Diagnostic scan acquisition support method and system

Abstract

To provide methods and systems for assisted diagnostic scan acquisition.SOLUTION: Various methods and systems are provided for an x-ray imaging system (100). In an example, a method for the system includes, in response to operation of the x-ray imaging system in an automated mode, generating a difference image (600) from a first image (400) and a second image (500). The difference image is displayed at a display device (232) to allow motion outside of a region of interest (ROI) (402) to be detected based on analysis of the difference image (1114). Further, the difference image is displayed before a contrast agent reaches the ROI (1122).SELECTED DRAWING: Figure 10

Description

Technical Field

[0001] Embodiments of the subject matter disclosed herein relate to the acquisition of X-ray images, particularly computed tomography images.

Background Art

[0002] Computed tomography (CT) can be used as a non-invasive medical imaging technique. Specifically, CT image data acquisition can include passing an X-ray beam through an object such as a patient so that the X-ray beam is attenuated, and collecting the attenuated X-ray beam with an X-ray detector array. Similar to high-contrast detection that enables a CT imaging system to visualize a contrast agent, imaging of various tissues including bone, soft tissue, etc. becomes possible.

[0003] In an example where a contrast agent is used, a CT imaging system may be configured to automatically detect changes in image contrast within a monitored region of interest (ROI). In response to a change in contrast exceeding a threshold, acquisition of a diagnostic scan may be automatically initiated. In some embodiments, movement in and / or near the ROI may trigger early or late acquisition of a diagnostic scan for an instance of an actual change in contrast that reaches the threshold.

Summary of the Invention

[0004] In one embodiment, the X-ray imaging system includes generating a difference image from a first image and a second image in response to the operation of the X-ray imaging system in automatic mode, and displaying the difference image on a display device. By displaying the image on the display device, motion outside the region of interest (ROI) may be detected based on the analysis of the difference image. The difference image can be displayed before the contrast agent reaches the ROI. In this way, the automatic initiation of diagnostic scan acquisition can be supported by an optional process for triggering diagnostic scan acquisition when motion is present outside the ROI. This makes the initiation of diagnostic scan acquisition more robust to motion that might otherwise cause false triggering of scan acquisition. As a result, diagnostic information about the subject can be obtained more efficiently while reducing the subject's X-ray exposure.

[0005] It should be understood that the above brief explanation is provided to simplify and introduce some of the concepts that will be further described in the detailed explanation. It is not intended to identify any important or essential features of the claimed subject matter, the scope of which will be independently defined by the claims that follow the detailed explanation. Furthermore, the claimed subject matter is not limited to embodiments that resolve any of the defects pointed out above or in any part of this disclosure. [Brief explanation of the drawing]

[0006] The present invention will be better understood by reading the following description of non-limiting embodiments with reference to the accompanying drawings. [Figure 1] This is a diagram illustrating an imaging system according to one embodiment. [Figure 2] This is a schematic block diagram of an exemplary imaging system according to an embodiment. [Figure 3] The image shows a scout image acquired by a CT imaging system according to an embodiment. [Figure 4] Figure 2 shows a first cross-sectional image acquired by a CT imaging system based on the scout image according to the embodiment. [Figure 5] A second cross-sectional image, obtained by a CT imaging system based on the scout image in Figure 2, is shown according to the embodiment. [Figure 6] The first difference image according to the embodiment is shown. [Figure 7] A second difference image according to the embodiment is shown. [Figure 8] This embodiment illustrates a high-level workflow for assisting scan acquisition in a CT imaging system. [Figure 9] This is a graph plotting the contrast in the region of interest of the difference image according to the embodiment. [Figure 10] This shows a display shown to the operator during scan acquisition support in a CT imaging system, according to one embodiment. [Figure 11A] This document describes a method for acquiring CT images using a CT imaging system configured with assisted scan acquisition according to one embodiment. [Figure 11B] This document describes a method for acquiring CT images using a CT imaging system configured with assisted scan acquisition according to one embodiment. [Modes for carrying out the invention]

[0007] The following description relates to various embodiments of diagnostic scan acquisition assistance in X-ray imaging systems. More specifically, diagnostic scan acquisition in a CT system may be automatically triggered and modified based on the detection of motion in a region of interest (ROI). An example of a CT system is shown in Figure 1, and a block schematic diagram of the components of the CT system is shown in Figure 2. During a diagnostic event, a scout image, as shown in Figure 3, may be acquired first to orient the CT system. The scout image may then be used as a guide to obtain a cross-sectional image focused on the region of interest (ROI) (examples of which are shown in Figures 4-5). A difference image may be generated to determine whether or not there is motion around the ROI in order to determine the mode of scan acquisition initiation performed by the CT system. Examples of difference images are shown in Figures 6-7. The difference image may be used to select the mode of scan acquisition initiation according to a higher-level workflow shown in Figure 8, in which scan acquisition assistance may be performed when motion is detected. In the initial stages of scan acquisition, as shown in the graph in Figure 9, the contrast difference in the cross-sectional images acquired by the CT system may be monitored, and this difference may be displayed to the operator along with the CT image, as shown in Figure 10. An example of how to acquire CT images using assisted scanning is shown in Figures 11A to 11B.

[0008] Before further discussing approaches to assist in scan acquisition, we provide a general overview of CT imaging. Figure 1 shows an exemplary CT system 100 configured for CT imaging. In particular, the CT system 100 is configured to image a subject 112, such as a patient, an inanimate object, one or more manufactured parts, and / or foreign objects such as dental implants, stents, and / or contrast agents present in the body. In one embodiment, the CT system 100 includes a gantry 102, which may further include at least one X-ray source 104 configured to project an X-ray emission beam 106 (see Figure 2) for use in imaging the subject 112 lying on a table 114. Specifically, the X-ray source 104 is configured to project the X-ray emission beam 106 toward a detector array 108 located opposite the gantry 102. Although Figure 1 depicts only a single X-ray source 104, in certain embodiments, multiple X-ray sources and detectors may be employed to project multiple X-ray emission beams 106 to acquire projection data at different energy levels corresponding to the patient. In some embodiments, the X-ray source 104 can enable dual-energy gemstone spectral imaging (GSI) by rapid peak kilovoltage (kVp) switching. In some embodiments, the employed X-ray detector is a photon counting detector capable of distinguishing X-ray photons of different energies. In other embodiments, two sets of X-ray sources and detectors are used to generate dual-energy projection, one at low kVp and the other at high kVp. Therefore, it should be understood that the methods described herein can be implemented not only with dual-energy acquisition techniques but also with single-energy acquisition techniques.

[0009] In certain embodiments, the CT system 100 further includes an image processing unit 110 configured to reconstruct an image of a target volume of a subject 112 using an iterative or analytic image reconstruction method. For example, the image processing unit 110 can use an analytical image reconstruction technique such as filtered back projection (FBP) to reconstruct an image of a target volume of a patient. As another example, the image processing unit 110 can reconstruct an image of a target volume of a subject 112 using iterative image reconstruction approaches such as advanced statistical iterative reconstruction (ASIR), conjugate gradient (CG), maximum likelihood expectation maximization (MLEM), or model-based iterative reconstruction (MBIR). As will be further described herein, in some embodiments, the image processing unit 110 can use both iterative image reconstruction approaches and analytical image reconstruction approaches such as FBP.

[0010] In some CT imaging system configurations, the X-ray source is collimated to be located in the XYZ plane of the Cartesian coordinate system, projecting a cone-shaped X-ray beam, commonly called the "imaging plane." The X-ray beam passes through the object being imaged, such as a patient or subject. After being attenuated by the object, the X-ray beam collides with an array of detector elements. The intensity of the attenuated X-ray beam received by the detector array depends on the attenuation of the beam by the object. Each detector element in the array generates a separate electrical signal, which is a measurement of the X-ray beam attenuation at the detector location. The attenuation measurements from all detector elements are acquired individually to create a transmission profile.

[0011] In some CT systems, the X-ray source and detector array rotate with the gantry in the imaging plane and around the object being imaged so that the angle at which the radiation beam intersects the object is constantly changing. A set of X-ray emission attenuation measurements from the detector array at a given gantry angle, e.g., projection data, is called a “view.” A “scan” of an object includes a series of views taken at different gantry angles, i.e., view angles, during one rotation of the X-ray source and detector. Since the advantages of the method described herein are intended to apply to medical imaging modalities other than CT, the term “view,” as used herein, is not limited to the use described above with respect to projection data from a single gantry angle. The term “view” is used to mean a single data acquisition whenever there are multiple data acquisitions from different angles, whether from CT, positron emission tomography (PET), or single-photon emission CT (SPECT), and / or other modalities, including modalities not yet developed, and combinations thereof in fused embodiments.

[0012] Projection data is processed to reconstruct images corresponding to two-dimensional slices passing through an object, and in some examples where the projection data includes multiple views or scans, a three-dimensional rendering of the object is reconstructed. One method for reconstructing an image from a set of projection data is called filtered back projection in the art. Transmission tomography and emission tomography reconstruction techniques also include statistical iterative methods such as maximum likelihood expectation maximization (MLEM) and ordered-subsets expectation-reconstruction techniques, as well as iterative reconstruction techniques. This process converts attenuation measurements from scans into integers called "CT numbers" or "Hounsfield units," which are used to control the brightness of the corresponding pixels on a display device.

[0013] To reduce the total scan time, a "helical" scan may be performed. A helical scan involves moving the patient while acquiring data for a predetermined number of slices. In such systems, a single helix is ​​generated from the cone-beam helical scan. The cone-beam-mapped helix generates projection data, from which images of each predetermined slice can be reconstructed.

[0014] As used herein, the phrase “reconstructing an image” is not intended to exclude embodiments of the invention in which data representing an image is generated, but a viewable image is not. Therefore, as used herein, the term “image” broadly refers to both a viewable image and the data representing a viewable image. However, many embodiments generate (or are configured to generate) at least one viewable image.

[0015] Figure 2 shows an exemplary imaging system 200 similar to the CT system 100 in Figure 1. According to aspects of this disclosure, the imaging system 200 is configured to image a subject 204 (e.g., subject 112 in Figure 1). In one embodiment, the imaging system 200 includes a detector array 108 (see Figure 1). The detector array 108 further includes a plurality of detector elements 202 that together sense an X-ray emission beam 106 (see Figure 2) passing through the subject 204 (e.g., a patient) to acquire corresponding projection data. Thus, in one embodiment, the detector array 108 is fabricated in a multislice configuration including a plurality of cells or columns of detector elements 202. In such a configuration, one or more additional rows of detector elements 202 are arranged in a parallel configuration to acquire projection data.

[0016] In certain embodiments, the imaging system 200 is configured to traverse different angular positions around the subject 204 in order to acquire desired projection data. Thus, the gantry 102 and the components mounted thereon may be configured to rotate about a rotation center 206, for example, to acquire projection data at different energy levels. Alternatively, in embodiments where the projection angle relative to the subject 204 changes as a function of time, the mounted components may be configured to move along a general curve rather than along a segment of a circle.

[0017] When the X-ray source 104 and the detector array 108 rotate, the detector array 108 collects data on the attenuated X-ray beam. The data collected by the detector array 108 undergoes preprocessing and calibration to adjust the data so that it represents the line integral of the attenuation coefficients of the scanned subject 204. The processed data is generally referred to as a projection.

[0018] In some examples, the individual detectors or detector elements 202 of the detector array 108 can include photon-counting detectors that register the interactions of individual photons into one or more energy bins. It should be understood that the methods described herein can also be implemented with energy-integrating detectors.

[0019] The set of acquired projection data can be used for basis material decomposition (BMD). During BMD, the measured projections are converted into a set of material density projections. The material density projections can be reconstructed to form a set or collection of material density maps or images of the respective basis materials, such as bone, soft tissue, and / or contrast agent maps. The density maps or images may then be associated to form volume renderings of the underlying materials, such as bone, soft tissue, and / or contrast agent, within the imaged volume.

[0020] Once reconstructed, the basis material images generated by the imaging system 200 reveal the internal features of the subject 204 represented by the densities of two basis materials. The density images can be displayed to show these features. In conventional approaches to the diagnosis of medical conditions, such as medical events, a radiologist or physician considers a hard copy or display of the density image to identify characteristic features of interest. Such features include lesions, the size and shape of specific anatomical structures or organs, and other features distinguishable from the image based on the individual physician's skill and knowledge.

[0021] In one embodiment, the imaging system 200 includes a control mechanism 208 that controls the movement of components such as the rotation of the gantry 102 and the operation of the X-ray source 104. In certain embodiments, the control mechanism 208 further includes an X-ray controller 210 configured to supply power and timing signals to the X-ray source 104. Additionally, the control mechanism 208 includes a gantry motor controller 212 configured to control the rotation speed and / or position of the gantry 102 based on imaging requirements.

[0022] In a particular embodiment, the control mechanism 208 further includes a data acquisition system (DAS) 214 configured to sample the analog data received from the detector element 202 and convert the analog data into a digital signal for subsequent processing. The DAS 214 can be further configured to selectively aggregate the analog data from a subset of the detector elements 202 into so-called macro detectors, as further described herein. The data sampled and digitized by the DAS 214 is transmitted to a computer or computing device 216. In one example, the computing device 216 stores the data in a storage device or mass storage device 218. The storage device 218 can include, for example, a hard disk drive, a floppy disk (trademark) drive, a compact disc read / write (CD-R / W) drive, a digital versatile disc (DVD) drive, a flash drive, and / or a solid state storage device drive.

[0023] Furthermore, the computing device 216 provides commands and parameters to one or more of the DAS 214, the X-ray controller 210, and the gantry motor controller 212 to control system operations such as data acquisition and / or processing. In certain embodiments, the computing device 216 controls system operations based on operator input. The computing device 216 receives operator input, including commands and / or scanning parameters, for example, via an operator console 220 operably coupled to the computing device 216. The operator console 220 may include a keyboard (not shown) or a touchscreen to allow the operator to specify commands and / or scanning parameters.

[0024] In some examples, as described herein, the computing device 216 may consist of instructions, e.g., algorithms, for initiating the acquisition of a diagnostic scan in response to a trigger according to an automatic scan acquisition mode. In one embodiment, the trigger may be a change in image contrast by at least a threshold amount, corresponding to the detection of a contrast agent in the ROI. The instructions may also enable the detection of motion in the vicinity of the ROI (e.g., the area surrounding the ROI), which may prompt an option to initiate acquisition according to an assisted scan acquisition mode presented to the operator. Further details of the automatic scan acquisition mode and the assisted scan acquisition mode are provided below with reference to Figures 3 to 11.

[0025] Although only one operator console 220 is illustrated in Figure 2, multiple operator consoles may be coupled to the imaging system 200 for purposes such as inputting or outputting system parameters, requesting examinations, plotting data, and / or viewing images. Furthermore, in certain embodiments, the imaging system 200 may be coupled to multiple displays, printers, workstations, and / or similar devices located either locally or remotely, for example, within an institution or hospital, or in entirely different locations, via one or more configurable wired and / or wireless networks, such as the Internet and / or virtual private networks, wireless telephone networks, wireless local area networks, wired local area networks, wireless wide area networks, and wired wide area networks.

[0026] In one embodiment, for example, the imaging system 200 includes or is coupled to a picture archiving and communications system (PACS) 224. In an exemplary embodiment, the PACS 224 is further coupled to remote systems such as a radiology information system, a hospital information system, and / or an internal or external network (not shown), allowing operators in different locations to supply commands and parameters and / or access image data.

[0027] The computing device 216 operates the table motor controller 226 using commands and parameters supplied by the operator and / or defined by the system, which can control the table 114, which may be an electric table. Specifically, the table motor controller 226 can move the table 114 to properly position the subject 204 within the gantry 102 in order to acquire projection data corresponding to the target volume of the subject 204.

[0028] As described above, the DAS214 samples and digitizes the projection data acquired by the detector element 202. The image reconstructor 230 then performs high-speed reconstruction using the sampled and digitized X-ray data. While Figure 2 illustrates the image reconstructor 230 as a separate entity, in certain embodiments, the image reconstructor 230 can form part of the computing device 216. Alternatively, the image reconstructor 230 may not be present in the imaging system 200, and instead, the computing device 216 may perform one or more functions of the image reconstructor 230. Furthermore, the image reconstructor 230 may be located locally or remotely and may be operably connected to the imaging system 200 using a wired or wireless network. In particular, in one exemplary embodiment, computing resources within a “cloud” network cluster can be used for the image reconstructor 230.

[0029] In one embodiment, the image reconstruction device 230 stores the reconstructed image in the storage device 218. Alternatively, the image reconstruction device 230 may transmit the reconstructed image to the computing device 216 to generate patient information useful for diagnosis and evaluation. In certain embodiments, the computing device 216 can transmit the reconstructed image and / or patient information to a display or display device 232 that is communicatively coupled to the computing device 216 and / or the image reconstruction device 230. In some embodiments, the reconstructed image may be transmitted from the computing device 216 or the image reconstruction device 230 to the storage device 218 for short-term or long-term storage.

[0030] Various methods and processes further described herein (such as those described later with reference to Figures 11A and 11B) can be stored as executable instructions in non-transient memory on a computing device (or controller) within the imaging system 200. In one embodiment, the computing device 216 may include such executable instructions to automatically generate a difference image based on two sequentially acquired images in the initial stages of scan acquisition. The difference image can be used to determine whether motion is occurring near but outside the ROI, and this determination can be used, according to the executable instructions, to adjust the start of scan acquisition between an automated and an assisted mode.

[0031] By enabling the CT system to operate in either automatic or assisted mode at the start of a diagnostic scan, the start of scan acquisition can be more closely correlated with the appearance of contrast agent within the ROI, potentially improving the quality of reconstructed images output by the CT system. In automatic mode, scan acquisition can be started without operator input, while in assisted mode, scan acquisition can be started by input from an operator or a machine learning model. For example, during CT imaging of a subject, the subject may move while the CT system is operating. In some embodiments, such as when a contrast agent is used to improve resolution, the CT system may be configured to automatically start scan acquisition when it detects that an increase in contrast (e.g., image contrast) in the acquired image of the ROI has reached a threshold compared to previously acquired images. However, if the subject moves, distortion of the acquired image may falsely trigger diagnostic scan acquisition. In some cases, inaccurate detection of contrast increase may result from image noise.

[0032] As a result, the reconstructed images may have poor image quality, such as resolution and sharpness, due to inaccurate scan acquisition triggers, potentially requiring further scans of the subject. Furthermore, because the automated scan acquisition start mode tends to be triggered by events other than the contrast increase reaching a threshold, operators may be hesitant to use the automated scan acquisition start mode. Consequently, operators may become entirely reliant on manual scan acquisition triggers, which can reduce workflow efficiency and increase the operator's burden.

[0033] In one embodiment, as described herein, the above-mentioned problems can be addressed at least in part by adapting the CT system to an assist mode for initiating scan acquisition, which can operate in conjunction with the automated mode. The assist mode may include instructions to generate at least one difference image during initial frame acquisition and evaluation, such as after a scout image has been obtained and before the contrast agent is expected to reach the ROI. The difference image may be generated from the actual current image and the initial mask image and may be displayed to the operator on a display screen of the operation console, such as the display device 232 in Figure 2. This allows the operator to override the operation in automated mode and instead select to operate in the assist mode for scan acquisition based on observation of the difference image. Additionally, an actual image of the ROI, such as a cross-sectional image acquired by the CT system, may be projected onto the display screen, thereby additionally or alternatively allowing the operator to monitor the movement of the actual image.

[0034] In another embodiment, an automated process can be used to evaluate the difference image for motion that may cause distortion, which could lead to an incorrect triggering of the scan acquisition. If motion is detected, the operator may be notified and offered the option to switch to an assist mode for scan acquisition, which allows the operator to manually trigger the scan acquisition. Simultaneously with the processing of the difference image by the image processing algorithm, the operator can also observe the difference image on the display screen and manually select operation in the assist mode based on a visual inspection of the difference image.

[0035] For example, machine learning algorithms can be used to analyze difference images and identify motion. Once motion is detected, the machine learning algorithm may include instructions to delay the start of scan acquisition until no more motion is detected. Thus, scan acquisition can be initiated via an automated mode assisted by machine learning.

[0036] Furthermore, machine learning algorithms may be used to adjust the position of the ROI from frame to frame during frame collection before scan acquisition begins. For example, a trained machine learning model may be configured to identify ROIs in an image, track the ROI's position based on the difference image, and output an updated image with the ROI in real time. As an example, a machine learning model may be trained on acquired data and, for example, moved data manipulated in known ways to simulate the movement of the ROI. For example, supervised learning can be achieved using motionless reference images along with multiple simulations of moving objects. The machine learning model may output an indicator of whether patient movement was detected between reconstructed time frames.

[0037] Alternatively, ROI realignment can be performed using fast image alignment algorithm motion. For example, an initial image, such as a mask image, can be used as a template for aligning subsequently acquired images. To achieve image alignment, image registration algorithms such as rigid registration and deformable registration approaches may be used, or a registration model may be trained in a manner similar to that described above for machine learning models. As the displayed actual image is updated, the ROI is repositioned within the actual image.

[0038] Another option for tracking ROI location is to register the entire image, such as a mask image, before placing the ROI. Subsequent images can be registered relative to landmarks, which are easily identifiable features adjacent to the ROI. Image registration allows the ROI's position to be updated in response to subject movement. Such automated ROI registration can provide accurate CT count measurements despite patient movement. Automated registration can be achieved using a model trained as described above with respect to machine learning models, but the model's output may instead be coordinates for translating the ROI, rather than a binary value indicating whether or not movement has occurred. Alternatively, ROI translation may not be required in cases where fast image alignment is used, but it may be required when fast image alignment is applied. In either case, the user can observe the images, and the ROI can be robustly positioned on the anatomy of interest, such as the aorta. However, image registration is a relatively slow process and may not be applicable if movement causes the ROI to move out of the imaging plane.

[0039] Details of the assisted scan acquisition initiation are provided in the following descriptions of Figures 3 to 11. First, looking at Figure 3, an example of a scout image 300 is shown. A scout image 300 may be used to indicate the target area to be acquired and imaged by the CT system at the start of an imaging event. For example, a scout image 300 may be a digital X-ray image, such as a two-dimensional X-ray image, containing an anatomical region of the subject from which cross-sectional images can be acquired. The scan range for subsequent CT scan acquisitions may be defined by the scout image 300.

[0040] Markers can be placed on the scout image 300 to indicate the desired scan plane for CT scan acquisition. For example, an operator can add a line 302 directly to the scout image 300, which can then be used to position the gantry of the CT system, for example, the gantry 102 in Figures 1 and 2, to target the indicated scan plane. Cross-sectional images may then be acquired according to the line 302. An example of a first cross-sectional image 400 corresponding to the line 302 on the scout image 300 is shown in Figure 4.

[0041] The first cross-sectional image 400 is a frame showing a slice of the subject's anatomical structure and may be acquired while the CT system gantry is stationary. Markers indicating ROI 402 may be placed on the first cross-sectional image 400, for example, by the operator, to indicate a specific location where data acquisition is desired. ROI 402 may, for example, be the aorta, which may be adjacent to landmark 404, which is another anatomical feature such as bone.

[0042] When automated scan acquisition is initiated, ROI 402 can monitor changes in visual parameters. For example, when a contrast agent is introduced into the subject, e.g., injected or ingested, frames acquired following the first cross-sectional image 400 (with the gantry still) can be monitored for changes in contrast in the ROI as the contrast agent circulates within the subject.

[0043] As an example, Figure 5 shows a second cross-sectional image 500 that captures the same field of view (FOV) as the first cross-sectional image 400, which includes ROI 402 and landmark 404. The second cross-sectional image 500 may be a frame collected immediately after the first cross-sectional image 400 was acquired, according to a predetermined image acquisition frequency. Alternatively, the second cross-sectional image 500 may be a frame acquired after the first cross-sectional image 400, but not immediately afterward (for example, additional frames may have been collected in between).

[0044] In the second cross-sectional image 500, the contrast at ROI402 may increase due to infiltration of the contrast agent. This change in contrast can affect X-ray attenuation, and the change in contrast can be observed in the scan when the contrast agent reaches ROI402. The scan is processed using an image processing algorithm, and the acquisition of the diagnostic scan may be automatically initiated when the change in contrast at ROI402 in the frame reaches a preset threshold.

[0045] However, the automated mode for initiating scan acquisition may not be robust to movement that could cause apparent contrast changes in the ROI, resulting in false triggers for scan acquisition. For example, if a subject moves during the initial period of image acquisition to monitor contrast changes between images, adjacent anatomical features may move within the marked ROI. As an example, landmark 404 may move into an area marked as ROI 402 by the operator if the subject moves. The contrast difference between bone and aorta may be sufficient to trigger the acquisition of a diagnostic scan because the contrast change reaches a threshold.

[0046] Movement around and outside the ROI can be detected from a difference image generated from the acquired cross-sectional image. The difference image may be an image resulting from subtracting one image frame from another image frame, which is collected at a different time. If no movement occurs outside ROI 402, the first difference image 600 shown in Figure 6 may be generated.

[0047] In one example, the first difference image 600 may be generated by subtracting a first cross-sectional image, acquired before the contrast agent is delivered to the subject, from a second cross-sectional image, acquired after the contrast agent is delivered to the subject. The first cross-sectional image may be called a mask image, and the second cross-sectional image may be called a contrast image. The mask image can be subtracted pixel by pixel from the contrast image to generate the first difference image 600, which can be displayed to the operator in real time. In the first difference image 600, the ROI 402 is a clear circular spot that shows high contrast to the region surrounding the ROI 402. However, it will be understood that variations in the appearance of the ROI in the difference image may vary depending on the type of anatomical feature targeted in the ROI.

[0048] However, if movement occurs in and / or around ROI402, a difference image such as the second difference image 700 shown in Figure 7 can be generated instead. The second difference image 700 can be generated, as described above, for example, by subtracting the first cross-sectional image (e.g., mask image) from a second cross-sectional image (e.g., contrast image) acquired later. In the second difference image 700, ROI402 also appears as a bright circular spot with high contrast against the surrounding area. Additional structures that may appear as distortion 702 are also visible in the difference image 700. Distortion 702 may result from displacement of anatomical features captured in the first and second cross-sectional images, which are a result of movement in and / or around ROI402.

[0049] In one embodiment, when the second difference image 700 is displayed to the operator, the operator may observe the distortion 702 and choose to override the automated scan acquisition initiation in the CT system to mitigate a false trigger for diagnostic scan acquisition (by disabling and replacing the automated scan acquisition initiation). In another example, difference images generated in real time during the initial monitoring period of a CT scan acquisition event can be processed using an image processing algorithm (e.g., software) to detect motion by identifying distortions such as the distortion 702 in Figure 7. Once distortions are identified from the difference images, the image processing algorithm may include a command to display a notification to the operator that motion has been detected. The operator can then decide how to proceed, for example, by continuing with the automated scan acquisition initiation or by switching to a manual, operator-activated initiation.

[0050] In addition or alternatively, machine learning algorithms in machine learning models can be trained to detect motion within an ROI based on difference images. For example, machine learning algorithms can be trained to analyze difference images for the presence of distortion using various techniques such as feature mapping, image registration, and image classification. Furthermore, machine learning algorithms can be trained to simulate patient motion by resampling volumes between subsequent time frames. Upon distortion detection, the machine learning algorithm may be configured to notify the operator of the detected motion, as described above, for example, with reference to processing via an image processing algorithm. This allows the operator to decide whether to proceed with automated or assisted (e.g., manual) scan acquisition.

[0051] As another example, a trained machine learning model may be integrated into an automated mode for starting scan acquisition. For instance, difference images may be automatically monitored and analyzed by the machine learning model to detect motion within an ROI based on distortion identification. If motion within an ROI is detected, the trained machine learning model can override the automated mode for starting scan acquisition (e.g., the default setting for the automated mode used when no motion is detected).

[0052] The machine learning model may also be configured to generate a vector map based on the difference image, which may be used to guide the repositioning of the ROI to reflect the actual location of the target anatomical feature being scanned. The contrast image and / or difference image can be updated with the repositioned ROI for each newly acquired image frame, and the contrast can be monitored. When the contrast reaches a threshold, scan acquisition may be initiated.

[0053] The machine learning model may be implemented in machine learning-assisted automatic mode, but the operator may be provided with the option to override the machine learning-assisted automatic mode and switch to manually initiated scan acquisition. If the operator chooses to manually initiate scan acquisition, the machine learning model can continue to reposition and update ROIs in contrast and / or difference images.

[0054] Alternatively, as described above, the ROI can be repositioned based on a high-speed image registration algorithm or image registration. When a high-speed image registration algorithm is used, contrast-enhanced and / or difference images may be displayed with the repositioned ROI as the images are updated. In the case of image registration, contrast cross-sectional images acquired in real time may be registered to mask cross-sectional images acquired at the start of the scan acquisition event. If motion is detected in or around the ROI, for example by an operator or a machine learning model, that motion can be used to reposition the ROI.

[0055] For example, as described above, images can be registered based on landmarks identified in the mask image and contrast image. For instance, the position of ROI 402 relative to landmark 404, which represents bone in Figures 4 and 5, can be determined in the mask image. Landmark 404 is more opaque, and its contrast may be independent of the contrast agent. Changes in the position of the landmark may be mapped and used to reposition the ROI according to its relative position to the landmark.

[0056] Image registration can provide an established method for resizing images, and is widely used for aligning two-dimensional images. However, in the case of CT systems, ROI movement can occur within the imaging plane or deviate from it depending on the subject's movement. Thus, image registration is reliable for compensating for motion within the imaging plane, but may not be able to track movement in directions other than those along the imaging plane. Similarly, fast image registration algorithms may only be efficient techniques for aligning two-dimensional images. Furthermore, image registration can be an iterative and computationally intensive process, potentially requiring a longer completion time than the frequency of frame captures. Thus, ROI resizing using machine learning models may be more efficient and robust.

[0057] Figure 8 shows a high-level workflow 800 for a CT system configured to have automated and assisted modes for initiating scan acquisition. For example, the CT system may be the imaging system 200 shown in Figure 2. In workflow 800, step 802, a contrast threshold increase is set as a trigger for initiating scan acquisition, which may depend on the kVp of acquisition and the contrast injection protocol used. The threshold may be set before the scout image is acquired, or after the scout image is acquired but before the acquisition of cross-sectional images begins (e.g., a cross-sectional image used as a trigger for diagnostic scan acquisition). After the threshold is set, the acquisition of cross-sectional images may begin, which may include acquiring contrast images following mask images. The operator can specify an ROI in the mask image, which is propagated to the subsequently captured contrast images.

[0058] A difference image may be generated from the mask image, and the contrast in the ROI may be determined from the difference image. The difference image may be updated at a rate corresponding to the image acquisition frequency of the contrast image. In 804 of workflow 800, the difference image may be used to detect motion outside the ROI and within the frame of the difference image, according to one of the strategies described above. If no motion is detected, workflow 800 proceeds to 806 and starts a diagnostic scan acquisition via an automated mode that does not depend on operator input. For example, a scan acquisition may be triggered when it is determined that the contrast of the difference image has reached a threshold.

[0059] However, if motion is detected in 804, workflow 800 proceeds to 808 instead and switches to assistance mode for initiating diagnostic scan acquisition. In some embodiments, instead of automatically switching to assistance mode in response to motion detection, a prompt requesting input from the operator may be displayed, providing an alert that motion has been detected and offering the option to switch to assistance mode. Assistance mode may depend on the operator initiating scan acquisition or on a machine learning model. In one embodiment, during operator assistance, the option may be provided to reposition the ROI by either fast image alignment or a machine learning model to better identify when the operator should trigger scan acquisition. If assisted by a machine learning model, the ROI is repositioned and changes in contrast can be monitored.

[0060] The algorithm providing commands for performing assist and automatic scan acquisition in a CT system may include, for example, commands for outputting and displaying information to the operator on a display screen. The displayed information may include a real-time graph 900 plotting the contrast in the ROI of the difference image over time, as shown in Figure 9, where the contrast in the difference image represents the change in contrast between the images used to generate the difference image. Each of several data points 902 may be generated when the difference image is updated based on acquisition with a new contrast image, and the difference image takes into account movement such that the ROI is repositioned between image frames. The contrast images may be collected according to a preset frequency, and the real-time graph 900 may be updated with new data points at a rate corresponding to the preset frequency.

[0061] The real-time graph 900 may include threshold contrast 904, as shown by the dashed line. Threshold contrast 904 may be an increase in contrast indicating that the contrast agent has reached the ROI. The real-time graph 900 may be displayed at least in the early stages of a CT imaging event, such as before a diagnostic scan acquisition is triggered.

[0062] The real-time graph 900 display can also provide the operator with a visual countdown of the period during which overriding the start of automatic scan acquisition is permitted. For example, during the period indicated by bracket 906, the operator may be presented with the option to switch to the assist mode for starting scan acquisition if motion is detected (by the operator or a machine learning model). This period may extend after the first cross-sectional image, which may be used as a mask image, has been acquired, and before the contrast reaches the threshold contrast 904. If a switch to the assist mode is not instructed before the timer period expires, the acquisition of the diagnostic scan may proceed once the contrast reaches the threshold contrast 904.

[0063] If the operator chooses to switch to assistance mode within the period indicated by bracket 906, the real-time display of graph 900 is maintained, and the display of images, as described later with reference to Figure 10, may be continuously updated as new images are collected and generated. However, after the period indicated by bracket 906, the option to switch to assistance mode is no longer available, and graph 900 may be removed from the display.

[0064] In one example of the display screen 1000, as shown in Figure 10, the real-time graph 900 may be displayed together with a cross-sectional image. This cross-sectional image may be the first cross-sectional image 400 in Figure 4, but in other examples it may be the second cross-sectional image 500 in Figure 5, or it may be any mask image or contrast image that was most recently acquired in the initial stages of the CT imaging event.

[0065] Furthermore, the display screen 1000 may display a difference image, which may, for example, be the second difference image 700 in Figure 7. The displayed difference image may be the most recently generated difference image and may be updated as new contrast-enhanced images are obtained. The cross-sectional image and the difference image may be presented in the initial stages of a CT imaging event, allowing the operator to observe changes in contrast and clarity in the cross-sectional image and the presence of distortion in the difference image.

[0066] For example, the cross-sectional image 400 is displayed and updated each time a new cross-sectional image is acquired. Each refresh of the cross-sectional image may correspond to one of the data points 902 of the real-time graph 900, which are added as the cross-sectional images are acquired. A difference image may be displayed after the first cross-sectional image has been acquired, which may be a mask image, as shown by bracket 906 in Figure 9. In some embodiments, movement around the ROI of the image may be visible in the cross-sectional image and can be used by the operator to determine whether an assistance mode is desirable. For example, the display screen 1000 may include a manual action request for switching to an assistance mode, such as a selectable button.

[0067] The period indicated by bracket 906 in Figure 9 allows the operator to monitor the updated difference image as each new cross-sectional image is acquired and used to generate the difference image. In one embodiment, if distortion is present in the difference image, it can be easily observed by the operator, which may additionally trigger the display of a notification regarding motion detected outside the ROI of the image. The display of the notification may be output by image processing software. Furthermore, an option may be presented to switch to a scan acquisition start assistance mode.

[0068] If the operator selects the assistance mode, the operator can monitor the difference image and initiate scan acquisition based on visual observation. For example, the operator can manually trigger scan acquisition when they observe that distortion is no longer present in the difference image. If the distortion persists throughout the period indicated by bracket 906 in Figure 9, the operator can either cancel the scan acquisition or allow scan acquisition to begin when data point 902 in graph 900 reaches the threshold contrast 904.

[0069] Alternatively, a machine learning model can be used from the start of cross-sectional image acquisition to monitor the difference image for distortion. If distortion indicating motion outside the ROI is detected, the machine learning model can automatically override the automatic scan acquisition start to control the trigger for scan acquisition. Scan acquisition can be triggered by the machine learning model when no motion is detected in the difference image. The machine learning model may also include instructions to monitor contrast according to data point 902 in graph 900. If a period of time elapses as shown in bracket 906 of Figure 9 and motion is still detected outside the ROI, the operator may be presented with options such as, for example, stopping or continuing scan acquisition. Graph 900 may be displayed until the contrast reaches a threshold contrast 904 or until no motion is detected in the difference image and scan acquisition is started. Similarly, once scan acquisition starts, the difference image may be removed from the display screen 1000.

[0070] In this way, the use of automated scan acquisition initiation in CT systems can be more robust to motion occurring outside the ROI, which could otherwise lead to false triggering of scan acquisition. By implementing an assist mode that can be used to override the automated mode when motion is detected, scan acquisition can be triggered at the operator's discretion or according to analysis by a machine learning model. Motion can be identified from difference images generated by subtracting a mask image from contrast images that are continuously acquired and updated until a change in contrast in the image reaches a threshold. By configuring an assist mode (e.g., by a machine learning model or operator) as an option to support the automated mode, the use of the automated mode may become more attractive to the operator. This may improve the overall workflow and efficiency while reducing the subject's X-ray exposure.

[0071] An example of method 1100 for acquiring a diagnostic CT scan is shown in Figures 11A-11B. Instructions for performing method 1100 may be executed by a processor in a computing device, such as the computing device 216 in Figure 2, based on instructions stored in the controller's memory. In one embodiment, the CT system may be the CT system 100 or imaging system 200 in Figure 2. Instructions may include a high-speed image alignment algorithm and / or an algorithm for a machine learning model trained to analyze images acquired and generated by the CT system. The CT system may be configured to operate in an automatic mode in which the contrast of the difference images can be automatically monitored, a manual mode in which image and scan acquisition is controlled by an operator, and an auxiliary mode that can be performed in conjunction with the automatic mode.

[0072] Prior to or at the start of Method 1100, the contrast agent may be delivered to the subject to be scanned. For example, the contrast agent may be injected into or ingested by the subject. An estimated duration for the contrast agent to circulate within the subject and reach the target region of the subject may be predetermined and used as a time frame reference for the execution of Method 1100. For example, Method 1100 may be performed within the time interval between the delivery of the contrast agent to the subject and the circulation of the contrast agent to the target region, so that enhanced contrast can be observed. Thus, certain elements such as difference images and graphs, as further described below, may be generated and displayed only before the contrast agent reaches the ROI.

[0073] Turning to Figure 11A first, in 1102, method 1100 includes checking whether an automated operating mode has been selected by the operator. If the automated mode has not been selected, method 1100 proceeds to 1104 to operate in manual mode. In manual mode, the operator can observe cross-sectional images acquired by the CT system and displayed on a display device such as the display device 232 in Figure 2. The operator can manually initiate the acquisition of a diagnostic scan if motion between time frames is evident based on the observation of the difference images.

[0074] For example, if the operator selects the automated mode, method 1100 proceeds to 1106 to acquire a scout image. For example, the scout image may be the scout image 300 in Figure 3, and is used to orient the CT system toward the subject and to indicate from where the cross-sectional image will be acquired. After the scout image is acquired, a mask image is acquired in 1108. The mask image may be a first cross-sectional image obtained in the region shown in the scout image with the gantry of the CT system stationary.

[0075] In step 1110, method 1100 includes checking whether an ROI is indicated (displayed, pointed out, suggested). For example, an ROI may be indicated by the operator selecting or placing a marker on the mask image. If an ROI is not indicated, method 1100 is held in step 1110 until an ROI is indicated. If an ROI is indicated by the operator, method 1100 proceeds to step 1112, where a contrast image is collected and a difference image is generated based on the contrast image and the mask image. The contrast image and difference image can be displayed to the operator along with a graph plotting the contrast at the ROI in the difference image over time, such as graph 900 in Figures 9 and 10. The difference image can be generated by subtracting the mask image pixel by pixel from the contrast image, resulting in an image like the example difference image depicted in Figures 6-7.

[0076] In 1114, method 1100 includes checking whether motion is detected in the difference image outside the ROI. For example, motion in a subject may be indicated by the presence of distortion in the difference image, such as distortion 702 in Figure 7. Distortion can be identified by an operator based on visual observation or by a machine learning model. If no motion is detected, method 1100 proceeds to 1116 to check whether the contrast in the ROI in the difference image reaches a threshold contrast. Threshold contrast may correspond to the expected change in contrast when the contrast agent reaches the ROI.

[0077] If the contrast does not reach the threshold, method 1100 returns to 1112 to collect another contrast image and generate a corresponding difference image. The newly acquired contrast image and its corresponding difference image can replace the previous contrast image and difference image, thereby updating the display on the display device. The graph may also be updated to include new data points corresponding to the contrast in the newly generated difference image. If the contrast meets the threshold at 1116, method 1100 proceeds to 1128 (as shown in Figure 11B) to remove the difference image and graph from the display as described below.

[0078] Returning to 1114, if motion is detected in the difference image, method 1100 proceeds to 1118 to display a notification or warning to the operator that motion is present. Displaying a notification or warning may also include displaying an option to switch to an assistance mode for the operation or prompting the operator to do so. In 1120, method 1100 includes checking whether an assistance mode has been requested. If no request for an assistance mode has been confirmed in 1120, method 1100 proceeds to 1116 as described above to check whether the contrast in the difference image has reached a threshold. If an assistance mode request has been confirmed in 1120, method 1100 proceeds to 1122 to continue acquiring the contrast image, generating the difference image, and updating the display of the image and graph.

[0079] In one embodiment, refreshing the display of a contrast image may involve repositioning the ROI frame by frame to compensate for the motion of the subject. The ROI may be repositioned using, for example, a fast image alignment algorithm. In another embodiment, a mask image may be registered before positioning the ROI, and the contrast image may be registered with the mask image instead of repositioning the ROI. In yet another embodiment, the ROI may be repositioned by a machine learning model based on a vector map generated by the machine learning model to predict the movement of the ROI. For example, the vector map may be generated from difference images, for example, by analyzing distortion. In one embodiment, using a machine learning model to update the position of the ROI in the displayed contrast image may be faster and less computing-intensive than image registration.

[0080] Next, looking at Figure 11B, at 1124, method 1100 includes checking whether the contrast of the ROI in the difference image reaches a threshold, as described above in 1116. If the contrast does not reach the threshold, method 1100 proceeds to 1126 to check whether a request to start a diagnostic scan has been received. If no request is received, method 1100 returns to 1124 to check again whether the contrast reaches the threshold. At 1126, if a request to start a diagnostic scan is received, method 1100 proceeds to 1128 to stop displaying the difference image and graph. In other words, no further difference images are generated, and both the difference image and graph may be removed from the display on the display device. The option to switch to assistance mode is no longer available if it has not already been selected. Method 1100 then proceeds to 1130 to start acquiring a diagnostic scan of the ROI. A request to start acquiring a scan can be received from the operator when the operator observes that there is no longer any distortion in the difference image. Alternatively, this request can be received from a machine learning model when, based on its analysis of the difference images, it determines that there is no longer any motion.

[0081] Returning to step 1124, if the contrast in the ROI of the difference image reaches a threshold, method 1100 proceeds to step 1132 to check whether a request to abort the scan acquisition has been received. For example, an operator may request to abort the scan acquisition if the motion does not stop and / or if the motion is so large that it impairs the ability to obtain useful information from the diagnostic scan. Similarly, a request to abort the scan acquisition may be received from the machine learning model if it determines that the motion is too large to reliably track the ROI, and that the resulting scan is likely to be unusable. If an abort request is received at step 1132, the scan acquisition event ends at step 1134. If no abort request is received, method 1100 proceeds to step 1128 to remove the difference image and graph from the display screen. Method 1100 then proceeds to step 1130 to begin acquiring the diagnostic scan.

[0082] Figure 1 shows an example of a configuration with the relative positions of various components. If components are shown to be in direct contact with or directly connected to one another, such components can be said to be in direct contact with or directly connected to one another, in at least one instance. Similarly, components shown to be continuous or adjacent to one another may be continuous or adjacent to one another, in at least one instance. For example, components laid in surface contact with one another may be said to be in surface contact. Another example is that, in at least one instance, components are spaced apart from one another, with only space between them and no other components present, and these may be referred to as such. Yet another example is that components shown above / below each other, on opposite sides of each other, or to the left / right of each other may be referred to as such relative to each other. Furthermore, as shown in the figure, in at least one instance, the topmost element or point on an element may be referred to as the “top” of the component, and the bottommost element or point on an element may be referred to as the “bottom” of the component. As used herein, top / bottom, top / bottom, and top / bottom are relative to the vertical axis of the figure and may be used to describe the relative positions of the elements in the figure. Thus, an element shown above another element is, in one example, positioned vertically above the other element. In yet another example, the shape of an element depicted in a figure may be described as having that shape (e.g., circular, straight, flat, curved, rounded, chamfered, angled, etc.). Furthermore, elements shown intersecting each other may, in at least one example, be described as intersecting elements or intersecting each other. Additionally, an element shown within another element, or outside of another element, may be referred to in this way, in one example.

[0083] The Disclosure also provides support for a method for an X-ray imaging system, comprising the steps of: generating a difference image from a first image and a second image in response to the operation of the X-ray imaging system in automated mode; and displaying the difference image on a display device to enable detection of movement outside a region of interest (ROI) based on analysis of the difference image, wherein the difference image is displayed before the contrast agent reaches the ROI. In a first embodiment of the Method, the difference image is displayed after the second image is acquired and is continuously updated on the display device as new images are acquired. In a second embodiment of the Method, the first embodiment is optionally included, but the difference image is no longer displayed once the contrast agent reaches the ROI. In a third embodiment of the Method, one or both of the first and second embodiments are optionally included, the first image is a mask image and the second image is a contrast image, both the first and second images are displayed on the display device, and the difference image is generated by subtracting the first image from the second image. In a fourth embodiment of this method, optionally including one or more of the first to third embodiments, motion outside the ROI is detected by the presence of distortion around the ROI in the difference image. In a fifth embodiment of this method, optionally including one or more of the first to fourth embodiments, the method further includes continuously acquiring new images and displaying the actual image using the new images, with new difference images being generated using each of the new images and displayed on the display device. In a sixth embodiment of this method, optionally including one or more of the first to fifth embodiments, the method further includes presenting a notification on the display device when motion is detected outside the ROI, the notification including an option to switch the operation of the X-ray imaging system to support mode. In a seventh embodiment of this method, optionally including one or more of the first to sixth embodiments, when the X-ray imaging system is operating in support mode, scan acquisition is initiated based on operator input based on the operator's visual observation of motion.In the eighth embodiment of this method, motion is detected by optionally including one or more of the first to seventh embodiments or each of each embodiment, using a machine learning model trained to identify motion in the difference image, or by an operator visually observing the difference image. In the ninth embodiment of this method, optionally including one or more of the first to eighth embodiments or each of each, a graph plotting the contrast in the ROI of the difference image against time is also displayed on the display device, with each data point on the graph corresponding to the acquisition of a new contrast image and the generation of a new difference image.

[0084] The disclosure also provides support for an X-ray imaging system comprising: an X-ray source and an X-ray detector positioned on opposite sides of a gantry; a subject positioned within the gantry between the X-ray source and the X-ray detector; and a processor of a computing device composed of executable instructions, which, when executed, causes the processor to perform the following steps: when operating in automatic mode, generate a difference image by subtracting a mask image from a contrast image; display the difference image on a display device; display a notification when motion outside the ROI is detected in the difference image; and acquire a diagnostic scan in response to receiving a request to start acquiring a diagnostic scan, the request to start acquiring a diagnostic scan is received when motion outside the ROI is no longer detected in the difference image. In a first embodiment of the system, the mask image is a first cross-sectional image acquired by the X-ray imaging device, the contrast image is a cross-sectional image acquired after the mask image, the contrast image is acquired at a predetermined frequency and displayed on a display device together with the difference image. In a second embodiment of the system, optionally including the first embodiment, the display of the difference image and contrast image is refreshed at a predetermined frequency until the contrast at the ROI in the difference image reaches a threshold contrast, which indicates that the contrast agent has reached the ROI. In a third embodiment of the system, optionally including one or both of the first and second embodiments, motion in the difference image is automatically detected by a machine learning model trained to detect motion in the difference image based on the presence of distortion. In a fourth embodiment of the system, optionally including one or more or each of the first to third embodiments, the machine learning model is further trained to determine the new position of the ROI based on the difference image and to reposition the ROI in the contrast image when the display of the contrast image on the display device is updated.

[0085] This disclosure also provides support for a method for operating an X-ray imaging system. The method includes the steps of generating a difference image from a mask image and a contrast image in response to the operation of the X-ray imaging system in automatic scan acquisition start mode, and displaying the difference image and the contrast image on a display device, and providing an option to switch the operation of the X-ray imaging system to an assisted scan acquisition start mode if motion is detected in the difference image outside of a region of interest (ROI); and, upon receiving confirmation to switch the operation to an assisted scan acquisition start mode, continuously refreshing the display of the difference image and the contrast image, wherein refreshing the display includes repositioning the ROI to compensate for motion; and, in response to receiving a request to start a diagnostic scan acquisition, starting a diagnostic scan acquisition. In a first embodiment of the Method, the Method further includes stopping the diagnostic scan acquisition when a stop request is received from an operator or a machine learning model. In a second embodiment of the Method, which optionally includes the first embodiment, the repositioning of the ROI includes updating the contrast image at the new position of the ROI between frames using a fast image alignment algorithm. A third embodiment of the method optionally includes one or both of the first and second embodiments, wherein the ROI location is tracked in the contrast image based on image registration, and image registration is performed before the ROI is indicated in the contrast image. A fourth embodiment of the method optionally includes one or more of the first to third embodiments or each of each embodiment, wherein ROI repositioning includes updating the ROI location in the contrast image frame by frame by a machine learning model, and the machine learning model is configured to generate a vector map based on difference images to determine the ROI location.

[0086] Where used herein, an element or step described in the singular and preceded by the word "a" or "an" should be understood not to exclude the plural form of that element or step unless such exclusion is expressly stated. Furthermore, a reference to "one embodiment" of the invention is not intended to be construed as excluding the existence of additional embodiments that also incorporate the mentioned features. Furthermore, unless the opposite is expressly stated, an embodiment that "includes," "equips," or "has" an element or a plurality of elements having a particular characteristic may include additional such elements that do not possess that characteristic. The terms "includes" and "has" are used as plain equivalents of the terms "includes" and "has," respectively. Furthermore, terms such as "first," "second," and "third" are used simply as labels and are not intended to impose numerical requirements or a specific positional order on their subjects.

[0087] This specification uses examples to disclose the present invention, including best modes, and to enable a person having ordinary skill in the relevant art to practice the invention, including the manufacture and use of any device or system, and the execution of methods incorporating the invention. The patentable scope of the present invention is defined by the claims and may include other examples that a person skilled in the art can conceive. Such other examples are intended to be included in the claims if they have structural elements that are not different from the language of the claims, or if they include equivalent structural elements that are substantially not different from the language of the claims.

[0088] Further aspects of the present invention are provided by the subject matter of the following clauses. [Embodiment 1] A method for an X-ray imaging system, Steps to respond to the operation of the X-ray imaging system in automatic mode, The steps include generating a difference image from the first image and the second image, The steps include displaying a difference image on a display device and enabling the detection of motion outside the region of interest (ROI) based on the analysis of the difference image, The differential image is displayed before the contrast agent reaches the ROI, including the method. [Embodiment 2] The method according to Embodiment 1, wherein the difference image is displayed after the second image is acquired and is continuously updated on the display device as a new image is acquired. [Embodiment 3] The method according to Embodiment 1, wherein once the contrast agent reaches the ROI, the difference image is no longer displayed. [Embodiment 4] The method according to Embodiment 1, wherein the first image is a mask image, the second image is a contrast image, both the first and second images are displayed on a display device, and the difference image is generated by subtracting the first image from the second image. [Embodiment 5] The method according to Embodiment 1, in which movement outside the ROI is detected by the presence of distortion around the ROI in the difference image. [Embodiment 6] Furthermore, the method according to Embodiment 1 includes the steps of continuously acquiring new images and displaying the actual image made up of the new images, wherein a new difference image is generated from each of the new images and displayed on a display device. [Embodiment 7] The method according to Embodiment 1, further comprising the step of presenting a notification on a display device when motion is detected outside the ROI, the notification including an option to switch the operation of the X-ray imaging system to an assist mode. [Embodiment 8] The method according to Embodiment 7, wherein when the X-ray imaging system is operating in support mode, scan acquisition is initiated based on input from the operator, which is based on the operator's visual observation of movement. [Embodiment 9] The method according to Embodiment 1, wherein motion is detected using a machine learning model trained to identify motion in the difference images, or by an operator visually observing the difference images. [Embodiment 10] The method according to Embodiment 1, wherein a graph plotting the contrast at the ROI in the difference image against time is also displayed on the display device, and each data point in the graph corresponds to the acquisition of a new contrast image and the generation of a corresponding new difference image. [Embodiment 11] An X-ray imaging system, An X-ray source and X-ray detector are located on the opposite side of the gantry, Between the X-ray source and the X-ray detector, the subject is placed inside the gantry, A processor of a computing device consisting of executable instructions, which, when executed, the processor When operating in automatic mode, the steps include: subtracting the mask image from the contrast image to generate a difference image; The steps include displaying the difference image on a display device, If movement outside the ROI is detected in the differential image, a notification is displayed. Steps include: acquiring a diagnostic scan in response to receiving a request to start acquiring a diagnostic scan; A processor and other components are used to perform this task. The X-ray imaging system receives a request to start a diagnostic scan when no motion outside the ROI is detected in the differential image. [Embodiment 12] The X-ray imaging system according to Embodiment 11, wherein the mask image is a first cross-sectional image acquired by an X-ray imaging device, the contrast image is a cross-sectional image acquired after the mask image, the contrast image is acquired at a predetermined frequency and displayed on a display device together with the difference image. [Embodiment 13] The display of the difference image and contrast image is refreshed at a predetermined frequency until the contrast of the ROI in the difference image reaches a threshold contrast, which indicates that the contrast agent has reached the ROI, according to Embodiment 12 of the X-ray imaging system. [Embodiment 14] The X-ray imaging system according to Embodiment 11, wherein motion is automatically detected in the difference image by a machine learning model trained to detect motion in the difference image based on the presence of distortion. [Embodiment 15] The X-ray imaging system according to Embodiment 14, wherein the machine learning model is further trained to determine a new position of the ROI based on the difference image and to reposition the ROI within the contrast image when the display of the contrast image on the display device is updated. [Embodiment 16] A method for operating an X-ray imaging system, In response to the operation of the X-ray imaging system in automatic scan acquisition start mode, the steps include generating a difference image from the mask image and contrast image, and displaying the difference image and contrast image on a display device, The step of providing an option to switch the operation of the X-ray imaging system to an assist scan acquisition start mode when motion is detected in the difference image outside the region of interest (ROI), Upon receiving confirmation to switch the operation to the assist scan acquisition start mode, the system continuously updates the display of the difference image and contrast image, wherein updating the display includes repositioning the ROI to compensate for motion, and the system further comprises the steps described above. A method comprising the steps of: initiating a diagnostic scan acquisition in response to receiving a request to initiate a diagnostic scan acquisition. [Embodiment 17] The method according to embodiment 16, further comprising the step of stopping the acquisition of a diagnostic scan when a stop request is received from an operator or a machine learning model. [Embodiment 18] The method according to Embodiment 16, wherein ROI repositioning includes the step of updating a contrast image at the new position of the ROI between frames using a high-speed image alignment algorithm. [Embodiment 19] The method according to embodiment 16, wherein the location of the ROI is tracked in a contrast image based on image registration, and image registration is performed before the ROI is indicated in the contrast image. [Embodiment 20] The method according to Embodiment 16, wherein repositioning the ROI involves updating the position of the ROI in the contrast image frame by frame by a machine learning model, the machine learning model is configured to generate a vector map based on the difference image to determine the position of the ROI. [Explanation of symbols]

[0089] 100: CT system 102: Gantry 104: X-ray source 106: X-ray beam 108: Detector array 110: Image processing unit 112, 204: Subject 114: Table 200: Imaging system 202: Detector element 206: Rotation center 208: Control mechanism 210: X-ray controller 212: Gantry motor controller 214: DAS 216: Computing device 218: Mass storage device 220: Operator console 224: PACS 226: Table motor controller 230: Image reconstructor 232: Display 300: Scout image 302: Line 400: First cross-sectional image 402: ROI 404: Landmark 500: Second cross-sectional image 600: First difference image 700: Second difference image 702: Distortion 900: Real-time graph 902: Data points 904: Threshold contrast 906: Real-time graph 1000: Display screen

Claims

1. A method for operating an X-ray imaging system (100) that operates in either an automatic mode that can start scanning without operator input, or an assist mode that starts scanning upon input from the operator, A step of generating a difference image (600) from a first image (400) of the subject and a second image (500) of the subject collected after the first image (400), The steps include displaying the difference image (600) on the display device (232), Step (1114) of detecting the movement of the subject outside the region of interest (ROI) (402) based on the analysis of the difference image (600), When the aforementioned movement is detected, the operator is prompted to switch to the automatic mode and to the support mode. Includes, The first image (400) is obtained before the contrast agent is delivered to the subject. The second image (500) is obtained after the contrast agent has been delivered to the subject. The difference image (600) is displayed before the contrast agent reaches the ROI (402) (1122), in this method.

2. The method according to claim 1, wherein the difference image (600) is continuously updated (1112) in response to the acquisition of a new image which becomes the second image (500).

3. The method according to claim 1, wherein when the contrast agent reaches the ROI (402), the difference image (600) is no longer displayed (1128).

4. The method according to claim 1, wherein the first image is a mask image (400), the second image is a contrast image (500), both the first image and the second image are displayed on the display device (232), and the difference image (600) is generated by subtracting the first image from the second image.

5. If distortion (702) exists around the ROI (402) in the difference image (600), the motion is detected. The method according to claim 1, wherein the distortion (702) arises from a displacement of the anatomical features of the subject captured in the first image and the second image.

6. The method according to claim 2, comprising the step of displaying the new image.

7. The method according to claim 1, further comprising the steps of: presenting a notification on the display device (232) when the motion is detected (1118); and presenting an option to switch the operation of the X-ray imaging system (100) to the support mode.

8. The method according to claim 1, wherein when no motion is detected and the X-ray imaging system (100) is operating in the automatic mode, scanning is automatically started when it is determined that the contrast of the difference image (600) has reached a threshold.

9. The method according to claim 1, wherein the motion is detected using a machine learning model trained to identify the motion in the difference image (600), or by the operator visually observing the difference image (600).

10. The method according to claim 1, wherein a graph (900) plotting the contrast at the ROI (402) in the difference image (600) against time is also displayed on the display device (232).

11. An X-ray imaging system (100) that operates in either an automatic mode that can start scanning without operator input, or an assist mode that starts scanning upon input from the operator, An X-ray source (104) and an X-ray detector (108) are positioned on opposite sides of the gantry (102), A processor of a computing device (216) consisting of memory for storing executable instructions, Includes, The subject (204) is placed in the gantry between the X-ray source and the X-ray detector. When the aforementioned executable instruction is executed, the processor will: An X-ray imaging system that performs the method according to any one of claims 1 to 10.

12. The first image is a mask image (400), and the second image is a contrast image (500). The X-ray imaging system (100) according to claim 11, wherein the mask image is a first cross-sectional image (400) acquired by the X-ray imaging system, the contrast image (500) is a cross-sectional image acquired after the mask image, the contrast image is acquired at a predetermined frequency and displayed on the display device (232) together with the difference image (600).

13. The display of the difference image (600) and the contrast image (500) is refreshed according to the predetermined frequency until the contrast of the ROI (402) in the difference image (600) reaches a threshold contrast, and the threshold contrast indicates that the contrast agent has reached the ROI (402), according to claim 12, X-ray imaging system (100).

14. If distortion (702) exists around the ROI (402) in the difference image (600), the motion is detected. The distortion (702) arises from the misalignment of the anatomical features of the subject captured in the first image and the second image. The X-ray imaging system (100) according to claim 12, wherein the motion is automatically detected in the difference image (600) by a machine learning model trained to detect the motion in the difference image (600) based on the presence of the distortion.

15. The X-ray imaging system (100) according to claim 14, wherein the machine learning model is further learned to determine a new position of the ROI (402) based on the difference image (600) and to reposition the ROI (402) within the contrast image when the display of the contrast image (500) on the display device (232) is updated.

Images