2D and 3D tomographic X-ray imaging using flat panel detectors

[0002] The present invention relates to the art of tomographic imaging. It finds more particular application in 3D and 2D tomographic imaging based on one or more layers in a human body being measured by an x-ray system. However, it will be appreciated that the invention has other applications such as imaging of an object with other types of penetrating radiation to determine the make up and constituent components of that object and it may be advantageously employed in other environments and applications.

[0003] Traditionally, x-rays are projected from an x-ray source and through an object for exposing an image onto x-ray film. The image is typically a compression of all of the structures in the imaged portion of the body exposed onto the x-ray film. The vast amount of superimposed information can make diagnosis difficult.

[0004] In one technique for emphasizing a selected plane of the object, the x-ray source and the x-ray film move contra-cyclically in parallel planes which are parallel to the plane of interest. More specifically, the x-ray source and the x-ray film are moved such that the ray which exposes each incremental element of the x-ray film pivots about a fixed point in the plane of interest or focal plane. In this manner, each incremental element of the x-ray film is consistently exposed by radiation passing through a constant point in the focal plane. However, because the ray is pivoting about the constant point, the same increment of x-ray film is also exposed by a variety of structures outside the focal plane. In this manner, information from structures outside of the focal plane become blurred on the same x-ray film on which image data attributable to structures on the focal plane remains sharp and crisp.

[0005] If the diagnostician wishes to view another selected plane, the position of the patient is shifted relative to the x-ray source and the x-ray film canister such that the focal plane passes through a different selected plane of the object. With sufficient time, a series of parallel planer film images can be generated through the object. However, because each image takes a significant amount of time to generate, there is a significant temporal offset or time evolution among the images.

[0006] The temporal offset can be greatly reduced with volumetric CT scanners. However, CT scanners are not only much larger and more complex then the traditional x-ray film imaging systems, they are also much more expensive. Moreover, CT scanners are a different piece of equipment often positioned in a different room or location within the facility and in some instances at a different facility completely. Typically, the CT scanners are operated by different technicians. Accordingly, there are time delays in transporting the patient from one piece of equipment to the other, sometimes involving the making of additional appointments. Thus, the film images and the CT images may have very large temporal offsets. Moreover, it is difficult to maintain the position and orientation of the patient constant during such transporting. Accordingly, the resultant CT images are often not of the same physiology as the film images.

[0007] The present invention contemplates a new and improved apparatus and method which overcomes the above-referenced problems and others.

[0008] In accordance with the present invention, a device for imaging an object generates a multi-dimensional image of the object. The device includes a movable light source for sending a radiation beam through the object. A detector detects the radiation beam after it has passed through the object and generates electronic data representative of the detected radiation intensity. A system non-linearly displaces the radiation source relative to the object during the imaging. A processor receives, sorts, and processes the data to generate a series of stacked planar images of the object, each of which has minimized blurring in a corresponding plane and enhanced blurring outside of the corresponding plane.

[0009] In accordance with a more limited aspect of the present invention, the x-ray source moves in an arc to minimize magnification distortion by aligning the center of the cone beam with the region of interest.

[0010] In accordance with another more limited aspect of the present invention, the radiation source moves transversely to and along the arc to enhance off-plane blurring.

[0011] In accordance with another more limited aspect of the present invention, the radiation source is movably mounted below the object and a flat panel detector is located above the object in a stationary position.

[0012] In accordance with a more limited aspect of the present invention, the radiation source and the detector are located on opposite sides of an object support. The gantry moves the x-ray source up and down and left to right in relation to a flat panel detector.

[0013] A first advantage of the present invention is that it concurrently generates data for processing a plurality of layers.

[0014] Another advantage of the present invention resides in a reconstruction technique which quickly and easily reconstructs 2D and 3D tomographic images.

[0015] Another advantage is that electronic image representations are generated.

[0016] Yet another advantage is that the imaged planes can be changed after the data collection process is complete.

[0017] Still further advantages of the present invention will become apparent to those of ordinary skill in the art upon reading and understanding the following detailed description of the preferred embodiments.

[0018] The invention may take form in various components and arrangements of components, and in various steps and arrangements of steps. The drawings are only for purposes of illustrating a preferred embodiment and are not to be construed as limiting the invention.

[0019] FIG. 1 illustrates a diagrammatic side-view of a first embodiment of the present invention;

[0020] FIG. 2A is a perspective view illustrating the motion of an x-ray source;

[0021] FIG. 2B illustrates a top view of an x-ray source track of FIG. 2A;

[0022] FIG. 2C illustrates a cosine or sine radiation source motion path;

[0023] FIG. 3 is a diagrammatic illustration of a second embodiment of the invention;

[0024] FIG. 4 is a pictorial representation illustrating a reconstruction algorithm used when the radiation source moves in 2 Dimensions;

[0025] FIG. 5 is a pictorial representation illustrating a reconstruction algorithm used when the radiation source moves in 3 Dimensions;

[0026] FIG. 6 shows a side-view of a tomographic imaging system according to a third embodiment of the invention;

[0027] FIG. 7 shows an end view of the third embodiment of the present invention with the x-ray source extended;

[0028] FIG. 8 shows the movement of the x-ray source according to the third embodiment of the present invention;

[0029] FIG. 9 shows a flow chart for determining and generating 2D and 3D tomographic images according to the several embodiments of the invention.

[0030] With reference to FIG. 1, an object 10, preferably a subject or patient, is supported on an x-ray table 12 or other support surface. An x-ray tube or other radiation source 14 is mounted below the table 12 on an x-ray source moving system 16. The x-ray source moving system 16 moves the x-ray source 14 back and forth along a arc of constant radius with respect to the object 10. A flat panel, solid state radiation detector 18 is disposed above the table 12 in alignment with an x-ray beam projected by the x-ray source 14. In the preferred embodiment, the flat paneled detector 18 is a rectangular grid of radiation sensitive elements, each of which generates an electrical signal indicative of an intensity or amount of received radiation.

[0031] As a radiation source 14 moves below the object 10, the elements of the flat panel detector 18 are periodically sampled. More specifically, an x-ray source position detector 20 monitors an angular position φ of the x-ray source 14 by monitoring its physical position along the x-ray moving system 16. The flat paneled detector 18 is read out at each of a plurality of preselected angular positions as the x-ray source 14 moves back and forth along the arc. The x-ray source 14 can be stepped incrementally for optimal precision. Alternatively, the x-ray source 14 can be moved continually with sampling occurring over a short arc with minimal image degradation.

[0032] A sorting routine 22 receives the sampled period of image data or view at each of the angular positions and stores it in a corresponding view memory 24a, 24b, . . . , 24n. In the preferred embodiment in which the x-ray source moves back and forth a plurality of times, a series of view adders 26a, 26b, . . . , 26nadd each new view from a corresponding angular position to the views already in the corresponding view memory in order to sum or average the plurality of collected views.

[0033] As described in greater detail below, a processor 28 processes views from the view memories 24a, 24b, . . . , 24n. As explained in greater detail below, each of the views is offset and summed with preceded views in such a manner that radiation attenuation attributable to portions of the object in a first focal plane L₁are coherent and have minimal blurring while radiation attenuation from portions of the object 10 outside of the first focal plane L₁have maximum blurring. In this manner, a first slice of volumetric image data is generated and stored in a volumetric image memory 30. Changing the offset and summing as described in detail below, data representative of a second focal plane L₂and plurality of subsequent focal planes L_nare generated and stored in the volumetric image memory 30. It is to be appreciated, that as long as the views are stored in the view memories 24a, 24b, . . . , 24nthe offsets can be reselected to select a different array of focal planes L₁, L₂, . . . , L_nwhich might better illustrate a region of interest.

[0034] An operator controls an image selection processor 32 which selects the image to be displayed. Typical images include a single slice through a region of interest, coordinated sagittal, coronal, and transverse slices through the region of interest, oblique slices, volumetric renderings, and the like. The image selection processor 32 causes the appropriate values to be read out of the volumetric image memory 30 to a video processor 34 which converts the data into appropriate format for display on a video monitor or other human readable display 36.

[0035] With reference to FIGS. 2A, 2B, and 2C, the under-table x-ray tube 14 moves along the arcuate trajectory with a motion that combines two degrees of freedom. In particular, the motion of the x-ray source 14 has components which are both parallel and transverse to a longitudinal-vertical plane, but symmetrical through the object as the overhead flat panel 18 remains stationary. To this end, the x-ray source movement system 16 includes a transverse slide bar or support 40 which is mounted on its edge to a pair of similar rails 42 which extend along the sides of the arcuate path. The x-ray source moving system 16 includes a first drive, such as a motor 44, and drive train for moving the slide bar 40 along the track 42. As indicated above, the motor 44 can be a stepper motor for moving the x-ray source 14 in fixed increments or a synchronous or servo motor for moving it at a controlled speed. The encoder 20 for indicating the angular position of the source 14 can conveniently be a tachometer. The slide bar 40 carries a second drive (not shown) for moving the x-ray source 14 transversely relative to the vertical-longitudinally plane.

[0036] A motor control controls the two drives such that the radiation source 14 follows a preselected trajectory, such as the sinusoidal trajectory illustrated in FIG. 2C. Other, more complex motions are also contemplated such as spiral, circular, elliptical, or hypcycloidial. In the illustrated sinusoidal embodiment, the radiation source trajectory has a width W and a sampling interval S, both of which are controllable.

[0037] During the tomographic procedure, a set of digital photo-spot or pulsed fluoroscopic images or views is acquired and stored. The image set is then processed by the 3D tomographic image reconstruction processor 28 to create 2D tomographic images which are stacked to form the 3D tomographic image. The magnification of the imaging geometry of the system is not constant so a distortion correction of the varied magnification is employed to suppress the image blurring and to increase the image visibility. This is described hereinafter with reference to FIG. 9 below.

[0038] It is to be appreciated that the system can also be configured for conventional radiographic and fluoroscopic applications. During the radiographic applications, the x-ray tube 14 is positioned in the middle of the track 42.

[0039] The use of the flat panel detector 18 offers higher dynamic range than that of the conventional screen film. During one tomographic procedure, a set of digital flat panel images is acquired. From this set of flat panel images, multi-layer tomographic images of the anatomy of the object 10 are constructed. A 3D tomographic image is constructed from this same set of images. The processor 28 uses digital processing for image enhancement and distortion correction to increase image contrast and visibility. The control system is simple. The two motors, one for the motion along the track and one for moving the x-ray source along the slide bar 40, along the trajectory of FIG. 2C. This design makes the motion control easier and reliable.

[0040] Turning now to the second embodiment shown in FIGS. 3-5, x-ray source 14 is mounted overhead and the flat panel detector 18 is mounted below the object 10 disposed on the x-ray table 12. The x-ray source 14 moves along transverse and longitudinal axes relative to the object 10.

[0041] With reference to FIGS. 3 and 4, a digital tomographic system with a fixed magnification is illustrated. This system has the simplest reconstruction algorithm since there is no vertical motion of the source in this example. The flat panel detector 18 is located under the object 10 and is stationary during the tomographic procedure. An overhead x-ray tube 14 moves across the region of interest of a patient 10. In the illustrated diagnostic procedure, the tube 14 moves parallel to the flat panel detector 18 such that the magnification remains constant. The processor 28 integrates the views to create a tomographic image of a selected focal plane L. The x-ray source 14 is kept at a constant distance k from the detector plane 18, where the focal plane L is at a constant distance (from the detector plane 18. To reconstruct an image of layer L, an initial x-ray source position (X_o, Y_o) is selected as a reference position. The ray from point (X_o, Y_o) through a selected element O in the focal plane L of the object 10 intersects the flat detector 18 at (x_o, y_o). When the x-ray tube 14 moves from position (X_o, Y_o) to position (X_i, Y_i) the ray through element O in focal plane L, intersects the detector 18 at (x_i, y_i). The displacement D between the position (x_o, y_o) and (x_i, y_i) is calculated by:

[0042] x_i=x_o+(l/(k−l))*(X_o−X_i)  (1a)

[0043] y_i=y_o+(l/(k−l))*(Y_o−Y_i)  (1b).

[0044] It is to be appreciated that an incremental element O which lies along the ray from (X_o, Y_o) to (x_o, y_o) also lies along the ray between (X_i, Y_i) to (x_i, y_i). Further, different incremental elements lie along these two rays. As the x-ray source 14 moves to additional positions, the only common incremental element of the additional rays is still the incremental element O. Although illustrated with respect to incremental element O, all incremental elements in the plane L are effected likewise. It will further be observed that those incremental elements such as O′, which do not lie on plane L change their pixel location, i.e., the pixel which represents the radiation attenuation through such incremental element, in each view. In this manner, the contribution of the out of focal plane incremental elements to the summed image become blurred or homogenous with sufficient blurring, the out of plane incremental elements O′ are reduced to a background noise, preferably merely a haze. Optionally, this haze can be reduced through filtering, such as reducing all gray scale values below a preselected or calculated level to a base background level, i.e., black.

[0045] The (x_i, y_i) view is shifted by vector D and added to the (x_o, y_o) view. This process is repeated for a multiplicity of x-ray source positions to build the image of the first focal plane L and the image of the object O can be enhanced by adding two images together. The images of any object point on layer L can be enhanced using the same method.

[0046] With reference to FIG. 5, a more general case in which the magnification is not constant is shown. The x-ray source 14 moves from the initial or reference position (X_o, Y_o, Z_o) from which an image falls at (x_o, y_o, 0) on detector 18 after passing through the incremental element O₁, of focal plane L₁, while the distance l₁between the flat panel detector 18 and the reconstruction layer L₁remains constant. The x-ray source 14 moves from the initial or reference position (X_o, Y_o, Z_o) to a first incremental position (X_i, Y_i, Z_i) from which an image falls at (x_i, y_i, 0) on detector 18 after passing through the incremental element O₁, of focal plane L₁, while the distance l₁between the flat panel detector 18 and the reconstruction layer L₁remains constant. When the x-ray source 14 moves along an arc of constant radius the z-component changes. The distance l₁is held constant and the detector plane is arbitrarily set as z_o=z_i=0. Equations (1a) and (1b) become in their dimensions:

[0047] x_i=x_o+(l₁/(Z_o−l₁))(X_o−X_i)  (2a)

[0048] y_i=y_o+(l₁/(Z_o−l₁))(Y_o−Y_i)  (2b)

[0049] z_i=0  (2c).

[0050] The position (x_i, y_i, 0) is calculated for each point of the focal plane L₁, as:

[0051] x_i=[Z_i(Z_o−l₁)/Z_o(Z_i−l₁)]x_o+[(Z_il₁/Z_o(Z_i−l₁)]*X_o−[l₁/(Z_i−l₁)]X_i  (3a)

[0052] y_i=[Z_i(Z_o−l₁)/Z_o(_i−l₁)]y_o+[(Z_il₁/Z_o(Z_i−l₁)]*Y_o−[l₁/(Z_i−l₁)]Y_i  (3b)

[0053] z_i=0  (3c)

[0054] It should be appreciated that the ray from the x-ray source 14 at (X_o, Y_o, Z_o) falling on a corresponding point (x_o, y_o, 0) on the flat panel detector 18 passes through elements on a plurality of planes, including element O₂of plane L₂. The ray from the x-ray source 14 at point (X_i, Y_i, Z_i) passes through the element O₂and strikes the detector plane 18 at a different point offset by the vector D₂from point (x_o, y_o, 0) By simply adjusting the offset vector D₂, the same series of views can be summed to generate images in each of a multiplicity of focal planes.

[0055] During a single volumetric tomographic imaging procedure, a single set of the digital views is collected in the view memories 24a, 24b, . . . , 24n. For a first plane L₁which is a distance l₁from the flat plate detector 18, the data value of each view is adjusted in accordance with Equations (3a) and (3b) to correct for the offset and magnification and the views are summed. Then, for a second focal plane, L₂a distance l₂from the detector plane 18, the same adjustments are again made and these are summed to generate a slice image for a second plane L₂. This same process of adjusting the magnification and offset adjustments to the same data are repeated until each of the planes of the volumetric image representation are generated. These steps can be repeated with smaller interfocal plane spacings to get greater detail in a separate region of particular interest.

[0056] During a volumetric tomographic imagery procedure, a set of digital views is acquired. From this set of flat panel views, multi-layer tomographic images of anatomy are constructed. A 3D or volumetric tomographic image is constructed by stacking additional slice images reconstructed from the same set of tomographic views. Digital processing is used for image enhancement and distortion correction to increase the image contrast and visibility.

[0057] Turning now to FIGS. 6 and 7, a third embodiment is shown where an x-ray tomographic system which has the x-ray source 14 attached to a rotatable gantry 50, which is, in turn, attached to a translating gantry 52. X-rays from the source 14 pass through the object 10 located on the x-ray table 12. The x-ray beam is detected by the flat panel detector 18 positioned below the table 12.

[0058] With reference to FIG. 8, the gantry 50 is rotated to align the flat panel detector 18 automatically with the overhead x-ray tube 14 during radiographic procedures. During the tomographic procedure, the overhead x-ray tube 14 is driven by servo or stepper motors (not shown) which allows it to scan relative to a region of interest of the object. The flat panel detector 18 remains stationary. The translation gantry 52 moves left and right while the rotational gantry 50 rotates counterclockwise and clockwise with respect to the translational gantry 52 to keep the collimated x-ray cone beam aligned with the region of interest.

[0059] With continuing reference to FIGS. 6-7, the rotating gantry 50 and the longitudinally translating gantry 52 are interconnected such that the x-ray tube 14 can move toward and away from the flat paneled detector 18. The position of the source point 54 is located on the rotation axis of the gantry 50. In this manner, the motions can be coordinated such that the focal spot of the x-ray tube 14 follows an arcuate trajectory that is equidistant from the flat paneled detector 18 at all locations to minimize magnification effects.

[0060] It is also to be appreciated, the rotating gantry 50 can also be configured to move inward and outward relative to the translating gantry 52. This enables the x-ray source 14 to move with both transverse and longitudinal components of motion relative to the object. As discussed above, the additional motional components causes the out of plane contributions of the image to be blurred more fully reducing the background contributions.

[0061] It is also to be appreciated that the object could be in a commercial warehouse environment and may be any object which needs to be measured for its constituent layer parts during processing. As another option, the surface that supports the object can move relative to the x-ray source, for example it could be a conveyor belt.

[0062] With reference to FIG. 9, there is a flow chart 60 detailing a preferred method for reconstructing the 2D and 3D tomographic image representations. At a step 62, the tomographic position of the selected slice is initially read. A step 64 reads out a view of an object from the detector 18 which a step 66 converts into a floating point view. Concurrently, a step 68 extracts an exact source position for the encoder 20. Next, a step 70 determines if this is the first view which was completed. If “yes”, a step 72 saves the initial or reference source position. If this determination step 70 is “no”, then a step 74 calculates the image shift vector D and the magnification errors as set forth in Equations (3a), (3b), (3c), a view summation or integration by which the step of 76 summed or integrated views build the resultant image. A step 78 determines if the system is finished detecting views for the current layer, i.e., whether the slice image is complete. If this determination in step 78 is “yes” then the slice image is reconstructed and image processing step 80 is performed.

[0063] First, the image processing step 80 performs image analysis of the image followed by a noise clipping and filtering to remove noise and blurred background, particularly contributions attributable to out of slice elements. The step 80 also performs scaling and normalization of the image to optimize the image for viewing.

[0064] Once these digital information processes are complete, the slice imaged (data is saved in a step 82) into the volume image memory 30 for later or immediate display. Further, a next layer determination step 84 determines whether this current image is the last slice of the volume to be reconstructed or, alternatively, whether there is another layer which is to be reconstructed and saved. If this determination is “yes”, then the next slice or layer is selected in step 62 and steps 64-82 are repeated on the same views stored in view memories 24a, 24b, . . . 24n. If this determination is “no”, then a step 86 ends the reconstruction process.

[0065] The invention has been described with reference to the preferred embodiment. Obviously, modifications and alterations will occur to others upon reading and understanding the preceding detailed description. It is intended that the invention be construed as including all such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.