Cone-beam computed tomography using continuous kV beams

Abstract

Cone-beam computed tomography (CBCT) methods use a continuous beam and a region detector to perform rapid acquisition of projection data. The acquired projection data is then reconstructed to generate a tomographic image. In the acquisition of projection data, a radiation source (102) continuously illuminates a subject (112) with a cone-shaped radiation beam (116) from multiple angles, and a region detector (104) continuously reads out the data. A CBCT system is also disclosed, comprising a source operatively generating a cone-shaped radiation beam and a region detector that is synchronously movable with the source to rapidly acquire projection data for CBCT reconstruction.

Description

Technical Field

[0001] This disclosure generally relates to radiographic imaging and processing. In particular, various embodiments of cone-beam computed tomography (CBCT) using a continuous kilovolt (kV) beam are described. Background Technology

[0002] Computed tomography (CT) is known in medical imaging and radiotherapy and has proven to be a useful tool for a number of clinical applications. During CT imaging, a radiation source emits X-rays from multiple angles toward a portion of the object, and a detector detects the radiation passing through that portion to acquire projection data. Reconstruction algorithms are used to reconstruct the acquired projection data to generate a tomographic image in a two-dimensional image slice or a three-dimensional image volume of the object's portion.

[0003] Conventional CBCT uses pulsed beams to acquire projection data. A radiation source emits beam pulses at a predetermined frequency, and a detector detects the signals generated by the beam pulses transmitted through the imaging section. The emission of pulses and the detection of transmitted pulses are operated sequentially. During the detection of transmitted pulses, no pulses are delivered to the torso.

[0004] Conventional CBCT using pulsed beam acquisition significantly limits the duty cycle of the radiation source during the portion of the radiation period generated by the source operation. For example, in kV-CBCT using pulsed beam operation, the duty cycle of a kV source can be limited to 30-40%. For large duty cycles, pulsed beam CT reduces the maximum possible frame rate of the detector during the acquisition of projection data due to the sequential operation of beam pulse emission and transmission detection. Furthermore, pulsed beams can cause measurement inaccuracies because the beam output does not actually stop immediately at the end of the pulse, but can take several milliseconds until the intensity drops to zero. This so-called tailing effect can produce imaging artifacts, which can be relatively large when using short pulses.

[0005] Therefore, there remains a need for improved computed tomography (CBCT) to alleviate the problems of conventional CBCT. It is desirable to provide CBCT methods and systems that allow the radiation source to operate at 100% duty cycle. It is desirable to provide CBCT methods and systems that use continuous beams to rapidly acquire projection data and generate reconstructed images with improved image quality. Summary of the Invention

[0006] A CBCT method using continuous kV beam acquisition is disclosed. The method includes acquiring multiple projections of at least a portion of an object using a radiation source and a region detector, wherein, in acquiring a dataset of multiple projections, the source continuously irradiates the portion of the object from multiple angles using a cone-shaped radiation beam, and the region detector reads out data during the continuous irradiation of the portion of the object using the cone-shaped radiation beam. A tomographic image of the portion of the object is constructed based on at least a portion of the dataset of multiple projections.

[0007] A CBCT system is disclosed. The system includes a radiation source, a region detector that is movable synchronously with the radiation source, and an image processing device. The radiation source is operable to generate a cone-shaped radiation beam and is movable relative to a subject to irradiate at least a portion of the subject from multiple angles. The region detector is operable to acquire multiple projections of a portion of the subject, while the source continuously irradiates the portion of the subject from multiple angles using the cone-shaped radiation beam. The image processing device is configured to reconstruct a tomographic image of the portion of the subject based on at least a portion of the dataset of the multiple projections.

[0008] The present invention is provided in a simplified form to describe selected embodiments and is not intended to identify key features or essential characteristics of the claimed subject matter, nor is it intended to be used to help determine the scope of the claimed subject matter. The selected embodiments are provided only to the reader as a brief overview of certain forms in which the invention may take and are not intended to limit the scope of the invention. Other aspects and embodiments of this disclosure are described in the Detailed Description section. Attached Figure Description

[0009] These and various other features and advantages of this disclosure will be better understood by reading the following detailed description and the appended claims in conjunction with the accompanying drawings, wherein:

[0010] Figure 1 An example radiation system according to an embodiment of the present disclosure is depicted.

[0011] Figure 2 This is an example exposure time graph of the preprocessing method according to embodiments of the present disclosure.

[0012] Figure 3 The image shows a reconstruction generated from standard CBCT according to an embodiment of this disclosure. Image of a phantom.

[0013] Figure 4 The image shows an embodiment of the present disclosure generated by iterative CBCT reconstruction. Image of the phantom.

[0014] Figure 5 An image of a torso phantom reconstructed from standard CBCT according to an embodiment of this disclosure is shown. Detailed Implementation

[0015] refer to Figures 1 to 5 Various embodiments of CBCT methods and systems will now be described. Typically, example CBCT methods use a continuous cone beam and a region detector to perform rapid acquisition of projection data. The acquired projection data is reconstructed to generate tomographic images with image quality superior to or at least comparable to conventional CBCT using a pulsed beam. Prior to the reconstruction of the tomographic images, the acquired projection data may optionally be preprocessed to correct or compensate for possible geometric errors caused by acquisition using a continuous beam and region detector.

[0016] According to embodiments of this disclosure, the acquisition of projection data using a continuous beam and a zone detector effectively increases the duty cycle of the radiation source from the currently common 30% to 40% duty cycle to 100%. Therefore, an increased kV imaging dose can be delivered over a limited time span. In conventional kV-CBCT acquisition on a radiotherapy (RT) machine, pulsed beam operation represents a kV source duty cycle ranging from a typical 30% to a maximum of 40%, for example, in acquisitions with a kV pulse length of 20 milliseconds (ms) and the imager operating at 15 frames per second (fps), or in acquisitions with a kV pulse length of 10 ms and the imager operating at 30 fps. For a given source power, pulsed operation limits the total dose or mAs exposure that can be delivered during a given time period. A milliampere-second (mA-second) or mAs is a measure of radiation (milliamperes) generated over a set number of seconds via an X-ray tube. With continuous kV beam operation according to this disclosure, a 100% source duty cycle allows for the delivery of approximately 2.5 times more dose in the same amount of time.

[0017] As an example, in a conventional kV-CBCT scan (e.g., a pelvic scan), acquiring 900 projections with a total dose of 1080 mAs requires a 60-second scan time using an X-ray tube operating at 125 kV, a beam pulse length of 20 ms, a pulse current of 60 mA, and an imager operating at 15 fps, for a total beam time of 18 seconds. According to an embodiment of the invention, using a continuous beam with the same current (60 mA), the same dose amount (1080 mAs) can be delivered within 18 seconds.

[0018] As another example, radiation processing (RT) machines that allow the gantry used for delivery to rotate 360° in less than 17 seconds have been developed. However, for CBCT on such RT machines, the CBCT scan mode with the highest dose (e.g., large pelvis) currently takes longer (41 seconds) to be delivered, mainly due to the limited duty cycle (40% to 50%) of the CBCT source.

[0019] The acquisition of projection data using continuous beams also improves the effective detection speed of area detectors, such as flat panel imagers. For a given readout time of a flat panel imager, the maximum frame rate can be increased. Unlike conventional pulsed beam operation, where the emission of kV beam pulses and the readout of imager frames are constructed sequentially, with continuous beam operation according to this disclosure, the readout of imager frames can actually begin immediately after the readout of the previous frame has ended. By way of example, in pulsed beam operation with a 20ms kV beam pulse and a 20ms imager readout time (40ms cycle time), the imager operates at a frame rate of 25fps. Using continuous beam operation according to this disclosure, an imager with the same readout time (20ms) allows a frame rate of 50fps (20ms cycle time).

[0020] Acquiring projection data using continuous beams and area detectors can extend the lifespan of radiation sources such as X-ray tubes (filaments). For example, the lifespan of a kV tube is extended because a given dose can be delivered with a lower continuous current rather than in a short pulse with a high peak current.

[0021] Acquiring projection data using a continuous beam avoids the kV pulse tailing effect. In conventional pulsed beam operation, the beam output does not stop immediately at the end of the pulse, but can take several milliseconds until the beam intensity drops to zero. This so-called tailing can affect measurement accuracy. For short pulses, the tailing can contribute relatively large to the overall signal or patient dose. Acquiring data using a continuous beam according to this disclosure significantly reduces the effects associated with pulse tailing.

[0022] Acquiring projection data using continuous beams also reduces aliasing artifacts in CBCT reconstruction. For objects moving across the image, such as those rotating on a gantry, exposure with a 100% duty cycle results in more blurred structures. In CBCT reconstruction, this reduces spatial resolution to some extent away from the axis of rotation, but helps reduce aliasing artifacts, which can otherwise be clearly visible even with very short pulses.

[0023] Figure 1 Examples of imaging or radiation systems 100 are depicted, in which embodiments of the methods of this disclosure can be implemented. It should be noted that although embodiments of this disclosure are described in conjunction with imaging systems, the principles of this disclosure can be implemented in processing systems, simulation systems, research and development systems, or any other suitable radiation system that includes imaging capabilities. For example, embodiments of this disclosure can be embodied in a radiation processing system comprising... Figure 1 The imaging system shown also includes a source operable in the megavolt (MV) range, for example, from 4 to 20 MV, to produce radiation suitable for processing.

[0024] like Figure 1 As shown, the example radiation system 100 generally includes a radiation source 102, a zone detector 104, and an image processing device 106. The operation of the radiation source 102, zone detector 104, and image processing device 106 can be controlled by a computer and control system 108. For example, a treatment bed support device 110 can place, for example, a patient 112 between the source 102 and the zone detector 104 to receive radiation for imaging and / or treatment.

[0025] Source 102 and area detector 104 can be arranged opposite each other and supported by a movable or rotatable stage 114. Therefore, source 102 and area detector 104 can be moved synchronously during the acquisition of projections of an object from multiple angles. Stage 114 can be a ring stage or a C-arm stage operable to rotate source 102 and area detector 104 in radians known in the art. Alternatively or additionally, processing bed 110 supporting object 112 can be moved in multiple degrees of freedom, such as horizontal and / or vertical translation, and / or rotation about various axes, to position object 112 relative to source 102 before or during the acquisition of projection data.

[0026] In a particular embodiment, source 102 is operable to generate a conical beam 116 and to continuously illuminate the object 112 from multiple angles via a rotating stage 114. During illumination using the continuous conical beam 116, a region detector 104, such as that of a flat-panel imager, is operable to continuously read out data, allowing for the rapid acquisition of multiple projections required for CBCT reconstruction. Typically, the number of projections required for CBCT reconstruction ranges from 400 to 900, depending on the acceptable image quality for a particular application. However, the claims of this disclosure are not limited thereto. For CBCT reconstruction, more than 900 or less than 400 projections can be acquired.

[0027] In a particular embodiment, the gantry 114 can rotate the source 102 and the area detector 104 at a relatively fast speed, such as two or more 360° rotations per minute, or four or more 360° rotations per minute. During the relatively high-speed rotation, the conical beam 116 from the source 102 remains on, continuously illuminating the object 112 from multiple angles, and the area detector 104 continuously reads out data, thereby allowing the source 102 to operate at 100% duty cycle. In a particular embodiment, during a single 360° rotation of the source 102, multiple projections of at least a portion of the object 112 are acquired, during which the source 102 continuously emits the conical beam and the area detector 104 continuously reads out data. In an alternative embodiment, the 360° rotation is divided into multiple sectors or blocks, for example, into five sectors, each sector being a 72° arc. In each of the multiple sectors or blocks, multiple projections are acquired. During this acquisition, source 102 continuously emits a cone-shaped beam, and area detector 104 continuously reads out the data, allowing source 102 to operate at 100% duty cycle. The projection data acquired in the multiple sectors or blocks can then be combined to form a complete 360° dataset.

[0028] In a particular embodiment, source 102 is an X-ray tube that is operable in the kilovolt (kV) range (e.g., ranging from 50 to 150 kV, with a peak kilovolt (kVp) of, for example, 125 kVp). In a particular embodiment, the area detector is an X-ray tube with a large active detection area (e.g., 20 × 20 cm). 2 Or larger, 30×30cm 2 Or larger, or 40×40cm 2 Or larger, or in sizes of 20×20 and 45×45cm 2 A flat panel imager (of any size between) is used. A conical beam 116 generated by source 102 has a conical beam angle. The conical beam angle can be defined in the direction of the pixel columns and / or rows of the flat panel imager. By way of example, this is for an imager with a size of 20 × 45 cm centered on the conical beam. 2 For a flat panel imager with an effective detection area, the conical beam angle can range from 7 to 17 degrees in both the pixel column and row directions of the imager. Other conical beam angles can be employed by combining area detectors of different sizes. In an alternative embodiment of this disclosure, the area detector can be laterally offset relative to the centerline of the conical beam.

[0029] The synchronized rotation of source 102 and region detector 104 around object 112 provides a reconstructed volume comprising at least a portion of object 112 (e.g., the pelvic region of a patient). The reconstructed volume is a typically cylindrical volume space in which a three-dimensional image can be reconstructed by CBCT. The diameter of the reconstructed volume is commonly referred to as the field of view (FOV). The length of the reconstructed volume is referred to herein as the scan length. The size of the FOV and the scan length of the reconstructed volume can depend on the distance between the focal point of source 102 and the detection surface of region detector 104, the cone angle of the beam 116 generated by source 102, and the size of the detection surface of region detector 104. In a particular embodiment, the distance between the focal point of source 102 and the detection surface of region detector 104, the cone angle of the beam generated by source 102, and the size of the detection surface of region detector 104 can be selected such that the reconstructed volume has an FOV ranging from 25 to 50 cm or greater, and / or a scan length ranging from 15 to 28 cm or greater. It should be noted that the size of the reconstructed volume is provided for illustrative purposes. As described above, the region detector can be laterally offset relative to the beam (e.g., in a half-fan configuration), and the FOV and scan length of the reconstructed volume can be greater than or less than the example values ​​above.

[0030] Image processing device 106 may include computer 118 and software 120 designed for reconstructing tomographic images based on acquired projections. Various reconstruction software or algorithms are known in the art and can be used in conjunction with embodiments of this disclosure. For example, standard CBCT reconstruction software based on the filtered back projection technique can be used to generate tomographic images. Iterative CBCT reconstruction software based on the algebraic reconstruction technique can also be used to generate tomographic images, and this software has become commercially available due to increased computing power. In a particular embodiment, the software 120 implemented in image processing device 106 can reconstruct projection data into tomographic images representing the three-dimensional volume of a portion of an object.

[0031] In certain embodiments, prior to reconstruction, the acquired projection data is preprocessed to interpret the radiation beam and detector properties, as well as the system's electronic properties. For example, the acquired projection data may be preprocessed to correct detector pixel defects, interpret electronic gain, or apply logarithmic transformations, etc. In certain embodiments according to this disclosure, the projection data may be preprocessed to correct or compensate for geometric errors or geometric twists caused by the acquisition of the continuous conical beam and the area detector, as will be described in more detail below.

[0032] CBCT reconstruction algorithms generate 3D volumes based on a series of projections (frames) acquired during gantry rotation. For multiple processing steps, current algorithms assume that all pixel data in a single frame represents the same time. For example, backprojection of pixel data assumes the same gantry position for all image rows in a given projection. If the systematic time delay introduced by successive beam acquisition is not considered, this delay will generate geometric errors, such as those in the upper rows where pixel data represents slightly different gantry positions. Without correction, this will result in geometric errors corresponding to cylinders with slight torsion.

[0033] By way of example, a CBCT system on an RT platform may include a device with a size or effective detector area of ​​43 × 43 cm. 2 The flat-panel imager, with a frame readout time of 20 ms, is positioned with lateral offset (semi-fan configuration) and provides a CBCT reconstruction volume of 49 cm in diameter. If the RT platform gantry rotates at 4 revolutions per minute (RPM) (24° per second), the readout time of the flat-panel imager will correspond to a 10 ms time delay between the center detector row and the top or bottom detector row of the imager. At 4 RPM, a structure located 24 cm from the isobath axis (i.e., near the edge of the reconstruction volume with a 49 cm diameter) will move approximately 1 mm in the reference frame of the rotating imaging system over a period of 10 ms. This means that if the structure is positioned such that its image is projected onto the top detector row, the measured position of the structure can be up to 1 mm away from the position observed in the pulsed system.

[0034] Due to the cone-shaped beam geometry, a given imager row contributes to voxels located on different slices of the volume, except for the central row of the imager which contributes only to the central slice. For this reason, the “de-twisting” of the reconstructed volume cannot be compensated for simply by rotating each slice of the volume back by a clearly defined small angle, which would only represent an inaccurate compensation.

[0035] According to specific embodiments of this disclosure, the preprocessing of projection data includes weighting data of a given projection acquired within a given time period with data of projections acquired after the given projection, and / or with data of projections acquired before the given projection. The preprocessing method according to this disclosure allows the use of the same reconstruction algorithm and thus simplifies the introduction of this acquisition mode into existing CBCT reconstruction frameworks.

[0036] Figure 2 This is an example exposure time map illustrating a preprocessing method according to a specific embodiment of this disclosure. The region detector 104 for acquiring projection data is a flat panel imager, comprising multiple detector pixels in columns and rows (the total number of rows is specified as h). Figure 2In the diagram, the vertical axis represents the detector row; the horizontal axis represents time. For a given projection (index i), the detector pixel readout begins at the first row (index y = 0) at time t0, then the second row at t1… the center row at tc, and ends at the last row at th. The time elapsed between consecutive row readouts (e.g., between the first and second rows, t1 and t0) is referred to herein as the line readout period (lineTime). The time taken for the imager to read out all rows is referred to herein as the detector readout time. In the example shown, the detector readout time can be approximated by multiplying lineTime by the number of rows.

[0037] After a given projection (index i) is read out, the next projection (index i+1) is read out starting at time t'0 and continuing until the center row at t'c, until the last row of the projection is completed. The time elapsed between consecutive projections or frame readouts is called the period or cycle time (T). The cycle time can be determined by the detector frame rate (T = 1 / frame rate). For consecutive kV beams, the cycle time also represents the total time (exposure time) during which the kV beam signal is accumulated (integrated) in the pixel. The cycle time is equal for all detector rows. The shortest cycle time (fastest frame rate) can be determined by the detector readout time, but the cycle time can also be much longer than the detector readout time.

[0038] The timestamp associated with a given projection and acquisition parameters such as the angular position of the gantry are constrained relative to the center pixel row of the detector. One objective of preprocessing is to generate projection data where all pixel rows correspond to the same time (and gantry position), arbitrarily constrained, for example, by the center row of the detector. Figure 2 In the diagram shown, for a given projection i, the first row is read out earlier than the center row. According to a specific embodiment of the invention, preprocessing of projection i can use pixel data from the next projection (index i+1) for linear interpolation. The weight (w) of the original data can be determined by the relative time delay. Similarly, for a given projection i, the last row is read out later than the center row. Therefore, preprocessing of projection i can also use pixel data from the previous projection (index i-1) for linear interpolation. The weight (w) of the original data can be determined by the relative time delay. By way of example, the time delay of reading out the data of a given row (y) relative to the reference row (y0) of the region detector can be used.

[0039] In a particular embodiment, the weighting of the pixel data for a given projection is performed based on the following linear interpolation:

[0040] Replace pi(.., y) with:

[0041] w * pi(.., y) + (1 - w) * pi+1(.., y), where

[0042] pi(.., y) represents the data value of the pixel at row y of a given projection i that includes h rows of pixels and y < h / 2 obtained by a region detector,

[0043] pi+1(.., y) represents the data value of the pixel at row y of projection i + 1, and

[0044] w represents a weighting factor, and W = 1 - (h / 2 - y) * (lineTime / CycleTime), or

[0045] Replace pi(.., y) with:

[0046] w * pi(.., y) + (1 - w) * pi-1(.., y), where

[0047] pi(.., y) represents the data value of the pixel at row y of a given projection i that includes h rows of pixels and y > h / 2 obtained by a region detector,

[0048] pi-1(.., y) represents the data value of the pixel at row y of projection i - 1, and

[0049] w represents a weighting factor, and W = 1 - (y - h / 2) * (lineTime / cycleTime).

[0050] Embodiments of the present disclosure will now be described in connection with working examples. It should be noted that while some specific details are set forth in the working examples to provide a thorough understanding of the present disclosure, it will be clear to those of ordinary skill in the art that some of these specific details may not be required to practice the embodiments of the present disclosure. In other instances, some well-known details may not be described in order to avoid unnecessarily obscuring the embodiments of the present disclosure.

[0051] Example [[ID=​​​​​​The effective detector area is a flat panel imager. The flat panel imager is read out in pixel-mixing mode (1×4) or at a resolution of 1280×320 pixels. A phantom was used in the evaluation. 504 and the body phantom were obtained from the Phantom Laboratory in Salem, New York. Scans were performed in both clockwise and counterclockwise directions for geometric accuracy testing. For CBCT using a continuous kV beam, the X-ray tube was operated at 100% duty cycle. For CBCT using a pulsed kV beam, the X-ray tube was operated at approximately 30% duty cycle or lower. Some other operating parameters are listed in Table 1.

[0053] Table 1: Operating Modes and Parameters

[0054]

[0055]

[0056] Example 1 (Example 1 in Table 1) was designed to test whether the results of continuous kV beam generation were comparable to or superior to those of pulsed beam acquisition at the same frame rate using a flat-panel imager. As shown in Table 1, in Example 1 (Example 1), the total dose delivered to the phantom in both continuous and pulsed beam acquisition was practically the same, approximately 498 mAs. The total number of acquired projections was also the same, 491. The count or value of the gray levels of individual pixels on the projections was also the same. Figure 3 The diagram shows the reconstruction generated by standard CBCT based on projection data acquired using a pulsed kV beam, shown in the upper left corner, under the operating parameters described in Table 1 for Example 1 (Example 1). The phantom image, at the upper center, is generated based on projection data acquired using a continuous kV beam. Image of the phantom. Figure 4 This shows the result generated by iterative CBCT (iCBCT) reconstruction. Image of the phantom. Figure 4 upper left and upper center Phantom image display and Figure 3 The results are similar to those in the previous section. Figure 5 An image of the torso phantom reconstructed from standard CBCT is shown. Under the operating parameters described in Table 1 for Example 1... Figure 5 The upper left shows a torso phantom image reconstructed using standard CBCT based on projection data acquired with pulsed kV beams, while the top center shows a torso phantom image generated based on projection data acquired with continuous kV beams. Figures 3 to 5The reconstructed images show that the results generated by continuous kV beam acquisition are better than or at least as good as those generated by pulsed beam acquisition.

[0057] Example 2 (Example 2 in Table 1) was designed to test whether the results generated using rapid acquisition with a continuous kV beam were comparable to or superior to those generated by pulsed beam acquisition. As shown in Table 1, in Example 2, the total dose delivered to the phantom in both continuous and pulsed beam acquisition was virtually the same, approximately 498 mAs. One difference was that with continuous beam acquisition, the flat-panel imager was operated at a much faster readout rate, approximately 49 frames per second, with a line time of approximately 57.66 μs, a readout time of approximately 19.2 ms, and a loop time of approximately 20.48 ms. Thus, with continuous beam acquisition, a higher total number of frames or projections (811) were read out, but the counts per projection were lower. However, the results generated by continuous kV beam acquisition (image quality with stripes and noise) were still superior to or at least as good as those produced by pulsed beam acquisition, as... Figures 3 to 5 As shown. Figure 3 The diagram shows the reconstruction generated by standard CBCT based on projection data acquired using a pulsed kV beam, shown in the upper left corner, under the operating parameters described in Table 1 for Example 2. The phantom image, in the upper right corner, is generated based on projection data acquired using a continuous kV beam. Image of the phantom. Figure 4 The image shows the result generated by iCBCT reconstruction. Image of the phantom. Figure 4 upper left and upper right The phantom images show the relationship with Figure 3 The results are similar to those in the previous section. Figure 5 An image of the torso phantom generated by standard CBCT reconstruction is shown. Figure 5 Under the operating parameters described in Table 1, the upper left shows a torso phantom image generated by standard CBCT reconstruction based on projection data acquired with pulsed kV beams, and the upper right shows a torso phantom image generated based on projection data acquired with continuous kV beams.

[0058] Example 3 (Example 3 in Table 1) was designed to test whether acquisition using a source operating at 100% duty cycle directly reduced scan time. As shown in Table 1, in Example 3, operating parameters were adjusted so that approximately the same total dose (595 mAs using a continuous beam versus 601 mAs using a pulsed beam) was delivered to the phantom. Approximately the same total number of projections were acquired (855 continuous beams versus 859 pulsed beams), with similar counts on the projections. The total scan time was significantly reduced from approximately 30.8 seconds using pulsed beam acquisition to approximately 17.5 seconds using continuous beam acquisition, even though the X-ray tube current used for continuous beam acquisition was less than half that used for pulsed beam acquisition (34 mA using a continuous beam versus 70 mA using a pulsed beam). However, as Figures 3 to 5 As shown, the results generated by continuous kV beam acquisition are still better than, or at least as good as, the results generated by pulsed beam acquisition. Under the operating parameters for Example 3 as described in Table 1, Figure 3 The image in the lower left shows a reconstruction generated using standard CBCT based on projection data acquired with pulsed KV beams. The phantom image, located in the lower right corner, is generated based on projection data acquired using a continuous beam. Image of the phantom. Figure 4 Generated by iCBCT reconstruction Image of the phantom. Figure 4 The middle part is in the lower left and lower right parts Phantom image display and Figure 3 The results are similar to those in the previous section. Figure 5 An image of the torso phantom reconstructed using standard CBCT is shown. Under the operating parameters described in Table 1 for Example 3... Figure 5 The lower left section shows a torso phantom image reconstructed using standard CBCT based on projection data acquired with a pulsed kV beam, while the lower right section shows a torso phantom image generated based on projection data acquired with a continuous kV beam.

[0059] Back Figure 1The radiation source 102 can be an operable source to produce photons, protons or other heavy ions, and electrons. By way of example, source 102 can include an x-ray tube comprising a target that generates x-rays based on electron impacts. The x-ray tube 102 can include collimating devices to collimate the beam, such that the beam exiting the x-ray tube is typically conical or pyramidal in shape. As another example, source 102 can include a linear accelerator comprising a target that generates x-rays based on electron impacts, and various collimating devices for shaping and / or scaling the x-rays. As another example, source 102 can include a nozzle that emits protons generated by a cyclotron or synchrotron and transported to the nozzle. Typically, source 102 can produce or emit radiation suitable for imaging. Alternatively, the source can operate in the kilovolt range, for example, from 50 to 150 kV, to produce radiation suitable for diagnostic imaging. Alternatively, the source may be operated at the megavolt (MV) level for use with, for example, an electronic field imaging device (EPID) to acquire images, such as for determining patient settings or for interventional procedures or treatment planning iterations. In some embodiments, source 102 may be a source operable in two or more modes, such as at the kV or MV energy level, to produce radiation for various applications.

[0060] In a particular embodiment, source 102 includes an operable x-ray tube to generate x-rays. X-ray tubes are known in the art and therefore their detailed description is omitted herein. Briefly and generally, an x-ray tube includes a source of electrons (cathode) (e.g., a filament) and a target (anode) (e.g., tungsten metal) on a rotating stage enclosed in a vacuum tube. A supply of current causes the filament to be heated to generate electrons. A voltage applied to the vacuum tube accelerates the electrons toward the anode to strike the target and generate x-rays. The small surface area on the target that generates the x-rays is called the focal point of the x-ray tube or source, and the size of this focal point can be determined by the electron beam from the cathode. The generated x-rays can then be shaped by a collimator and exit through a window in the tube. Although in Figure 1 Although not shown, the signals of the x-ray controller and generator assembly (e.g., current, voltage, exposure time) and other operating parameters of the x-ray tube, as well as the voltage to be supplied to the x-ray tube, are included. The x-ray controller and generator can then be controlled by a computer and control system 108.

[0061] Still referencing Figure 1 Area detector 104 detects radiation transmitted through object 112, indicating attenuation characteristics or the structure of the imaged portion of object 112. Area detector 104 is coupled to data acquisition system 105, which includes electronic circuitry for providing control signals, receiving and processing data signals, and outputting data to image processing device 106. Although in Figure 1The data acquisition system 105 and the area detector 104 are shown as separate units, but they can be integrated into a single unit controlled by a computer and control system 108.

[0062] The area detector 104 includes a plurality of detection elements arranged in rows and columns over a two-dimensional region. In a particular embodiment, the area detector 104 includes a flat-panel imager with a flat detector surface. In an alternative embodiment, the area detector 104 has a curved detector surface. The area detector 104 may have an effective detection area or a size, for example, from 20 × 20 cm. 2 Up to approximately 45×45cm 2 Various area detectors are available from Varex Imaging Corporation in Salt Lake City, UTAH.

[0063] In a particular embodiment of this disclosure, the area detector 104 is a flat panel imager. The flat panel imager may include a radiation conversion layer and a detector array. The conversion layer converts radiation, such as X-ray photons, into visible light. The detector array detects the visible light and converts it into an electrical signal. The conversion layer may include components capable of generating visible light photons in response to X-ray radiation. Therefore, the detector array may include a photosensitive element capable of generating an electrical signal in response to visible photons generated by a scintillator material. Suitable scintillator materials include gadolinium oxysulfide (Gd₂O₂S:Tb), cadmium tungstate (CdWO₄), and bismuth germanate (Bi₄Ge₃O₄). 12 The photosensitive element may include, but is not limited to, mercuric iodide (BGO), cesium iodide (Csl), cesium thallium iodide (Csl:Tl), thallium-doped sodium iodide (TI), or any combination thereof. Suitable photosensitive elements may include photodiodes, photogates, or phototransistors. Alternatively, the conversion layer may include a photoconductor material that can directly convert X-ray photons into charges (electron-hole pairs). Therefore, the detector array may include electrodes on either side of the photoconductor material to collect the charges generated by the photoconductor material. Suitable photoconductor materials include, but are not limited to, mercuric iodide (Hg12), lead iodide (Pb12), bismuth iodide (Bil3), zinc cadmium telluride (CdZnTe), and amorphous selenium (a-Se).

[0064] In certain embodiments, the example flat panel imager 104 may include a large number (e.g., hundreds of thousands or millions) of detector pixels. The large number of detector pixels may be arranged to form multiple rows and columns forming an effective detector region. In some specific embodiments, each detector pixel may include an addressable photosensitive element such as a photodiode and a switching transistor such as a thin-film transistor (TFT) or complementary metal-oxide-semiconductor (CMOS) transistor.

[0065] The detector array may also include multiple address lines and multiple data lines. Each of the multiple address lines connects multiple detector pixels in a row to a driver control component of the data acquisition system. Each of the multiple data lines connects multiple detector pixels in a column to a readout control component of the data acquisition system 105. The driver control component provides control signals for accessing selected pixel rows. The readout control component provides control signals for reading signals from the pixels. By way of example, when it is desired to capture an image signal from the detector array, the control signals from the driver control component drive the gates of switching elements, such as TFTs, in the selected pixel rows, and the signals stored in the selected pixel rows are read out by the readout control component. The signals from the selected pixels can be buffered, amplified, and converted by an analog-to-digital converter (ADC) using the electronics of the data acquisition system 105. The resulting digitized data signals can then be multiplexed, buffered, and transmitted to an image processing device 106 for further processing.

[0066] In the operation of the detector array, control signals from the driver control circuitry for pixel rows can be asserted on address lines for a predetermined time period or row time. During the control signal assertion, signals from each pixel in the selected row are transmitted to the readout control circuitry via column data lines, where signals on each data line are received and buffered by a corresponding charge-sensitive amplifier. Thus, an entire row of image data can be captured within the row time period. For each subsequent row time period, the subsequent rows of image data are captured. At the end of the "frame time" period, the entire image can be captured. In this way, pixels included throughout the entire effective detection area can be read out row by row within the row time period. In an alternative embodiment, the flat panel imager can use separate data lines, where the upper and lower halves of the array are read out simultaneously. This allows the flat panel imager to read out faster, for example, requiring only half the "frame time" to read out pixels. Therefore, the preprocessing of the projection data described above can be adjusted according to the specific readout timing implemented in this readout scheme. By way of example, flat panel imagers can operate at speeds ranging from 15 to 50 frames per second (fps), depending on dose requirements, image resolution, and the number of projection datasets acquired.

[0067] Still referencing Figure 1 The image processing apparatus may include a computer 118 and software 120, the software 120 being designed to process acquired projection data and reconstruct tomographic images based on the projection data. The computer 118 may include a processor, memory, and optionally a user interface and a network interface. The processor may include a central processing unit (CPU) known in the art, such as... processor or Processor, or graphics processing unit (GPU), such as GPU, or other types of processing units. The processor can retrieve and execute computer-executable instructions from memory, which can cause the processor to perform any methods and / or steps according to the embodiments of this disclosure described above.

[0068] Memory may include any one or a combination of volatile memory elements and non-volatile memory elements. Memory may include random access memory (RAM) or other dynamic storage devices for storing information and instructions executed by a processor, as well as for storing temporary variables or other intermediate information during the execution of instructions by the processor. Memory may also include read-only memory (ROM) or other static storage devices for storing static information and instructions specific to the processor. Memory may also include data storage devices such as magnetic disks or optical disks for storing information and instructions. Memory (e.g., non-transitory computer-readable media) may include programs (logic) for operating a computer system and for performing applications or other processing plans, including projection and dose calculations such as dosimetry as described above. Furthermore, memory may include a database storing any information that can be selected by a user (such as a radiation oncologist or radiation therapist).

[0069] User interface devices can include components that allow users to interact with the computer system, such as keyboards, pointing devices, pens, touch input devices, and voice input devices. Output devices such as display devices, printers, and speakers can also be included in the computer system.

[0070] A network interface allows a computer system to communicate with other devices or systems over a communication network such as the Internet or an intranet (e.g., a local area network). A network interface may include a Wi-Fi interface, an Ethernet interface, a Bluetooth interface, or other wireless or wired interfaces. A network interface allows a computer system to receive and transmit electrical, electromagnetic, or optical signals carrying data streams representing various types of information. For example, a network interface may allow a computer system to receive data streams representing software programs used for processing plans over a communication network.

[0071] Various embodiments of the CBCT method and system are described with reference to the accompanying drawings. It should be noted that some drawings are not necessarily drawn to scale. The drawings are intended only to facilitate the description of particular embodiments and are not intended to be exhaustive or to limit the scope of this disclosure.

[0072] Unless otherwise specifically defined, all technical and scientific terms used herein have the meanings commonly understood by one of ordinary skill in the art. As used in the specification and appended claims, the singular forms “a,” “an,” and “the” include multiple references unless the context clearly indicates otherwise. The term “or” refers to a non-exclusive “or” unless the context clearly indicates otherwise. The terms “first” or “second” are used to distinguish one element from another when describing various similar elements and should not be construed as being in any particular order unless the context clearly indicates otherwise.

[0073] Those skilled in the art will recognize that various other modifications can be made. All such and other variations and modifications were conceived by the inventors and are within the scope of this invention.

Claims

1. A method comprising: A dataset of multiple projections of at least a portion of an object is acquired using a radiation source and a region detector capable of moving synchronously with the source, wherein in the acquisition of the dataset of multiple projections, the source continuously illuminates the portion of the object from multiple angles using a conical radiation beam, and the region detector reads out data during the continuous illumination of the portion of the object using the conical radiation beam. as well as A tomographic image of the object is reconstructed based on at least a portion of the dataset of the multiple projections; The projected data is preprocessed by weighting the pixel data values ​​of the given row (y) based on the time delay of reading the data of the given row (y) relative to the reference row (y0) of the region detector.

2. The method of claim 1, wherein in the acquisition of the plurality of projected datasets, the source is operated at the kilovolt (kV) level.

3. The method of claim 1, wherein the dataset of the plurality of projections is acquired when the source rotates 360 degrees or more around the object and continuously illuminates the portion of the object during the rotation.

4. The method of claim 1, wherein the source includes a focal point and generates the conical radiation beam having a cone angle, and the area detector includes an effective detection surface spaced at a distance from the focal point of the source, such that synchronous rotation of the source and the area detector provides a reconstructed volume including the portion of the object, the reconstructed volume having a field of view ranging from 25 to 50 cm and / or a scan length ranging from 15 to 28 cm.

5. The method of claim 1, wherein the area detector comprises a flat panel imager, the flat panel imager comprising a plurality of detector pixels in rows and columns, each of the plurality of detector pixels having an addressable photosensitive element and a switching TFT or CMOS transistor.

6. The method of claim 5, wherein the flat panel imager comprises approximately 40 × 40 cm 2 Or a larger effective detection area.

7. The method according to claim 1, wherein the preprocessing further comprises: The data of the given projection is weighted together with the data of the projections previously obtained for the given projection.

8. The method of claim 1, wherein the weighting of the data for the given projection is performed according to the following linear interpolation: p i (..,y) is replaced with: w * p i (.., y) + (1 - w) * p i+1 (.., y), where p i (.. , y) represents the data value of the pixel at row y of a given projection i that includes h rows of pixels and y < h / 2, obtained by the region detector. p i+1 (.., y) represents the data value of the pixel at row y of projection i+1, and w represents the weighting factor, and w = 1 - (h / 2 - y) * (lineTime / cycleTime). lineTime represents the time period between consecutive line reads, and cycleTime represents the time period between consecutive readouts of the projection. or p i (..,y) is replaced with: w * p i (.. , y) + (1 - w) * p i-1 (.. , y), where p i (.., y) represents the data value of the pixel at row y of a given projection i, which includes h rows of pixels and y > h / 2, obtained by the region detector. p i-1 (.., y) represents the data value of the pixel at row y of projection i-1, and w represents the weighting factor, and w = 1 - (yh / 2) * (lineTime / cycleTime). lineTime represents the time period between consecutive line reads, and cycleTime represents the time period between consecutive readouts of a projection.

9. A system comprising: A source operable to generate a cone-shaped radiation beam, the source being movable relative to an object to irradiate at least a portion of the object from multiple angles; A region detector capable of moving synchronously with the source, the region detector being operable to acquire multiple projections of the portion of the object, while the source continuously illuminates the portion of the object from the multiple angles using the conical radiation beam; as well as An image processing device is configured to reconstruct a tomographic image of the portion of the object based on at least a portion of a dataset of the plurality of projections; The system is further configured to preprocess the projected data by weighting the data values ​​of the pixels in the given row (y) based on the time delay of reading the data of the given row (y) relative to the reference row (y0) of the region detector.

10. The system of claim 9, wherein the source comprises an x-ray tube capable of operating at kilovolt levels to generate the conical radiation beam.

11. The system of claim 9 further includes a source capable of operating at megavolt levels to generate a radiation beam suitable for the treatment of a disease.

12. The system of claim 9, wherein the area detector comprises a flat panel imager, the flat panel imager comprising a plurality of detector pixels in rows and columns, each of the plurality of detector pixels having an addressable photosensitive element and a switching TFT or CMOS transistor.

13. The system of claim 12, wherein the flat panel imager comprises approximately 40 × 40 cm 2 Or a larger effective detection area.

14. The system of claim 9, wherein the source includes a focal point and generates the conical radiation beam having a cone angle, and the area detector includes an effective detection surface spaced at a distance from the focal point of the source, such that synchronous rotation of the source and the area detector provides a reconstructed volume including the portion of the object, the reconstructed volume having a field of view ranging from 25 to 50 cm and / or a field of height ranging from 15 to 28 cm.

15. The system of claim 9, wherein the image processing device is further configured to preprocess the plurality of projections by weighting data of the given projection acquired within the given time period with data of a projection acquired prior to the given projection.

16. The system of claim 9, wherein the image processing device is configured to weight the data of the given projection according to the following linear interpolation: p i (..,y) is replaced with: w * p i (.. , y) + (1 - w) * p i+1 (.. , y), where p i (.. , y) represents the data value of the pixel at row y of a given projection i that includes h rows of pixels and where y < h / 2, obtained by the region detector. p i+1 (.., y) represents the data value of the pixel at row y of projection i+1, and w represents the weighting factor, and w = 1 - (h / 2 - y) * (lineTime / cycleTime). lineTime represents the time period between consecutive line reads, and cycleTime represents the time period between consecutive readouts of the projection. or p i (..,y) is replaced with: w * p i (.. , y) + (1 - w) * p i-1 (.. , y), where p i (.., y) represents the data value of the pixel at row y of a given projection i, which includes h rows of pixels and y > h / 2, obtained by the region detector. p i-1 (.., y) represents the data value of the pixel at row y of projection i-1, and w represents the weighting factor, and w = 1 - (yh / 2) * (lineTime / cycleTime). lineTime represents the time period between consecutive line reads, and cycleTime represents the time period between consecutive readouts of a projection.

Images