System and method for obtaining accurate measurements and quantification of x-ray images from estimates of key anatomical locations

Abstract

An artificial intelligence (AI) measurement system (1000) for X-ray images (102, 202) is used as a component of or separately from an X-ray imaging system (10) to automatically scan exposed X-ray images (102, 202) to detect and locate various landmarks (304) of anatomical structures (306) presented within the X-ray images (102, 202). A set of key image features (302) located close to each other with known distance relationships are superimposed on the X-ray images (102, 202). The positions of the key image features (302) are then adjusted to correspond to the landmarks (304) within the X-ray images (102, 202). These adjustments are made relative to previously known distance relationships between the key features (302), thus enabling the measurement system (1000) to easily calculate desired angle and length measurements (112) between the landmarks (304).

Description

Technical Field

[0001] This disclosure relates to X-ray imaging systems, and more particularly to X-ray imaging systems including auxiliary image processing systems to improve workflow and the quality of images produced by the X-ray system. Background Technology

[0002] Various X-ray imaging systems of different designs are known and currently in use. These systems are generally based on the generation of X-rays from an X-ray source directed towards a subject of interest. The X-rays traverse the subject and strike a detector (e.g., film, imaging plate, or portable cassette). The detector detects the X-rays that are attenuated, scattered, or absorbed by the subject's interventional structures. For example, in the context of medical imaging, such systems can be used to visualize a subject's internal structures, tissues, and organs for screening or diagnostic purposes.

[0003] Regarding X-ray images produced by an X-ray system, radiologists or other physicians are expected to obtain various angle and length measurements of the anatomical parts shown in the X-ray images for diagnostic and other purposes. To obtain these measurements, radiologists must be able to calculate the distances between different points and / or regions of the anatomical structures presented in the X-ray image. This process is normally achieved using measurement algorithms that are either part of or separate from the X-ray system that provides these measurements.

[0004] When calculating the expected measurement, such as Figure 1 As shown, a prior art AI model / algorithm 100 examines anatomical structures presented within an X-ray image 102 input to the model 100 and performs anatomical landmark segmentation on the anatomical structures. Anatomical landmark segmentation 105 identifies multiple specific and important regions, points, or structures 104 of the anatomical structures in the X-ray image 102, which serve as designated landmarks known to the measurement algorithm. Once identified within the X-ray image 102, the regions 104 are processed by a complex geometric / triangulation post-processing algorithm / computation 106 that provides the desired measurement 112 between the regions 104 shown in the X-ray image 102.

[0005] like Figures 1 to 2 As shown, existing techniques and systems for providing this type of measurement from X-ray image 102 include a single AI model 100 that performs multiple marker segmentation on X-ray image 102. Figure 1 The system 108, and the combination of multiple AI models 100 Figure 2 Alternative system 110, where each AI model 100 is configured to provide anatomical segmentation 105 of a single region 104 of the anatomical structure shown in the X-ray image 102, which is subsequently stitched together before post-processing / measurement calculations.

[0006] See now Figure 3 In an exemplary prior art segmentation process performed on an X-ray image 202 of the pelvis, an AI model 200 analyzes the image 202 to locate various landmarks / structures known to be present in the image of the pelvis, such as the femoral head, femoral axis, pelvic teardrop, acetabular origin, and acetabular lateral margin. The AI ​​model 200 proceeds to identify and locate landmarks within the image 202 and segments the image 202 into various representations 204 of important and / or desired structures of the pelvis. Figure 4 As shown, these representations 204 are then combined with their positions in image 202 to form a composite structure 206 of the pelvis from image 202, and measurements will be based on this composite structure. Geometric / triangulation post-processing 208 is performed on this structure 206 to provide desired angle and length measurements from the structure 206, as presented / shown in measurement image 210.

[0007] This measurement process results in highly accurate measurements of the anatomical structures represented in the X-ray image 202. However, the computational complexity for image segmentation and post-processing requires a significant amount of memory and processing power for the system to perform segmentation and post-processing, as well as a significant amount of time to complete.

[0008] Therefore, it is desirable to develop a system and method for automatically calculating various measurements of anatomical structures presented in X-ray images, which minimizes the computational complexity and time constraints of existing technologies. Summary of the Invention

[0009] According to one aspect of an exemplary embodiment of this disclosure, an artificial intelligence (AI) measurement system for X-ray images is used as a component of or separately from an X-ray imaging system to automatically scan exposed X-ray images to detect and locate various landmarks of anatomical structures presented within the X-ray images. A set of key image features whose locations are close to each other and have known distance relationships are superimposed on the X-ray image. The positions of the key image features are then adjusted to correspond to the landmarks within the X-ray image. These adjustments are made relative to previously known distance relationships between the key features, thus enabling the measurement system to easily calculate desired angle and length measurements between the landmarks.

[0010] According to another aspect of the exemplary embodiments of this disclosure, the AI ​​measurement system may use key points, key lines, or key regions individually, or by combining these key points, key lines, or key regions with each other, as key features to calculate various angle and length measurements of anatomical structures shown in X-ray images.

[0011] According to another exemplary embodiment of the present disclosure, an X-ray system includes an X-ray source, an X-ray detector configurable to be aligned with the X-ray source, and a processing unit operatively connected to the X-ray source and the X-ray detector to generate an X-ray image from data emitted from the X-ray detector. The processing unit includes an X-ray measurement system configured to provide a stack of one or more key features corresponding to one or more landmarks of an anatomical structure within the X-ray image, and to calculate measurements of the anatomical structure based on the location of the key features within the stack.

[0012] According to another aspect of an exemplary embodiment of the present disclosure, a method for determining measurements between landmarks of an anatomical structure within an X-ray image includes the steps of: providing an X-ray system having an X-ray source, an X-ray detector locating to be aligned with the X-ray source, and a processing unit operatively connected to the X-ray source and the X-ray detector to generate an X-ray image from data emitted from the X-ray detector, wherein the processing unit includes an X-ray measurement system configured to provide a stack of one or more key features corresponding to one or more landmarks of an anatomical structure within the X-ray image, and to calculate measurements of the anatomical structure based on the positions of the one or more key features within the stack, thereby applying the stack to the X-ray image, and calculating measurements of the anatomical structure in the X-ray image based on the positions of the one or more key features in the stack applied to the X-ray image.

[0013] These and other exemplary aspects, features and advantages of the invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings. Attached Figure Description

[0014] The accompanying drawings illustrate the currently conceived best mode for practicing the present invention.

[0015] In the attached diagram:

[0016] Figure 1 This is a schematic diagram of the first prior art measurement system used for X-ray imaging.

[0017] Figure 2 This is a schematic diagram of a second prior art measurement system for X-ray imaging.

[0018] Figure 3 This is a schematic diagram of the image segmentation process in a conventional measurement system.

[0019] Figure 4 This is a schematic diagram of the measurement process in an existing technology measurement system.

[0020] Figure 5 This is a schematic diagram of an X-ray imaging system employing an AI measurement system according to an exemplary embodiment of the present disclosure.

[0021] Figure 6 This is a flowchart of an operation method of an AI measurement system according to an exemplary embodiment of the present disclosure.

[0022] Figures 7A to 7B This is a schematic diagram of the output of an AI measurement system according to an exemplary embodiment of the present disclosure, compared to prior art segmentation measurements.

[0023] Figures 8A to 8B This is a schematic diagram of the output of an AI measurement system according to another exemplary embodiment of the present disclosure, compared with the segmentation results of existing technologies.

[0024] Figures 9A to 9B This is a schematic diagram of the output of an AI measurement system using key points and key lines according to another exemplary embodiment of this disclosure. Detailed Implementation

[0025] One or more specific implementations will be described below. To provide a concise description of these implementations, not all characteristics of the actual implementation may be described in the specification. It should be understood that, as in any engineering or design project, numerous implementation-specific decisions must be made in the development of any such implementation to achieve the developer's specific objectives, such as complying with system-related and business-related constraints that may differ between implementations. Furthermore, it should be understood that such development efforts may be complex and time-consuming, but remain routine tasks of design, fabrication, and manufacturing for those skilled in the art who benefit from this disclosure.

[0026] When describing elements of various embodiments of the invention, the articles “a,” “an,” “the,” and “described” are intended to mean the presence of one or more such elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that additional elements may be present in addition to those listed. Furthermore, any numerical examples in the following discussion are intended to be non-limiting, and therefore the additional values, ranges, and percentages are within the scope of the disclosed embodiments. As used herein, the terms “substantially,” “largely,” and “approximately” indicate a condition within reasonably achievable manufacturing and assembly tolerances relative to the ideal desired condition suitable for achieving the functional purpose of a component or assembly. Moreover, as used herein, “electrically coupled,” “electrically connected,” and “electrically communicated” mean that the referenced elements are directly or indirectly connected such that current can flow from one to the other. This connection may include a direct conductive connection (i.e., without intervening capacitors, inductors, or active elements), an inductive connection, a capacitive connection, and / or any other suitable electrical connection. Intervening components may be present. As used in this article, the term "real-time" refers to a level of processing responsiveness that is sufficiently immediate for the user to perceive or that enables the processor to keep up with external processes.

[0027] See Figure 5 A block diagram of an x-ray imaging system 2000 according to an embodiment is shown. The x-ray imaging system 2000 includes an x-ray source 111 radiating x-rays, a stand 132 for a subject 105 to stand on during examination, and an x-ray detector 134 for detecting x-rays radiated by the x-ray source 111 and attenuated by the subject 105. As a non-limiting example, the x-ray detector 134 may include a scintillator, one or more ionization chambers, a photodetector array, an x-ray exposure monitor, an electrical substrate, etc. The x-ray detector 134 is mounted on a stand 138 and configured to be vertically movable according to the imaging area of ​​the subject.

[0028] The operation console 160 includes a processor 161, a memory 162, a user interface 163, a motor driver 145 for controlling one or more motors 143, an X-ray power unit 114, an X-ray controller 116, a camera data acquisition unit 121, an X-ray data acquisition unit 135, and an image processor 150. X-ray image data emitted from the X-ray detector 134 is received by the X-ray data acquisition unit 135. The collected X-ray image data is processed by the image processor 150. A display device 155, communicatively coupled to the operation console 160, displays the image-processed X-ray image on it.

[0029] The X-ray source 111 is supported by a support column 141, which may be mounted to the ceiling (e.g., as depicted) or on a movable stand for positioning within an imaging chamber. The X-ray source 111 is vertically movable relative to the subject or patient 105. For example, one of one or more motors 143 may be integrated into the support column 141 and may be configured to adjust the vertical position of the X-ray source 111, for example, by increasing or decreasing the distance between the X-ray source 111 and the ceiling or floor. For this purpose, a motor driver 145 of an operating console 160 may be communicatively coupled to one or more motors 143 and configured to control one or more motors 143. The one or more motors 143 may be further configured to adjust the angular position of the X-ray source 111 to change the field of view of the X-ray source 111, as further described herein.

[0030] The x-ray power unit 114 and x-ray controller 116 provide power to the x-ray source 111 with appropriate voltage and current. A collimator (not shown) can be fixed to the x-ray source 111 to specify the field of view of the x-ray beam. The x-ray beam radiated from the x-ray source 111 is applied to the subject via the collimator.

[0031] The X-ray source 111 and camera 120 can pivot or rotate relative to the support column 141 in the angular direction 129 to image different parts of the subject 105.

[0032] Memory 162 is a suitable electronic storage medium and / or computer-readable medium storing X-ray image 170 and executable instructions 172, which, when executed, cause one or more of processor 161 and image processor 150 to perform one or more actions. Regarding Figure 5 The X-ray measurement system 1000 and AI application 1002 are further described herein as example methods that can be stored as executable instructions 172.

[0033] Processor 161 also includes an automated X-ray measurement system 1000, optionally stored in memory 162 as part of executable instructions 172 adopted by processor 161 to perform the functions of measurement system 1000. While measurement system 1000 is shown and described in conjunction with X-ray system 2000, it is also contemplated for use with other types of imaging systems, including but not limited to computed tomography (CT) systems, magnetic resonance imaging (MRI) systems, and ultrasound (US) systems, as well as other compatible imaging systems. X-ray measurement system 1000 is formed by artificial intelligence (AI) application 1002, which scans and detects X-ray images 1004 after exposure. Figure 6 Various types of information associated with (see Figure 9). For example, it could be an AI application 1002 based on a deep learning neural network, which is an image-based object detection application configured to detect various attributes of the exposed X-ray image 1004, such as information about various landmarks located in the anatomical structures presented within the X-ray image 1004.

[0034] Now for reference Figure 6 In an exemplary embodiment of the operation method 400 of the X-ray measurement system 1000, initially in block 402, the AI ​​application 1002 creates a stack 300 of multiple key features 302. Figure 7A , Figure 8A The key feature corresponds to the estimated or proximate location of each of a corresponding number of landmarks 304 found within a selected anatomical structure 306 presented in X-ray image 1004. The estimation of the location of the key feature 302 can be accomplished in any suitable manner, such as by using a training set of X-ray images so that system 1000 / AI application 1002 can determine the location of each landmark 304 in the training images and determine the average location of the landmarks 304 in the training image set for use in locating the key feature 302 in a stack 300 of anatomical structure 306 or one or more different anatomical structures 306 presented in X-ray image 1004. In the stack 300, the distances, angles, and other measurements between each key feature 302 are known, making it easy and direct to determine any desired measurement from the known relationships between the locations of the key features 302 relative to each other.

[0035] In box 404, X-ray image 1004 is supplied or input to system 1000, and system 1000, such as on display 46, positions the overlay 300 over the anatomical structures 306 within the X-ray image 1004 to provide a visual indication of the overlay 300 of the X-ray image 1004. In box 406, system 1000 analyzes the X-ray image 1004 using any suitable image viewing process or method employed by AI application 1002 to determine the exact location of markers 304 associated with each key feature 302 forming the overlay 300.

[0036] After determining the exact location of each marker 304 within the X-ray image 1004, in box 408, the AI ​​application 1002 proceeds to adjust or edit the position of each associated key feature 302 of the marker 304, if necessary, such that the position of the key feature 302 in the overlay 300 corresponds exactly to the position of the associated marker 304 in the X-ray image 1004. Adjustments to the key features 302 may alternatively or in association with the AI ​​application 1002 be performed by the user, such as by moving the position of one or more key features 302 in the overlay 300 on the X-ray image 1004 via the user interface 44.

[0037] By adjusting this and corresponding known quantities of known distances, angles, or other measurement parameters between each key feature 302 in the stack 300, the desired measurement between any combination of key features 302 and associated markers 304 within the X-ray image 1004 can be determined and provided in box 410. Based on changes in the relative positions of key features 302 corresponding to the positions of markers 304 in the X-ray image 1004 within the stack 300, the complexity of geometric / triangulation calculations required in prior art measurement systems is replaced in system 1000 by relatively simple adjustments to existing and known measurement parameters between the individual key features 302 in the stack 300. In this way, measurement system 1000 and AI application 1002 enable the rapid and direct determination of various desired measurements between markers in the X-ray image 1004.

[0038] Furthermore, in an alternative embodiment of method 400, the steps in blocks 404 through 408 may be reversed or altered in sequence, such that, for example, AI application 1002 may determine the exact location of the marker 304 before positioning the overlay 300 onto the anatomical structure 306 in the X-ray image 1004, and / or may perform adjustments to the position of the key feature 302 within the overlay 300 before or without positioning the overlay 300 onto the X-ray image 1004. Additionally, the step in block 204 positioning the overlay 300 onto the X-ray image 1004 may be moved after adjusting the position of the key feature 302, so that a visual representation of the overlay 300 onto the anatomical structure 306 in the X-ray image is provided only after all adjustments have been performed. Alternatively, the entire process of method 400 may be maintained as an internal process within system 1000, wherein the visual representation of the overlay 300 is not presented on, for example, display 46.

[0039] Regarding key features 302, the form of feature 302 can be selected by the user and / or AI application 102 as needed, such as depending on the anatomical structures 306 presented within the X-ray image 1004, where different anatomical structures 306 have different key features 302. In an exemplary embodiment, one or more key features 302 may take the form of one or more key points 312, one or more key lines 314, and / or one or more key regions 316, and combinations thereof. Various types of key features 302 can be displayed on display 46 in various ways, such as by changing the type (e.g., point, intersection, etc.), color, size, blink, or other attributes of the key feature 302 currently being analyzed and / or viewed by AI application 102.

[0040] As the first example of method 400, in Figure 7A In the X-ray image 1004, the anatomical structure 306, including anatomical landmarks 304, is shown positioned adjacent to a stack 300 containing multiple key features 302 in the form of keypoints 312. The stack 300 is also shown positioned above the anatomical structure 306 in the X-ray image 1004, wherein the positions of the keypoints 312 are adjusted from their positions in individual stacks 300 to be positioned directly above the landmarks 304 detected in the X-ray image 1004 associated with each keypoint 312. This X-ray image 1004 in Figure 7B The images are juxtaposed, which have similar X-ray images 202 that have undergone prior art segmentation, to identify markers 204 in the images 202 and form a segmentation structure 206 (also superimposed on the images 202), on which complex post-processing calculations must be performed to provide the required measurements from the structure 206.

[0041] The second example of the process of method 400 is in Figures 8A to 8B As shown in the figure, Figure 8A The stack 300 includes a small number of key features 302 in the form of keypoints 312 within the stack 300, which are located above the marker 304 in the X-ray image 1004. Again with Figure 8B Compare layer 300 on image 1004. Figure 8B The segmentation structure 206 produced in the prior art method is shown. Furthermore, a distance measurement line 1006 obtained by the system 1000 in the current method 400 is shown, corresponding to a known distance between two (2) key points 312 associated with a marker 304 on the anatomical structure 306 in the X-ray image 1004. Figure 8A )compared to, Figure 8A and Figure 8B Each is shown as an exemplary measurement line, wherein the line is a distance measurement line 212 (cut between two (2) marks 204 on the segmented structure 206 via prior art methods. Figure 8B ).

[0042] in addition, Figures 9A to 9B An alternative embodiment of an X-ray image 1004 output by system 1000 / AI application 1002 is shown, wherein the overlay 300 provided on image 1004 includes key features 302 in the form of keypoints 312, keylines 314, and key regions 316 corresponding to various markers 304 in X-ray image 1004. As previously described, the representations of keypoints 312, keylines 314, and key regions 316 corresponding to various markers 304 in X-ray image 1004 have been determined in training suitable for system 1000 / AI application 1002, wherein the average positions of keypoints 312, keylines 314, and key regions 316 corresponding to various markers 304 in a particular anatomical structure 306 are determined from a dataset including multiple X-ray images 1004 of anatomical structures 306 supplied to system 1000 / AI application 1002. Various key points 312, key lines 314, and key regions 316 in the anatomical structure 306 are then automatically adjusted by the system 1000 / AI application 1002 and / or manually adjusted by the user to directly position the key points 312, key lines 314, and key regions 316 above their corresponding landmarks in the anatomical structure 306 to form a stack 300. In this configuration, based on previously known relationships between the various key points 312, key lines 314, and key regions 316, and adjustments made to these relationships in forming the stack 300, the system 1000 / AI application 1002 can easily calculate the required measurements within the anatomical structure 306 presented in the X-ray image 1004. Furthermore, the system 1000 / AI application 1002 can be trained to...

[0043] Finally, it should be understood that System 1000 / AI Application 1002 may include necessary computers, electronics, software, memory, storage devices, databases, firmware, logic / state machines, microprocessors, communication links, displays or other visual or audio user interfaces, printing devices, and any other input / output interfaces for performing the functions described herein and / or achieving the results described herein. For example, as previously described, the system may include at least one processor / processing unit / computer and system memory / data storage structure, which may include random access memory (RAM) and read-only memory (ROM). At least one processor of the system may include one or more conventional microprocessors and one or more auxiliary coprocessors, such as math coprocessors. The data storage structures discussed herein may include suitable combinations of magnetic, optical, and / or semiconductor memories, and may include, for example, RAM, ROM, flash drives, optical discs such as compact discs, and / or hard disks or drives.

[0044] Additionally, software applications / algorithms adapted to perform the methods disclosed herein can be read from a computer-readable medium into the main memory of at least one processor. As used herein, the term "computer-readable medium" refers to any medium that provides or participates in providing instructions for execution to at least one processor of System 10,1000 (or any other processor of the device described herein). Such media can take many forms, including but not limited to non-volatile and volatile media. Non-volatile media include, for example, optical, magnetic, or optical disks, such as memory. Volatile media include dynamic random access memory (DRAM), which typically constitutes main memory. Common forms of computer-readable media include, for example, floppy disks, flexible disks, hard disks, magnetic tape, any other magnetic media, CD-ROMs, DVDs, any other optical media, RAM, PROMs, EPROMs or EEPROMs (electronically erasable programmable read-only memory), FLASH-EEPROMs, any other memory chips or cassette tapes, or any other media from which a computer can read.

[0045] Although in the embodiments, execution of a sequence of instructions in a software application causes at least one processor to perform the methods / processes described herein, hardwired circuitry may be used in place of or in combination with software instructions to implement the methods / processes of the present invention. Therefore, embodiments of the present invention are not limited to any particular combination of hardware and / or software.

[0046] It should be understood that the compositions, apparatus, and methods described herein are not limited to the specific embodiments and methods, as these are subject to variation. It should also be understood that the terminology used herein is for the purpose of describing specific exemplary embodiments only and is not intended to limit the scope of this disclosure, which will be limited only by the appended claims.

Claims

1. A method for determining measurements between landmarks (304) of anatomical structures (306) within an X-ray image (102, 202), the method comprising the steps of: - Provides an X-ray system (10), the X-ray system comprising: -X-ray source (14); - An X-ray detector (18) capable of being positioned and aligned with the X-ray source (14); and - A processing unit (40) operatively connected to the X-ray source (14) and the X-ray detector (18) to generate X-ray images (102, 202) from data emitted from the X-ray detector (18), wherein the processing unit (40) includes an X-ray measurement system (1000) configured to provide a stack (300) of one or more key features (302) corresponding to one or more landmarks (304) of an anatomical structure (306) within the X-ray images (102, 202), and to calculate a measurement (112) of the anatomical structure (306) based on the location of the one or more key features (302) in the stack (300). - Apply the stack (300) to the X-ray image (102, 202); - Adjust the position of the one or more key features (302) within the stack (300) to align with the associated marker (304) in the X-ray images (102, 202); and -Based on the location of the one or more key features (302) in the stack (300) applied to the X-ray images (102, 202), calculate the measurement (112) of the anatomical structure (306) in the X-ray images (102, 202).

2. The method according to claim 1, wherein the X-ray measurement system (1000) is formed by artificial intelligence (AI) (200).

3. The method of claim 2, wherein the AI ​​(200) is trained to provide the stack (300) including the average position of the one or more key features (302) corresponding to the one or more landmarks (304) of the anatomical structure (306).

4. The method of claim 3, wherein the AI ​​(200) is trained to provide the stack (300) including the average position of the one or more key features (302) of the one or more signs (304) corresponding to the plurality of anatomical structures (306).

5. The method of claim 2, wherein the AI ​​(200) is configured to provide the stack (300) with at least one of key points (312), key lines (314), and key regions (316) corresponding to the one or more key features (302), the one or more key features corresponding to the one or more markers (304) of the anatomical structure (306).

6. The method of claim 1, wherein the step of applying the stack (300) to the X-ray image (102, 202) comprises presenting the stack (300) in conjunction with the X-ray image (102, 202) on a display (46).

7. The method of claim 1, wherein the location of the one or more key features (302) within the stack (300) is known, and wherein the step of calculating the measurement (112) of the sign (304) in the anatomical structure (306) based on the location of the one or more key features (302) in the stack (300) comprises determining the measurement (112) of one key feature (302) relative to another key feature (302).

8. The method of claim 1, wherein the step of adjusting the position of the one or more key features (302) within the stack (300) to align with the associated marker (304) in the X-ray image (102, 202) is performed automatically by the X-ray measurement system (1000).

9. The method of claim 1, wherein the step of calculating the measurement (112) of the sign (304) in the anatomical structure (306) based on the location of the one or more key features (302) relative to the sign (304) in the X-ray images (102, 202) comprises determining the measurement (112) of a key feature (302) relative to the adjusted key feature (302).

10. The method of claim 1, wherein the step of adjusting the position of the one or more key features (302) within the stack (300) relative to the one or more markers (304) in the X-ray images (102, 202) comprises: - Move the one or more key features (302) to align the one or more key features (302) with corresponding markers (304) in the anatomical structure (306); and - Determine the adjustment position of one or more key features (302) within the stack.

11. The method of claim 1, wherein the step of calculating the measurement (112) of the sign (304) in the anatomical structure (306) based on the location of the one or more key features (302) relative to the sign (304) in the X-ray image (102, 202) comprises determining a measurement (112) of an adjusted key feature (302) relative to another adjusted key feature (302).

12. An X-ray system (10), the X-ray system comprising: -X-ray source (14); - X-ray detector (18), which can be positioned to be aligned with the X-ray source (14); and - Processing unit (40), which is operatively connected to the X-ray source (14) and the X-ray detector (18) to generate X-ray images (102, 202) from data emitted from the X-ray detector (18). The processing unit (40) includes an X-ray measurement system (1000) formed by artificial intelligence (AI) (200), which is configured to provide a stack (300) of one or more key features (302) corresponding to one or more landmarks (304) of an anatomical structure (306) in the X-ray images (102, 202), adjust the position of the one or more key features (302) in the stack (300) to align with the associated landmarks (304) in the X-ray images (102, 202), and calculate a measurement (112) of the anatomical structure (306) based on the position of the key features (302) in the stack (300).

13. The X-ray system of claim 12, wherein the AI ​​(200) is trained to provide the stack (300) including the average position of the one or more key features (302) corresponding to the one or more landmarks (304) of the anatomical structure (306).

Images