are modeled by edges.

[0051] Further, for example, to determine the distance d, the device 100, e.g., the data processor 110, may be configured to construct an axis between two adjacent bones and measure the distance d along that axis.

[0052] Figure 4 shows in sub-figures 4A and 4B the geometric parameters for determining the soft tissue related properties of the bone constellation from radiographic data in each radiographic image. In Figures 4A and 4B, the distance d, translation t and / or rotation r are related to the scapho-lunate enlargement, which can be considered as an indicator of degenerative arthritis.

[0053] As described above, the apparatus 100 is configured to utilize the 3D pose estimation computer model CM, augmented by including at least one additional free parameter, i.e., at least one geometric parameter, such as distance d, translation t, and / or rotation r, to determine scapholunate expansion as an example of a soft tissue-related characteristic. Comparing the scapholunate expansion to normal values, statistics, etc., can be utilized by the apparatus 100 to determine injuries, lesions, abnormalities, etc. in the soft tissue, even though the soft tissue is not visible in the radiographic data.

[0054] 5 is a flowchart illustrating a computer-implemented method for determining soft-tissue-related properties of a bone constellation from X-ray imaging data. The method can be performed by the above-described apparatus 100. The method includes, as step S1, acquiring X-ray imaging data of a region of interest including a bone constellation, and, as step S2, determining a pose estimate of the bone constellation and relative distances between bones of the bone constellation by utilizing a 3D pose estimation computer model CM fed with the X-ray imaging data, the relative distances being assigned to at least one soft-tissue-related property of the bone constellation. As explained above, the at least one soft-tissue-related property can be used for further diagnostic purposes, etc.

[0055] According to a further aspect, there is provided a computer program element for controlling an X-ray imaging system of the third aspect, configured to perform the method steps of the fourth aspect when executed by a processing unit.

[0056] According to a further aspect, there is provided a computer readable medium having stored thereon the computer program element of the fifth aspect.

[0057] The computer program element may therefore be stored in a computing unit, which may be an embodiment of the present invention. This computing unit may be configured to perform or direct the execution of the steps of the above-mentioned method. Furthermore, the computing unit may be configured to operate each component of the above-mentioned apparatus. The computing unit may be configured to operate automatically and / or to execute a user's order. The computer program may be loaded into the working memory of a data processor. The data processor may therefore be configured to perform the method of the present invention.

[0058] This exemplary embodiment of the present invention covers both computer programs that have intervention installed from the beginning, and computer programs that convert existing programs into programs that use the present invention by means of an update.

[0059] The computer program may be stored and / or distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems, however, the computer program may also be presented over a network such as the World Wide Web and can be downloaded into the working memory of a data processor from such a network.

[0060] According to a further exemplary embodiment of the present invention, a medium for making available for downloading a computer program element is provided, the computer program element being configured to perform a method according to one of the aforementioned embodiments of the present invention.

[0061] It should be noted that embodiments of the present invention are described with reference to different subject matters. In particular, some embodiments are described with reference to method-type claims, and other embodiments are described with reference to apparatus-type claims. However, those skilled in the art will infer from the above and following description that, unless otherwise indicated, any combination of features belonging to one type of subject matter, as well as other combinations between features relating to different subject matters, are considered to be disclosed in the present application.

[0062] All features can be combined to provide a synergistic effect greater than the simple sum of the features.

[0063] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive. The invention is not limited to the disclosed embodiments.

[0064] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the dependent claims.

[0065] In the claims, the word "comprise" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be interpreted as limiting the scope. [Explanation of symbols]

[0066] 100 devices 110 Data Processor CM 3D posture estimation computer model 200 Imaging Data Source d, t, r geometric parameters

Claims

1. 1. An apparatus for determining soft tissue related properties of a bone constellation from x-ray imaging data, comprising: acquiring the x-ray imaging data of a region of interest including the bone constellation; determining a pose estimate of the bone constellation and at least one geometric parameter defined between the bones of the bone constellation by utilizing a 3D pose estimation computer model fed with the X-ray imaging data to determine the soft tissue-related properties of the bone constellation; 1. An apparatus having a data processor configured to:

2. The device of claim 1 , wherein the device is further configured to trigger an output of at least the soft tissue-related property.

3. The apparatus according to claim 1 or 2, wherein the apparatus is further configured to describe and / or visualize the soft tissue-related properties in or based on the 3D pose estimation computer model.

4. The apparatus according to claim 1 , wherein the apparatus is further configured to determine soft tissue-related lesions and / or abnormalities from the at least one soft tissue-related characteristic.

5. The apparatus of claim 4 , wherein the apparatus is further configured to compare the geometric parameters and / or pose estimates with corresponding thresholds to determine the soft tissue-related lesions and / or abnormalities.

6. The apparatus of claim 5 , wherein the apparatus is further configured to adapt the respective threshold values ​​in the case of an implant present in or associated with the bone constellation.

7. 7. The apparatus according to claim 5 or 6, wherein the apparatus is further configured to combine the at least one geometric parameter and / or the soft tissue-related property with patient-related data to determine the soft tissue-related lesions and / or abnormalities.

8. The device according to claim 3 , wherein the device is further configured to trigger an alert to be output based on the determined soft tissue-related lesions and / or abnormalities.

9. 9. The apparatus of claim 1, wherein the 3D pose estimation computer model is parameterized by parameters including one or more of a flexion angle of the bone constellation, a gaze direction of the bone constellation, an X-ray detector position relative to the bone constellation, a translation in an X-ray detector plane, a rotation in an X-ray detector plane, a scaling in an X-ray detector plane, and the geometric parameters, and the 3D pose estimation computer model is trained to regress the one or more parameters.

10. The apparatus of claim 1 , wherein the at least one geometric parameter comprises one or more of a distance between the bones, an orthogonal translation, and a rotation.

11. 11. The apparatus of claim 1, wherein the apparatus is further configured to determine an axis between two adjacent bones of the bone constellation and to measure the distance between the two adjacent bones along the axis.

12. 12. The apparatus of claim 1, wherein the at least one geometric parameter comprises a distance between the bones, and wherein a tree or graph is used to encode distances between bones in the 3D pose estimation computer model, the bones being vertices and the distances being modeled by edges.

13. The apparatus of claim 1 , wherein the at least one geometric parameter is a variable parameter of the 3D pose estimation computer model.

14. 1. A computer-implemented method for determining soft tissue-related properties of a bone constellation from x-ray imaging data, comprising: acquiring the x-ray imaging data of a region of interest including the bone constellation; determining a pose estimate of the bone constellation and relative distances between bones of the bone constellation by utilizing a 3D pose estimation computer model fed with the X-ray imaging data, the relative distances being assigned to at least one soft tissue related property of the bone constellation; A method having the following.

15. A computer program configured to perform the method of claim 14 when executed by a processor.

Images