CT motion compensation with slice-wise definition of reference state

Abstract

A method for processing three-dimensional (3D) projection data includes receiving 3D projection data and reconstructing initial axial slices from the 3D projection data to cover a total volume with each initial axial slice corresponding to a slice of the total volume. At least one initial axial slice contains a motion artifact. The method then retrieves a local reference state for estimating motion in the axial slice such that the motion artifact is minimized. Each local reference state is determined for a corresponding slice that contains the motion artifact across the total volume. The method then includes reconstructing the axial slice using motion compensation based on the local reference state of the axial slice to generate a corresponding compensated axial slice. A system is provided for implementing the method.

Description

CT MOTION COMPENSATION WITH SLICE-WISE DEFINITION OF REFERENCE STATEFIELD

[0001] The present invention generally relates to systems and methods for motion compensation in medical imaging. In particular, the invention relates to motion compensation in computed tomography (CT) images.BACKGROUND

[0002] Patient motion is a problem that frequently occurs in computed tomography (CT). Motion-compensated reconstruction (MCR) aims at reducing motion artifacts by using an estimate of the underlying motion during reconstruction, namely a back- projection process. In general, the accuracy of such a motion estimate will vary from case to case, and in the case of an error, the resulting MCR will have deteriorated quality and may be worse than providing the default reconstruction.

[0003] Typical MCR implementations are based on filtered back-projection (FBP) type algorithms, such as aperture weighted wedge reconstruction, which use a rebinning step to generate a so-called wedge geometry. Such an approach is illustrated in Figures 3A and 3B.

[0004] The computationally most demanding step in such a FBP reconstruction is the back-projection, which has two aspects, namely address calculation and weight calculation. The address calculation is needed to provide for a given voxel to reconstruct, and each given re-binned projection, the address on the projection where the correct filtered projection value is located. The weight calculation accounts for the contributions of all redundant rays to ensure that they properly sum up to 1. The computational burden of motion compensated helical cone-beam CT reconstruction is prohibitively high if the full motion vector field is used for both address and weightcalculation. In some existing approaches, an approximation is created to reduce the processing load in the calculations. For example, the motion vector field may be limited to take only motion in the z direction into account for the weight calculation.

[0005] The approximation in the weight calculation can lead to two effects. In some instances, a weight of 0 is obtained for a voxel that is inside the cone. This case compromises the noise level in the reconstructed image. Second, it may happen that a non-zero weight is obtained for a voxel that is outside the cone. In this case, extrapolated projection data must be provided to avoid artifacts. However, extrapolation is also not straightforward and can easily itself lead to artifacts.

[0006] Further, the image volume is reconstructed in a specific motion state that usually does not correspond to the motion state of the reconstruction without motion correction. In other words, the image volume reconstruction may be warped.Therefore, it is not straight-forward to assess the quality of the MCR by comparing complete images generated from default reconstructions to motion compensated reconstructions. Further, deformation may be expected with respect to some body parts, such as a user's chest and rib cage, and a physician or analyst may be able to evaluate such a warped presentation. However, for body parts not typically deformed, motion may result in an unexpected and difficult to evaluate shape. For example, a skull presented with motion-based deformations may be difficult to evaluate.

[0007] There is therefore a need for a system and method in which a user can evaluate directly the quality of a motion compensated reconstruction against a default reconstruction, and thereby confirm that motion compensation has improved quality. There is a further need for a system and method in which high quality motion compensation can be provided in a manner less computationally demanding than existing algorithms.SUMMARY

[0008] A computer implemented method is provided for processing three-dimensional (3D) projection data. The method includes receiving 3D projection data and reconstructing a plurality of initial axial slices from the 3D projection data to cover a total volume with each initial axial slice corresponding to a slice of the total volume. At least one initial axial slice contains a motion artifact.

[0009] The method then retrieves a local reference state for estimating motion in the at least one axial slice such that the motion artifact is minimized. Each local reference state is determined for a corresponding slice that contains the motion artifact across the total volume. The method then includes reconstructing the at least one axial slice using motion compensation based on the local reference state of the axial slice to generate a corresponding compensated axial slice.

[0010] In some embodiments, the method further includes assembling the reconstructed axial slices of the total volume into a first 3D image.

[0011] In some such embodiments, the initial axial slices are assembled into the first 3D image for presentation to the user as a default reconstruction.

[0012] In some embodiments in which the axial slices are assembled into a first 3D image, the method further includes replacing the at least one axial slice in the initial axial slices with the corresponding compensated axial slice and assembling a motion compensated second 3D image from the at least one compensated axial slice combined with the initial axial slices for any remaining slices of the total volume.

[0013] In some such embodiments, the method includes displaying both the first 3D image and the second 3D image to a user and receiving an indication of preference from the user.

[0014] In some embodiments in which certain axial slices are replaced with motion compensated axial slices, the motion compensated reconstruction of the at least one axial slice is a warped reconstruction of the at least one axial slice.

[0015] In some such embodiments, upon receipt of a selection by a user, the method presents one of the second 3D image assembled from a standard motion compensated reconstruction or a third 3D image assembled at least partially from the warped reconstruction.

[0016] In some embodiments in which the motion compensated reconstruction is a warped reconstruction, the warped reconstruction is based on the local reference state, a motion state associated with a time at which corresponding projection data was acquired, a motion state determined based on registration of the at least one axial slice with an immediately adjacent slice, or a pre-existing model of expected anatomy.

[0017] In some such embodiments, the method includes generating a deformation field based on the local reference state, and wherein the warped reconstruction is based on a resampling of the compensated axial slice based on the deformation field.

[0018] In some embodiments, the local reference state for the at least one axial slice is registered to an immediately adjacent axial slice or can be registered within the context of a subject of the projection data or corresponds to a known landmark or surface.

[0019] In some embodiments, the method includes associating a known displacement with the at least one axial slice based on the corresponding local reference state.

[0020] In some such embodiments, the known displacement is determined based on a registration algorithm, a machine learning algorithm, a segmentation based on a shape model, or segmentation based on identification of a landmark.

[0021] In some embodiments, a system is provided for processing three-dimensional (3D) projection data in computed tomography. The system includes a memory forstoring a plurality of instructions, a display for presenting images to a user, and at least one processor that couples with the memory and is configured to execute the instructions.

[0022] The processor is further configured to execute the instructions to receive 3D projection data and reconstruct a plurality of initial axial slices from the 3D projection data to cover a total volume, with each initial axial slice corresponding to a slice of the total volume. At least one initial axial slice contains a motion artifact.

[0023] The processor is further configured to execute the instructions to retrieve for each initial axial slice that contains motion artifacts, a local reference state for estimating a motion such that the motion artifact is minimized. Each local reference state is determined for a corresponding initial axial slice that contains the motion artifact across the total volume. The processor then executes the instructions to reconstruct the at least one initial axial slice using motion compensation based on the local reference state of the axial slice to generate a corresponding compensated axial slice.

[0024] The processor is further configured to execute the instructions to assemble the reconstructed axial slices of the total volume into a first 3D image for presentation to a user at the display.

[0025] In some such embodiments, the reconstructed axial slices assembled for presentation to the user include the corresponding compensated axial slice replacing the at least one slice of the initial axial slices.BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a schematic diagram of a system according to one embodiment of the present invention.

[0027] Figure 2 illustrates an exemplary imaging device according to one embodiment of the present invention.

[0028] Figure 3A is an illustration of a substantially cylindrical detector for use in an imaging device.

[0029] Figure 3B is a wedge geometry used in existing motion-compensated reconstruction methods.

[0030] Figure 4 schematically illustrates aspects of a method for motion compensation in accordance with this invention.

[0031] Figure 5 illustrates a method for motion compensation in accordance with this invention.

[0032] Figures 6A-6C illustrate default reconstructions for three slices of 3D projection data from a 3D image.

[0033] Figures 1 -1Q illustrate a standard motion compensated reconstruction for the three slices of 3D image of Figures 6A-6C.

[0034] Figures 8A-8C illustrate a motion compensated reconstruction for the three slices of 3D image of Figures 6A-6C in accordance with this invention.DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0035] The description of illustrative embodiments according to principles of the present invention is intended to be read in connection with the accompanying drawings, which are to be considered part of the entire written description. In the description of embodiments of the invention disclosed herein, any reference to direction or orientation is merely intended for convenience of description and is not intended in any way to limit the scope of the present invention. Relative terms such as "lower," "upper," "horizontal," "vertical," "above," "below," "up," "down," "top" and "bottom" as well as derivative thereof (e.g., "horizontally," "downwardly," "upwardly," etc.) should be construed torefer to the orientation as then described or as shown in the drawing under discussion. These relative terms are for convenience of description only and do not require that the apparatus be constructed or operated in a particular orientation unless explicitly indicated as such. Terms such as "attached," "affixed," "connected," "coupled," "interconnected," and similar refer to a relationship wherein structures are secured or attached to one another either directly or indirectly through intervening structures, as well as both movable or rigid attachments or relationships, unless expressly described otherwise. Moreover, the features and benefits of the invention are illustrated by reference to the exemplified embodiments. Accordingly, the invention expressly should not be limited to such exemplary embodiments illustrating some possible non-limiting combination of features that may exist alone or in other combinations of features; the scope of the invention being defined by the claims appended hereto.

[0036] This invention describes the best mode or modes of practicing the invention as presently contemplated. This description is not intended to be understood in a limiting sense, but provides an example of the invention presented solely for illustrative purposes by reference to the accompanying drawings to advise one of ordinary skill in the art of the advantages and construction of the invention. In the various views of the drawings, like reference characters designate like or similar parts.

[0037] It is important to note that the embodiments disclosed are only examples of the many advantageous uses of the innovative teachings herein. In general, statements made in the specification of the present application do not necessarily limit any of the various claimed inventions. Moreover, some statements may apply to some inventive features but not to others. In general, unless otherwise indicated, singular elements may be in plural and vice versa with no loss of generality.

[0038] Figure 1 is a schematic diagram of a system 100 according to one embodiment of the present invention. As shown, the system 100 typically includes a processing device 110 and an imaging device 120.

[0039] The processing device 110 may apply processing routines to images or measured data, such as projection data, received from the image device 120. The processing device 110 may include a memory 113 and processor circuitry 111. The memory 113 may store a plurality of instructions. The processor circuitry 111 may couple to the memory 113 and may be configured to execute the instructions. The instructions stored in the memory 113 may comprise processing routines, as well as data associated with processing routines, such as machine learning or deep learning algorithms, and various filters for processing images.

[0040] The processing device 110 may further include an input 115 and an output 117. The input 115 may receive information including three-dimensional images or measured data, such as three-dimensional projection data, from the imaging device 120. It is noted that while a CT imaging method is discussed herein, and while a CT scanning unit 200 is discussed in detail below, similar methods may be implemented using different forms of three-dimensional imaging, such as magnetic resonance imaging (MRI) or positron emission tomography (PET) imaging.

[0041] In some embodiments, the processing device 110 may relate to the imaging device 120 directly. In alternate embodiments, the processing device 110 may be distinct from the imaging device 120, such that the processing device 110 receives images or measured data for processing by way of a network or other interface at the input 115.

[0042] Figure 2 illustrates an exemplary imaging device 200 according to one embodiment of the present invention. It will be understood that while a CT imaging device 200 is shown, and the following discussion is generally in the context of CT images, similar methods may be applied in the context of other imaging devices, and images to which these methods may be applied may be acquired in a wide variety of ways.

[0043] In an imaging device 200 in accordance with embodiments of the present invention, the CT scanning unit may be adapted for performing one or multiple axial scans and / or a helical scan of an object in order to generate the 3D projection data. In an imaging device 200 in accordance with embodiments of the present invention, the CT scanning unit may comprise an energy-resolving photon counting or spectral dual-layer image detector. Spectral content may be acquired using other detector setups as well. The CT scanning unit may include a radiation source that emits radiation for traversing the object when acquiring the 3D projection data.

[0044] In the example shown in Figure 2, the CT scanning unit 200, e.g., the CT scanner, may include a stationary gantry 202 and a rotating gantry 204, which may be rotatably supported by the stationary gantry 202. The rotating gantry 204 may rotate about a longitudinal axis around an examination region 206 for the object when acquiring the 3D projection data. The CT scanning unit 200 may include a support 207 to support the patient in the examination region 206 and configured to pass the patient through the examination region during the imaging process.

[0045] The CT scanning unit 200 may include a radiation source 208, such as an X-ray tube, which may be supported by and configured to rotate with the rotating gantry 204. The radiation source 208 may include an anode and a cathode. A source voltage appliedacross the anode and the cathode may accelerate electrons from the cathode to the anode. The electron flow may provide a current flow from the cathode to the anode, such as to produce radiation for traversing the examination region 206.

[0046] The CT scanning unit 200 may comprise a detector 210. The detector 210 may subtend an angular arc opposite the examination region 206 relative to the radiation source 208. The detector 210 may include a one- or two-dimensional array of pixels, such as direct conversion detector pixels. The detector 210 may be adapted for detecting radiation traversing the examination region 206 and for generating a signal indicative of an energy thereof.

[0047] The CT scanning unit 200 may include generators 211 and 213. The generator 211 may generate 3D projection data 209 based on the signal from the detector 210. The generator 213 may receive the 3D projection data 209 and, in some embodiments, generate three-dimensional imaging data 311 of the object based on the tomographic projection data 209. In some embodiments, the 3D projection data 209 may be provided to the input 115 of the processing device 110, while in other embodiments the three-dimensional imaging data 311 is provided to the input of the processing device.

[0048] Figure 3A is an illustration of a substantially cylindrical detector 210, 300 for use in an imaging device 200. Figure 3B illustrates a wedge geometry 310 used in existing motion-compensated reconstruction methods. As shown, the data is acquired initially using a substantially cylindrically shaped (curved) detector 210, 300 which receives radiation from radiation sources 208. The 3D projection data from several source positions along the helical path is then re-sorted into a 3D projection data set in a fanshaped wedge geometry 310, where all rays 320 output from radiation source positions 208 are parallel to each other except for differences in the cone angle 330.

[0049] As discussed above, traditional approaches to motion compensated reconstruction (MCR) require address calculation for voxels to reconstruct and weight calculations during the so-called back-projection step, i.e., accumulating the correct filtered projection value to obtain the correct voxel value.

[0050] Figure 4 schematically illustrates aspects of a method for motion compensation in accordance with one embodiment of the present invention. Figure 5 illustrates a method for motion compensation in accordance with one embodiment of the present invention. The method is typically a computer implemented method for processing three-dimensional (3D) projection data in CT applications.

[0051] Generally, as shown in Figure 4, the method includes first retrieving 3D projection data 400 and generating a default reconstruction 410. In parallel, the method retrieves multiple short scans 450 from the 3D projection data 400, which can be reconstructed into initial axial slabs 440a, 440b, and 440c. A short scan refers usually to a 3D projection data set obtained from source positions traveling over 180° plus the fan-angle of the system. However, other ranges of more than 180° and less than 360° are considered as short scans in this invention. From each short scan segment, a slab of a few slices is reconstructed and slabs from adjacent short scans may overlap. Each initial slab 440a, 440b, 440c then corresponds to a slab of the total volume captured by the 3D projection data, and may be utilized for estimating corresponding motion.

[0052] When discussing axial slices of the entire volume, each slice is a substantially planar segment of the overall volume. In the embodiments discussed, where slabs 440a, 440b, 440c are used as registration input, each slab has a known thickness (i.e., it contains a known number of slices) determined based on a cone angle of the corresponding 3D projection data and reconstruction parameters.

[0053] The initial slabs 440a, 440b, 440c representing slices of the total volume may then contain some motion artifacts. They may also suffer from motion in the sense that even though they represent the same image location from the examination region 206, they may show different parts of the patient since the patient may have moved during the data acquisition. To account for such artifacts, each slab may be overlayed on an adjacent slab and may then be registered to the adjacent slab. Accordingly, the second initial slab 440b may be overlayed and registered relative to the first initial slab 440a. The third initial slab 440c may then be overlayed and registered relative to the second initial slab 440b.

[0054] The registration process provides a motion vector field that can be used for a motion compensated reconstruction. Yet for a motion compensated reconstruction, it is required to define a reference state which is the time at which the patient's position should be reconstructed. In the related-art motion compensated reconstruction, this reference state is a global state. However, in the present invention, a local reference state is retrieved for estimating motion for one or some of the initial axial slices within the initial slabs and for generating corresponding slice-wise normalized motion vector fields. The corresponding axial slices may be reconstructed or modified based on the corresponding local reference states to generate corresponding motion compensated axial slices, which may then be used to generate a motion compensated 3D image 420 with the same overall geometry as the default reconstruction 410. As discussed in more detail below, the motion compensated image 420 may be further processed to warp the resulting 3D image to better conform with anatomically plausible shapes for the total volume 430.

[0055] As shown, the method includes first receiving (at 500), at the processing device 110, the 3D projection data 209, 400.

[0056] In a typical CT implementation, the 3D projection data 400 is used to reconstruct axial slices which are then assembled into a first 3D image 410. However, the 3D projection data 400 is subject to motion, and the resulting first 3D image 410 is therefore subject to motion, which can result in artifacts like distortion, ghosting, and blurring. Such motion may be a result of a subject who moved during scanning, for example.

[0057] The method proceeds by reconstructing (510) a plurality of initial slabs 440a, 440b, 440c from the 3D projection data 400. This may be by way of multiple short scans, as shown in Figure 4. The plurality of initial slabs 440a, 440b, 440c are combined to cover a total volume corresponding to the 3D image being processed, and each initial axial slice corresponds to a slice of the entire volume.

[0058] The proposed method may be applied if at least one slice from the plurality of initial slabs contains a motion artifact (520) or if they show differences indicative of motion during the acquisition. In a typical implementation, multiple slices of the initial axial slices of the total volume, or even all slices of the total volume, would be identified as containing motion artifacts (at 520).

[0059] Any axial slices having motion artifacts may then be further processed to minimize the impact of those motion artifacts. In doing so, each axial slice may be provided with a motion estimation by registering the corresponding slab containing the axial slice to an immediately adjacent slab, as discussed above.

[0060] Accordingly, for any axial slice identified (at 520) as containing motion artifacts, the method proceeds to provide (530) a local reference state. In some embodiments, all axial slices of the plurality of initial axial slices may be determined to have motion artifacts, in which case all axial slices will be provided (at 530) with local reference states.

[0061] In some embodiments, the method will determine, for each axial slice, if the motion artifacts are sufficiently severe to require motion compensation. As such, in some embodiments, thresholds may be utilized to determine whether local reference states and further reconstruction should be utilized.

[0062] So long as additional axial slices of the entire volume remain (540), the method continues to identify such axial slices (at 520) and provide (at 530) corresponding local reference states such that each local reference state is determined for a corresponding axial slice.

[0063] The local reference state may be derived and provided (at 530) in various ways. As discussed above, the local reference states may be extracted from a result of a registration process in which each axial slice is overlayed and registered relative to an adjacent axial slice. This allows for the consecutive registration of each slab 440a, 440b, 440 c. The local reference state may be the local (i.e ., determined per slice) average motion state obtained from the registration process.

[0064] Alternatively, or in addition, the method may rely on known landmarks or surfaces of the anatomy of the subject. Accordingly, the method may use a pre-existing model of a subject's expected anatomy, or may be provided with a model based on earlier or alternative imaging for the subject. In this way, the local reference state may be based, for example, on an expected location of a surface of a subject's head or skull.

[0065] The motion compensation is based on a known displacement and / or rotation associated with each axial slice identified (at 520) as containing motion artifacts. The known displacement is then determined based on the corresponding local reference state (provided at 530). The known displacement may then be determined based on a registration algorithm, a machine learning algorithm, a segmentation based on a shape model, or a segmentation based on identification of a landmark.

[0066] Once each axial slice having motion artifacts is provided with a corresponding local reference state (at 530) and a motion vector field that transforms the actual geometry of the axial slice to the geometry of the local reference state, the method proceeds to reconstruct (550) each corresponding slice using motion compensation based on the corresponding local reference state in order to generate (560) a corresponding compensated axial slice.

[0067] The method may then proceed to assemble the reconstructed axial slices of the entire volume into the first 3D image 410 (570). In some embodiments, this involves assembling the plurality of initial axial slices (generated at 510) into the first 3D image 410 for presentation to the user (at 580) as a default reconstruction 410. Such an initial presentation is optional, and in many embodiments, the initial presentation to a user may be following the assembly of both the first and second 3D images 410, 420.

[0068] In some embodiments, the method may first replace (590) any slices in the initial axial slices determined to have motion artifacts with the corresponding motion compensated axial slice. The method may then assemble a motion compensated second 3D image 420 (600) from the compensated axial slice, or slices (generated at 560), combined with the initial axial slices (generated at 510) for any remaining axial slices of the entire volume.

[0069] As noted above, in some embodiments, all axial slices of the initial axial slices may be determined to have motion artifacts (at 520), and as such, in preparing the second 3D image 420 (600), it is understood that all initial axial slices may have been replaced by corresponding compensated axial slices (generated at 560).

[0070] It is noted that while the first 3D image in this description (assembled at 570), shown at 410 in FIG. 4, refers to the default reconstruction image while the second 3D image (assembled at 600), shown as 420 in FIG. 4, refers to the motion compensatedreconstruction (MCR) image, in some embodiments, the method may initially prepare just the MCR image. Accordingly, the method may assemble the MCR image 420 as the first 3D image (at 570) and present it to the user.

[0071] In some embodiments, where both described images 410, 420 are assembled, the method may proceed to display (610) both the first 3D image 410 and the second 3D image 420 to the user, such that the user can determine which image is preferred. Accordingly, the user may review the second 3D image 420 in order to determine if motion artifacts present in the first 3D image 410 were indeed removed. Such images 410, 420 may be presented side by side, or in a display that toggles between the two images, as is discussed in more detail below with respect to Figures 6A-8C.

[0072] The method may then receive an indication of a preference (620) from the user and may then proceed with either the first 3D image 410 or the second 3D image 420.

[0073] Typically, where motion artifacts exist in a number of slices from the initial image scans, the first 3D image 410 contains deformations, as visible in Figure 4 (where a straight vertical hole is present in the physical phantom, which shows up as a curved hole), such that the overall assembled image does not conform to an expected shape of the anatomy of the subject of the 3D images.

[0074] The second 3D image 420 may be an MCR image, in that the motion compensation eliminates motion artifacts. However, particularly in cases with strong motion, the choice of a locally varying reference state retains the deformation of the subject itself, which may not be desirable, i.e., the hole in the phantom shows up as a clear hole, but still it is not straight). In the example shown, the first 3D image 410 and the second 3D image 420 each show a similar deformed skull shape despite the fact that most motion artifacts have been eliminated in the second 3D image 420.

[0075] Accordingly, in some embodiments, the method may present a user with an additional option (630) of warping the resulting MCR image 420. Alternatively, the MCR itself may utilize a warped reconstruction by default. Accordingly, when generating the motion compensated axial slices (at 560), the reconstruction of the corresponding axial slice may be warped to better conform to the subject's actual anatomy.

[0076] The warped reconstruction of each axial slice may then be based on, for example, the local reference state itself, a determined motion state associated with a time at which the corresponding projection data was acquired, or a pre-existing model of the expected anatomy of the subject.

[0077] In some embodiments, rather than applying this type of deformation correction to the individual axial slices during the motion compensated reconstruction (at 560), the warping is instead performed during or following assembly of the second 3D image 420. Accordingly, the method may generate a deformation field (640) based on the local reference states. The method may then proceed with resampling (650) of the second 3D image 420 based on the deformation field, resulting in a third 3D image or set of 3D images 430 corresponding to anatomically plausible shapes from which the user can select a preference. The resulting third 3D image 430 may then be displayed (660) to the user as the final assembled 3D image.

[0078] The selection of a warped reconstruction may be determined, in some embodiments, on the specific anatomy being evaluated. Accordingly, for an anatomy for which deformation is standard and well understood and recognized by physicians, such as a subject's chest, warping may not be necessary or desirable. However, for anatomy that is typically rigid, such as a subject's head, warping may be more desirable.

[0079] Figures 6A-6C illustrate default reconstructions 700a, 700b, 700c for three axial slices of 3D projection data from a 3D image 410, in this case, slices of a skull of aphysical phantom. As shown, the default reconstructions 700a, 700b, 700c each correspond to a single axial slice of a total volume, and each axial slice contains a motion artifact, typically blurring and ghosting of edges, identified by arrows 710a, 710 b, 710c.

[0080] In addition to the discrete motion artifacts, the shape of the subject's skull is distorted, as a result of the motion that generated the motion artifacts. Accordingly, in the context of the method of Figure 5, the three axial slices shown would be identified as containing motion artifacts (at 520), provided with local reference states (at 530), and reconstructed into compensated axial slices (at 560)

[0081] Figures 1 -1Q illustrate standard (related art) motion compensated reconstructions 720a, 720b, 720c for the three axial slices of 3D projection data shown in the default reconstructions 700a, 700b, 700c of Figures 6A-6C. As shown, the discrete motion artifacts 710a, 710b, 710c of the default reconstructions 700a, 700b, 700c have been partially eliminated at the corresponding locations 730a, 730b, 730c of the motion compensated reconstructions 720a, 720b, 720c. However, the changes cannot be safely attributed to the motion compensation because the default MCR reconstructs the image in a different motion state than the default recon.

[0082] According to the embodiments of the present invention, Figures 8A-8C illustrate the motion compensated reconstructions 800a, 800b, 800c for the three slices of 3D image shown in the default reconstructions 700a, 700b, c of Figures 6A-6C. As shown, in addition to eliminating motion artifacts 710a, 710b, 710c, such as blurring and ghosting, the image shows - unlike the default MCR - the same anatomical features as the default recon. Thus, the proposed method allows to attribute differences between images 700a and 800a solely to the effectiveness of the MCR. The proposed method also improves overall sharpness of the image, as is most apparent in Figure. 8B.

[0083] As noted above, the various reconstructions can be presented to users in various ways. In some embodiments, a default reconstruction, such as 700a shown in Figure 6A, may be presented to a user as a slice of a first 3D image 410 alongside a proposed MCR based reconstruction 800a for a corresponding slice of the second or third 3D image 420, 430. Alternatively, the images 700a, 800a may be toggled in a user interface to give a user a clear view of the differences between the two.

[0084] In some embodiments, if the MCR based image presented to the user is the second 3D image 420, and is therefore distorted, and the user chooses to utilize the MCR based reconstruction, the method may proceed to present the user with the axial slice of the MCR based reconstruction 800a alongside a corresponding axial slice of a default reconstruction 700a. This may similarly be presented in a user interface that allows the user to toggle between the two proposals. In this manner, a user may choose to use or reject any proposed modification to the reconstruction based on motion compensation on a case-by-case basis.

[0085] The methods according to the present invention may be implemented on a computer as a computer implemented method, or in dedicated hardware, or in a combination of both. Executable code for a method according to the present invention may be stored on a computer program product. Examples of computer program products include memory devices, optical storage devices, integrated circuits, servers, online software, etc. Preferably, the computer program product may include non- transitory program code stored on a computer readable medium for performing a method according to the present invention when said program product is executed on a computer. In an embodiment, the computer program may include computer program code adapted to perform all the steps of a method according to the present inventionwhen the computer program is run on a computer. The computer program may be embodied on a computer readable medium.

[0086] While the present invention has been described at some length and with some particularity with respect to the several described embodiments, it is not intended that it should be limited to any such particulars or embodiments or any particular embodiment, but it is to be construed with references to the appended claims so as to provide the broadest possible interpretation of such claims in view of the prior art and, therefore, to effectively encompass the intended scope of the invention.

[0087] All examples and conditional language recited herein are intended for pedagogical purposes to aid the reader in understanding the principles of the invention and the concepts contributed by the inventor to furthering the art, and are to be construed as being without limitation to such specifically recited examples and conditions. Moreover, all statements herein reciting principles, aspects, and embodiments of the invention, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future, i.e., any elements developed that perform the same function, regardless of structure.

Claims

What is claimed is:

1. A computer implemented method for processing three-dimensional (3D) projection data, comprising: receiving 3D projection data; reconstructing a plurality of initial axial slices from the 3D projection data to cover a total volume, each initial axial slice corresponding to a slice of the total volume, wherein at least one initial axial slice contains a motion artifact; retrieving a local reference state for estimating motion in the at least one axial slice such that the motion artifact is minimized, wherein each local reference state is determined for a corresponding slice that contains the motion artifact across the total volume; and reconstructing the at least one axial slice using motion compensation based on the local reference state of the axial slice to generate a corresponding compensated axial slice.

2. The computer implemented method of claim 1, further comprising assembling the reconstructed axial slices of the total volume into a first 3D image.

3. The computer implemented method of claim 2, wherein the initial axial slices are assembled into the first 3D image for presentation to the user as a default reconstruction.

4. The computer implemented method of claim 2, further comprising replacing the at least one axial slice in the initial axial slices with the corresponding compensated axial slice and assembling a motion compensated second 3D image from the at least one compensated axial slice combined with the initial axial slices for any remaining slices of the total volume.

5. The computer implemented method of claim 4, further comprising displaying both the first 3D image and the second 3D image to a user and receiving an indication of preference from a user.

6. The computer implemented method of claim 4, wherein the motion compensated reconstruction of the at least one axial slice is a warped reconstruction of the at least one axial slice.

7. The computer implemented method of claim 6, wherein, upon receipt of a selection by a user, the method presents one of the second 3D image assembled from a standard motion compensated reconstruction or a third 3D image assembled at least partially from the warped reconstruction.

8. The computer implemented method of claim 6, wherein the warped reconstruction is based on the local reference state, a motion state associated with a time at which corresponding projection data was acquired, a motion state determined based on registration of the at least one axial slice with an immediately adjacent slice, or a preexisting model of expected anatomy.

9. The computer implemented method of claim 8, further comprising generating a deformation field based on the local reference state, and wherein the warped reconstruction is based on a resampling of the compensated axial slice based on the deformation field.

10. The method of claim 1, wherein the local reference state for the at least one axial slice is registered to an immediately adjacent axial slice or can be registered within the context of a subject of the projection data or corresponds to a known landmark or surface.

11. The method of claim 1, further comprising associating a known displacement with the at least one axial slice based on the corresponding local reference state.

12. The method of claim 11, wherein the known displacement is determined based on a registration algorithm, a machine learning algorithm, a segmentation based on a shape model, or segmentation based on identification of a landmark.

13. A system for processing three-dimensional (3D) projection data in computed tomography comprising: a memory for storing a plurality of instructions; and at least one processor coupled to the memory and configured to execute the instructions to: receive 3D projection data; reconstruct a plurality of initial axial slices from the 3D projection data to cover a total volume, each initial axial slice corresponding to a slice of the total volume, and wherein at least one axial slice contains a motion artifact; retrieve a local reference state for estimating a motion in the at least one axial slice such that the motion artifact is minimized, wherein each local reference state is determined for a corresponding axial slice that contains the motion artifact across the total volume; reconstruct the at least one axial slice using motion compensation based on the local reference state of the axial slice to generate a corresponding compensated axial slice; and assemble the reconstructed axial slices of the total volume into a first 3D image for presentation to a user at the display.

14. The system of claim 13, wherein the reconstructed axial slices assembled for presentation to the user include the corresponding compensated axial slice replacing the at least one slice of the initial axial slices.

15. A non-transitory computer-readable medium for storing executable instructions, which cause a method to be performed to process three-dimensional (3D) projection data, the method comprising: receiving 3D projection data; reconstructing a plurality of initial axial slices from the 3D projection data to cover a total volume, each initial axial slice corresponding to a slice of the total volume, wherein at least one initial axial slice contains a motion artifact; retrieving a local reference state for estimating motion in the at least one axial slice such that the motion artifact is minimized, wherein each local reference state is determined for a corresponding slice that contains the motion artifact across the total volume; and reconstructing the at least one axial slice using motion compensation based on the local reference state of the axial slice to generate a corresponding compensated axial slice.

Images