JP5194095B2 - Apparatus and method for use in a computed tomography system using a reduced size detector covering only the half-image area - Google Patents

Description

  The present invention covers only half of the field of view, thereby increasing the size and cost of an area detector without increasing (or substantially increasing) artifacts. The present invention relates to a method and an apparatus used in a volumetric computed tomography (VCT) system that uses a reduced-size area-type detector capable of reducing the above.

  In computed tomography (CT), it is common to apply X-rays to a patient, collect digital X-ray projection data of a portion of the patient's body, and process and backproject this digital X-ray projection data, and then a CT system. It is necessary to display an image on the display monitor. A CT system typically includes a gantry, a table, an x-ray tube, an x-ray detector array, a computer, and a display monitor. The computer sends commands to the gantry controller that causes the gantry to rotate the x-ray tube and / or the detector array at a specific rotational speed.

  In the third generation CT system, there is a relative rotational movement between the patient's body and the gantry, part of which is composed of a detector array and an X-ray tube. As this relative rotational movement occurs, the computer acquires a digital radiograph by controlling the data acquisition process performed by the x-ray tube and the detector array. The computer then executes the reconstruction algorithm to process and backproject the digital radiographic data and display the reconstructed CT image on the display monitor.

  Many CT systems currently in use utilize a single row detector within the gantry, and this single row detector is usually referred to as a linear array of detector elements. In a further improved CT system, a multi-row detector is fabricated using a 2-4 row linear array of detectors. Both detector arrangements can be used in a helical scan protocol. However, a multi-row detector facilitates patient scanning because the specified axial coverage for the patient can be scanned in a shorter time by increasing the helical pitch of the detector array. The helical pitch is typically defined as the ratio of the displacement of the table supporting the patient during one gantry rotation to the detector pitch. For example, a helical pitch of 1 means that the patient table translates by an amount equal to the detector pitch during one rotation of the CT gantry of the CT system.

  Typically, a linear detector or an array of multi-row detectors covers the entire field of view for the X-ray fan beam emitted by the X-ray source. In other words, X-rays that have passed through the subject being scanned (which may or may not be a patient) or have irradiated that region of the subject are absorbed by the detector array.

  In some CT imaging systems, it is desirable to reduce the size of the detector array, and in some cases it is necessary to reduce the size of the detector array. For example, recent developments in CT technology use an area detector array composed of multiple rows of linear detector arrays for CT data collection. At present, detector panels that cover the entire imaging area, ie the entire range of the patient being imaged, are not yet available. In addition, some systems that use linear detector arrays provide an extremely wide field of view compared to the patient being scanned. Even in this situation, it is desirable to reduce the size and cost of the detector array.

  One technique that has been utilized to overcome these limitations is to translate a detector array that is smaller in size and half its width. For example, assume that the original size of the detector array required to cover the desired imaging area of the patient is 80 cm. A smaller detector can be used that has a width equal to half the original detector width (in this case 40 cm). The detector is offset by half the detector width (20 cm in this example) to cover approximately half of the CT imaging system imaging area. According to this example, the same imaging area of the patient can be collected by a detector having a width equal to half the original value.

  Furthermore, it is possible to expand the imaging area with a system including a detector having a fixed width. Usually, the center of rotation projection of the CT imaging system coincides with the center of the detector panel. The center of rotation of CT imaging is the physical position around which the X-ray source and detector array are rotated. However, the imaging field (FOV) of the system can be expanded by shifting the detector by half the detector width with respect to the original position. The projection of the center of rotation of the imaging system is near the edge of the shifted linear or multi-row detector. However, this detector can still measure projection data from x-rays that pass through the physical center of rotation of the imaging system (ie, the isocenter location). On the other hand, with this arrangement, the imaging area of the imaging system is virtually double the original configuration, which can greatly expand the imaging area of the imaging system. A system configuration that shifts the detector by half its width is typically called a half-detector shift.

  In the fan beam CT system, which is a CT system that irradiates only the detector panel as the X-ray source and emits X-rays having a fan-like shape with a certain angle aperture, the CT gantry is rotated at all angles. It is necessary to collect projection data for a part of Specifically, it is necessary to collect projection data while the gantry rotates around the patient by an angular region equal to 180 degrees plus the fan angle. Again, the fan angle is a measure of the X-ray angular aperture that only illuminates the detector array in the axial plane of the imaging system. It is clear that some of the projection data is redundant because the measurement of the projection does not necessarily have to be made during the entire 360 degree rotation of the gantry around the patient.

  In the half detector shift configuration of the CT system, data is collected for a full 360 degree rotation of the gantry. At each view angle of the gantry, only half of the projection data is measured. Data from other views of the gantry is used to complete the projection data for a given view angle. Schemes for performing this process are well known in the art. However, when combining measurement projection data that covers half of the imaging system's field of view with data from other views of the gantry, the resulting projection data may not match well near the center of the projection data. is there. These mismatches, if not reduced or eliminated, can cause undesirable artifacts in the reconstructed image.

  One technique currently used to mitigate artifacts caused by projection data discontinuities within the field of view utilizes a weighting function to smooth the data discontinuities within the transition region. This technique requires the detector to have some additional detector elements that extend beyond the projection of the center of rotation of the imaging system onto the detector. When the gantry rotates 360 degrees around the patient, an area of the shifted detector panel that extends in both directions slightly beyond the projection of the center of rotation on the detector is called a transition area. Actual data is measured by the detector in half of the transition region. Also, within the second half of the transition area, data can be created from another view of the gantry. Data in the transition region is multiplied with a weighting factor to smooth the discontinuities. In general, when the transition area is large, the image quality is improved, but the imaging area of this system configuration is slightly smaller than the area provided by the half-detector shift configuration, which leads to an increase in system cost.

  It is necessary to improve the integration of the measurement data and the data created in the transition area so that the entire imaging area can be realized with this detector array.

  Stereoscopic CT systems that utilize a half-detector shift configuration require projection data to be measured within half of the imaging system's field while the other half of the projected radiograph must be generated from opposite rays. There is. However, the projection data measured at another projection angle of the CT gantry has the same azimuth as the ray when measured with a detector that is twice the original width and does not have a deviation (offset). Absent. Therefore, there is a need for a VCT system that utilizes area detectors in a half detector shift configuration, realizes its benefits, and overcomes the above problems.

JP-A-10-290798

  A computed tomography (CT) system comprising an X-ray source and a detector for acquiring projection data of a subject. The detector is shifted by half its width with respect to the center position corresponding to the projection of the center of rotation of the CT system onto the detector. According to the method of the present invention, for each projection view, one detector element value Va is selected from the detector elements closest to the isocenter of the CT system. The detector element value Vb is then estimated for this selected detector element from the opposite direction or from the forward projection in the same direction. A smoothing function is then selected that can remove the difference between Va and Vb. This smoothing function is then applied to remove the difference between Va and Vb. Next, a smooth transition region is generated by applying a weight function to smoothen the step when combining the true projection data and the estimated projection data.

  Prior to describing the method and apparatus of the present invention, a general discussion of the VCT system of the present invention will be presented with reference to FIG. FIG. 1 is a block diagram of a stereoscopic CT scanning system suitable for implementing the method and apparatus of the present invention. While this stereoscopic CT scanning system will be discussed with respect to use in image reconstruction for patient anatomical features, the present invention is limited to imaging any particular object. Please understand that it is not. Furthermore, as will be appreciated by those skilled in the art, the present invention can also be used for industrial processes. Furthermore, the present invention is not limited to a medical CT apparatus, but an industrial system in which the geometric configuration of the X-ray source and detector is fixed, and the subject rotates during the scan time. Is also included.

  In a stereoscopic CT scan system, the gantry rotates around a subject such as a patient and projection data is collected. The computer 1 controls the operation of this stereoscopic CT scan system. In this specification, the term “gantry rotation” is intended to represent the rotation of the X-ray tube 2 and / or the rotation of the detector 3 (preferably, a high-resolution area detector). The X-ray tube 2 and the area detector 3 are included in the gantry. The controllers 4A and 4B are controlled by the computer 1 of the stereoscopic CT scanning system and are coupled to the X-ray tube 2 and the detector 3, respectively. The controllers 4A and 4B provide a suitable relative rotational movement with respect to the X-ray tube 2 and / or the detector 3. The control device is not necessarily required individually. A single controller component can also be used to rotate the gantry. It should also be noted that the computer 1 controls fluctuations in image scanning time, image resolution and / or axial coverage to implement the method of the present invention.

  The computer 1 controls the data collection process by instructing the data collection system 6 when to sample the detector 3 and controlling the speed of the gantry. In addition, the computer 1 instructs the data collection system 6 to set the resolution of the radiograph obtained by the area detector 3, thereby changing the resolution of the system. The data collection system 6 includes readout electronics as shown.

  The area-type detector 3 is constituted by an array (not shown) composed of detector elements. Each detector element measures an intensity value corresponding to the respective element related to the amount of X-ray energy incident on that detector element. By incorporating the apparatus and method of the present invention into a stereoscopic CT scanning system, a novel stereoscopic CT scanning system is created. Accordingly, a novel stereoscopic CT scanning system can be provided by the present invention.

  It should be noted that the present invention is not limited to any particular computer in order to perform data collection and perform processing according to the present invention. As used herein, the term “computer” is intended to represent any device capable of performing the calculations and calculations necessary to perform the processing according to the present invention. Is intended. Accordingly, the computer used to execute the control algorithm 10 according to the present invention can be any device capable of executing necessary processing.

  In the context of the present invention, it has been found that according to an alternative data smoothing scheme, it is not necessary to use additional detector elements to cover the transition region. In addition, one alternative method uses an iterative algorithm to estimate projection data that would be measured if it was assumed that the data was collected using a large detector array that covers the entire field of view. ing. Since errors in the transition region are handled in a similar manner, the two schemes will be considered below within the same context.

In this technique, a set of x-rays is obtained by either forward projecting the reconstruction data obtained from the preceding iteration step or by interpolating redundant detector data obtained from a set of oppositely directed rays. Form the projection {P _a }. Here, another set of projection data in the opposite direction to {P _a } is formed for the ray in the opposite direction. The forward projection technique is the process of emitting rays from a virtual X-ray source, which traverses the reconstructed volume toward individual detector elements. The line attenuation values of the values reconstructed along this ray are summed along the ray, and this is called the line integral of the line attenuation coefficient.

  Techniques for forward projection of reconstruction data (referred to as FPT) are generally suitable for creating projection data corresponding to a larger cone angle (ie, as used in a VCT system), while redundant projection data Is more suitable for projection data closer to the intermediate plane (that is, closer to the isocenter). Similar to the fan angle, the cone angle refers to the angular range of X-rays emitted from the X-ray source in a direction perpendicular to the fan angle direction. Distortion in the image near the isocenter due to the difference between the detector estimate obtained using either FPT or PDT and the original value that would have been obtained if the data were actually measured May occur.

  In order to reduce such distortion, a technique using a smoothing function has been developed. The smoothing function can be described as follows with reference to FIG.

1. For each projection view, select the known detector element closest to the isocenter (21). Hereinafter, this is represented by Va.
2. For the same detector element, an estimate is obtained (ie via interpolated projection data (PDT) from an alternative view or by forward projection (FPT) of the reconstruction data) (22). Hereinafter, this is represented by Vb.
3. An appropriate smoothing function is generated (23).
4). The difference between Va and Vb is reduced by a smoothing function, and this is smoothed in steps within a region near the center of the imaging area of the imaging system (24).

  This can be understood from the following considerations. That is, the amount of discontinuity of the projection data at the center of the shooting area is set as d = Va−Vb. As an example, one of the smoothing functions that can be used to smooth this difference in stages can be an exponential function defined by

V = 0.5 ^de -ax ₀ (Formula 1)
In the above equation, x ₀ is the absolute value of the distance from the detector position corresponding to the detector element value Va, and a is a coefficient for controlling the slope of the curve relating to this smoothing function. This exponential function is added (subtracted) to the projection value located on one side of the central ray position (detector position corresponding to the projection of the center of rotation of the imaging system onto the detector) and lower ( The exponential function is added (subtracted) to the projected value on the other side of the central ray position to lower (increase) the higher (lower) original value, while increasing (decreasing) the estimated value Yes. In other words, by combining the true projection data and the estimated projection data, it is possible to provide a method for reducing the mismatch of the projection data at the position of the central ray. Thus, this process can provide a smooth data transition region with reduced or eliminated artifacts (25).

  At present, there is no known prior art technique for using an area detector in a half detector shift configuration within a stereoscopic CT (VCT) system. Having described above various known techniques for removing discontinuities in projection data within the imaging system's field of view by creating a transition region, the VCT system has now been described with reference to another aspect of the present invention. I will explain.

Variable fθ And fθ ′ To represent the two functions that are obtained at the source angle θ and represent the signal strengths of the forward ray and the opposite ray (the ray with the same angle direction but the opposite transverse direction), respectively. To.

fθ (n) = 0 (when N ₂ <n <N) (Formula 2)
fθ ′ (n) = 0 (when 1 <n <N ₂ ) (Formula 3)
Since both of them cross the same position of the subject, ideally, fθ (N ₂ ) is fθ It should be exactly the same as ′ (N ₂ ). However, this cannot happen for the following reasons.

(A) The actual shape of each ray is a shallow tetrahedron with the source as the starting point and the detector element as the ending point. Except for the subject being homogeneous throughout and rotationally symmetric, There should be no identical rays crossing the same position.
(B) Subject / patient movement during the scan cycle introduces additional errors for each forward / opposite ray pair.
(C) To date, no efficient complete interpolation scheme has been developed. That is, errors are introduced by the interpolation process.

Fθ at the source angular position θ (N ₂ ) and fθ Assuming the difference of ′ (N ₂ ) is d (θ), if d (θ) is completely random, the error introduced into the reconstructed image is probably related to other random errors in CT. It should be obscured by quantum noise. However, if the error is systematic to some extent, obvious artifacts will be introduced in the reconstructed image. For this reason, it is necessary to use a smoothing process in the transition region. In other words, this smoothing function is expressed as fθ and fθ. It has been developed to make the step error of 'smaller around the central ray position represented by detector element N ₂ .

fθ And fθ Let the smoothing function for each of 'be expressed using W and W'. In deriving W and W ′, certain criteria must be considered as those skilled in the art will understand. Further, those skilled in the art will appreciate that various smoothing functions such as the following examples are suitable for this purpose other than those specifically listed herein. .

W (n) + W ′ (n) = 1 (for all n) (Formula 4)
δW / δn = δW ′ / δn = 0 (at the position of n = N ₂ ± Δn) (Formula 5)
In the above equation, Δn sets the degree of smoothness of W and W ′, and δ is a differential operator. It is further noted that the conventional smoothing function for this type of application is also commonly known as a feathering function. W and W ′ are used to find a transition region between forward and opposite rays. Note that Δn must be an integer greater than zero for the smoothing function to work correctly. In fact, the greater the Δn, the better the smoothing function works. However, if Δn is made too large, it is necessary to use an additional detector element so as to extend beyond N ₂ which is the detector element at the central ray position. Therefore, in selecting Δn, Δn must be large enough but not so large that it requires the addition of additional detector elements to the transition region. Therefore, fθ and fθ ′ Has the following limit:

fθ (n) = 0 (when N ₂ + Δn <n <N) (Formula 6)
fθ ′ (N) = 0 (when 1 <n <N ₂ −Δn) (Formula 7)
The actual detector signals used in the backprojection process are Wfθ and W′fθ. '. A wider transition region tends to eliminate much of the mismatching error between forward and opposite rays, so it is necessary to construct opposite rays for each half-field (FOV) projection data. Note further that there is no. In other words, each half-FOV data (and each additional Δn detector value) is zeroed to obtain detector data with a length of N, followed by the conventional filtered correction projection procedure. Interpolation is not necessary.

  Increasing the number of detector rows in the CT system, ie, the area detector, motivates more detector elements (Δn times the number of rows) to minimize Δn There is a need to improve upon conventional schemes by devising methods and apparatus that can be used.

  In the method according to the present invention, Δn is decreased to 1. On the other hand, computer simulations show that Δn should be approximately 20 in the conventional smoothing method in order to achieve an equivalent artifact level. This advantage becomes even more important in VCT applications where the number of detector elements required in the area detector transition region is on the order of three orders of magnitude greater than the number of elements required in the linear array.

There may be systematic errors in forward rays as well as in opposite directions obtained by interpolation or forward projection. fθ (N ₂ ) and fθ By detecting the difference between '(N ₂ ), the stepped error can be measured. Where d (θ) = fθ (N ₂ ) −fθ ′ (N ₂ ), d (θ) becomes fθ (N ₂ ) and fθ ′ (N ₂ ) is observed as a step error (where θ is the angular position of the X-ray source).

In this method, the stepped error is smoothed using the exponential function described in (Equation 1) (where a is a control coefficient for controlling the smoothing of the exponential function). Function of forward and opposite rays, fθ (N) and fθ ′ (N) is two different functions represented by the following equations, gθ (N) and gθ ′ (n).

gθ (n) = fθ (n) −pθ (N ₂ −n) (Formula 8)
gθ ′ (n) = fθ ′ (n) + pθ (n−N ₂ ) (Formula 9)
From (Equation 8) and (Equation 9), gθ (N ₂ ) = gθ '(N ₂ ).

The following procedure is executed for each projection image.
1. Padding zero according to (Equation 2), fθ Acquire original half-FOV projection data called (n).
2. Opposite ray set fθ '(N) is acquired and zeros are embedded according to (Equation 3).
3. In accordance with (Equation 8) and (Equation 9), a smoothing function is applied based on the stepped error d (θ) (where d (θ) = fθ (N ₂ ) −fθ ′ (N ₂ )).
4). gθ (N) and gθ ′ (N) is integrated to form one set of N detector data, that is, hθ (n). Here, when N ₂ <n <N, hθ (n) = gθ (n), and when 1 <n <N ₂ , hθ (n) = gθ ′ (n).
5. Apply conventional filtered back projection to hθ (n).
6). Repeat steps 1-5 for all projection angles.

  The above procedure is valid for any CT scanner that can acquire rays in the opposite direction by interpolating other projection data from various angles. In the case of a 2D fan beam in which another half-FOV data can always be roughly calculated from the redundant fan beam projection data, this is the complete case. If this technique is extended to 3D-VCT, complete interpolation can only occur on the intermediate plane when using circular orbits.

  When the same technique is applied to VCT using a circular orbit according to our simulations, this technique can be applied to cone angles in the range of ± 1.5 degrees (20 beyond the center ray position). It has been shown to perform better compared to conventional smoothing techniques (using an additional detector). However, this is not true for large cone angles. This is because the angular orientation of the opposite rays is significantly different from the flash data measured when using a perfect detector. To cope with this situation, iterative techniques that improve image quality can be used. The procedure is shown below.

1. An initial three-dimensional image is acquired according to steps 1 to 6 described above.
2. In the second iteration of the method, the forward projection method is used to obtain “opposite rays” for each set of semi-FOV projection data, and these are followed according to steps 3-6 outlined above. Combine to form a complete set of projection data.
3. Continue step 2 until this process has converged, ie until the quality of the image is no longer improved.

  It should be noted that the present invention has been discussed with respect to specific embodiments. Therefore, the present invention is not limited to these embodiments. For example, the three cases studied do not mean that all of the methods that can acquire the proper operation mode of the VCT system using the trade-off relationship of the above parameters are covered. . These cases have been considered to illustrate the concept of the present invention and how these basic parameters are in a trade-off relationship to achieve a proper scanning protocol. Furthermore, this trade-off relationship is not limited to a single scan protocol, ie, the trade-off relationship is for both axial and helical scan protocols (which do not move the patient table during the scan). Can be applied. Those skilled in the art will understand how to take advantage of these concepts and extend them to implement other area-based detector scanning protocols that are useful for specific applications.

It is a block diagram of CT system of the present invention. FIG. 4 is a diagram illustrating the deviation of a detector utilized in accordance with the method of the present invention. FIG. 2 is a block diagram illustrating the method of the present invention according to a preferred embodiment.

Claims (7)

A stereoscopic computed tomography (VCT) system for acquiring projection data of a subject,
An X-ray source that projects X-rays onto the subject;
A detector comprising a plurality of detector elements, receiving X-rays projected from an X-ray source and generating an electrical signal in response to incident X-rays, the rotation of the VCT system on the detectors A detector shifted by half the width of the detector with respect to the center position corresponding to the center projection; and
A data acquisition system that reads the electrical signal from the detector and converts it into a digital signal;
For each projection view, select the value Va of the detector element closest to the center position on the detector of the VCT system;
For each projection view, a value Vb corresponding to the value Va of the detector element of the detector is estimated, where the estimated value Vb is interpolated from projection data collected from the opposite direction, or the same An estimate of detector elements detected when the detector is a detector having a double width around the center position, through interpolation from reconstruction data of forward projections of directions ,
Once the value Va and value Vb are obtained for each projection view, a smoothing function is selected and applied to remove the difference between Va and Vb.
A computer configured to, and
A stereoscopic computer tomography system comprising:   The stereoscopic computer tomography system according to claim 1, wherein the detector is a high-resolution detector. Prior to applying the smoothing function, the computer is configured to weight the smoothing function;
Applying a weighted smoothing function to the detected or estimated detector element value reduces the difference between the value Va and the value Vb at the center position of the detector, and the center of the VCT system. Within the region near the position, the difference between the detected detector element value and the estimated detector element value is smoothed;
The three-dimensional computed tomography system according to claim 1 or 2,   The stereoscopic computer tomography system according to any one of claims 1 to 3, wherein the smoothing function is based on a half value of the difference between Va and Vb. A method for acquiring projection data of a subject using a computed tomography (CT) system comprising:
Projecting X-rays from an X-ray source onto a subject;
Receiving at the detector X-rays projected by the X-ray source in each projection view of the detector, the detector comprising a plurality of detector elements and incident on the detector; In response to the detector, wherein the detector is shifted by half the width of the detector relative to the center position corresponding to the projection of the center of rotation of the CT system onto the detector. Is a step,
Digitizing the electrical signal;
For each projection view, selecting the value Va of the detector element of the detector closest to the central position of the CT system;
For each projection view, the detector element value Vb associated with the detector element value Va of the detector is interpolated from projection data collected from the opposite direction, or reconstruction data for forward projection in the same direction. Estimating via interpolation from:
Selecting a smoothing function that can eliminate the difference between Va and Vb;
Reducing the difference between the value Va and the value Vb when combining the value Va and the value Vb to generate a smooth transition region using a smoothing function;
Including methods.   6. The method of claim 5, wherein the step of utilizing the smoothing function is performed by a weighted smoothing function.   The method according to claim 5 or 6, characterized in that the smoothing function is based on half the difference between the Va and Vb.