CN100453044C - CT beam hardening correction method based on original projected sinogram - Google Patents

Abstract

The CT beam hardening correction method based on the original projected sinogram, the steps are: (1) perform three generations of CT scans to obtain the original projected sinogram arranged in a matrix; (2) perform dark current and inconsistency correction on the original projected sinogram; (3) the minimum value of each row of pixel projection values in the projection sinogram that search step (2) obtains; (4) determine a group of positive real number correction coefficients less than 1; (5) use the projection sinogram that step (2) obtains The projection value of each row of pixels is subtracted from the product of the corresponding row minimum value obtained in step (3) and the corresponding row correction coefficient obtained in step (4) to complete the beam hardening correction of all row projections; (6) project the sine The image is sent to the CT reconstruction dynamic library. The invention does not need to perform modeling scanning and calculation in advance, is not limited by scanning objects and scanning conditions, and has a simple and real-time correction program without specific correction hardware.

Description

CT beam hardening correction method based on original projected sinogram

technical field

The invention relates to a CT beam hardening correction method, in particular to a CT beam hardening correction method based on an original projected sinogram, and belongs to the CT technical field.

Background technique

In the X-ray CT system, the X-ray source emits X-rays, which pass through a certain cross-section of the detected object from different angles, and the detector placed opposite the ray source receives them at corresponding angles, and then attenuates the rays to different degrees according to each angle. , using a certain reconstruction algorithm and computer operations to reconstruct the ray attenuation coefficient distribution mapping image of the detected section of the object, so as to realize the reconstruction of the image by projection and non-destructively reproduce the medium density, composition and structural shape of the object in the section. feature.

The CT reconstruction algorithm assumes that the X-ray source emits a single-energy X-ray beam. Real CT systems have polychromatic X-ray beams, and this assumption is not satisfied. When a polychromatic X-ray beam interacts with matter, the lower-energy X-ray photons are preferentially absorbed by the matter, and the higher-energy photons are less attenuated. As the transillumination thickness increases, the effective energy of the beam moves to the direction of higher energy. , the beam becomes harder to attenuate and "harder", this is beam hardening. Beam hardening poses serious problems for CT reconstruction. Using the reconstruction algorithm based on the monochromatic ray assumption, the reconstructed object line attenuation coefficient is smaller than the actual value, and the reconstructed image will produce black center artifacts (also known as cup artifacts). In medicine, these artifacts often lead to missed diagnosis of diseased tissues such as tumors; in industry, they cover up defects such as tiny cracks and looseness.

A method for suppressing hardening artifacts is called a beam hardening correction method. Beam hardening correction has been a research hotspot in the field of international CT imaging since 1975. At present, the calibration methods are mainly divided into two categories: single-energy and dual-energy. In the dual-energy method, the linear attenuation coefficient is represented by Compton scattering, photoelectric effect and cross-sectional material distribution function, and the monochromatic projection is estimated by two polychromatic projections under different energies, so as to achieve the purpose of correction. Due to the operational complexity, this method is rarely used in engineering practice. The single-energy method only needs to scan the object once, is easy to realize, and has better practical application effect, so it has been extensively studied.

Single-energy methods can be divided into pre-processing and post-processing according to the different processing sequences. The pre-processing method directly corrects the original projected sinogram, and then uses the monochrome reconstruction algorithm to reconstruct it; the post-processing method first uses the monochrome reconstruction algorithm to reconstruct the original projected sinogram, and then performs hardening correction on the reconstructed image .

The basic route of the post-processing correction method is: according to certain criteria, perform threshold segmentation on the reconstructed image to determine the scanning angle and area where hardening will occur; then, at the above scanning angle and area, use reprojection technology to generate a projection to replace the corresponding data of the original projected sinogram; finally, the corrected projected sinogram is reconstructed using a monochromatic reconstruction algorithm. Post-processing methods have disadvantages such as time-consuming calculation and loss of image information. Although many post-processing methods have appeared in the world, pre-processing methods should also be considered for practical applications.

The general route of the pre-processing method is: establish a hardening curve for a certain detection object by scanning some special phantoms; calculate the correction function according to a certain criterion; adjust the projected sinogram according to the function to complete the hardening correction. The fitting method proposed in Yang Min et al., Research on Correction of Radiation Hardening in CT Reconstruction, Optical Technology, 2003, Vol.29(2): 177-182. is a typical preprocessing beam hardening method, and its correction idea is: Under a certain detection condition, the multicolor beam is used to transilluminate the same material with different thicknesses, and the relationship curve between the multicolor beam transillumination data and the transillumination thickness is established, and then the curve is fitted, and then from the coordinates A tangent is made to the curve at the origin, and the tangent is used as the relationship between the monochromatic beam transillumination data and the transillumination thickness, so as to establish the correction relationship between the actual polychromatic beam data and the monochromatic beam data. Under certain conditions, this method can produce a good correction effect on objects composed of a single material, but it has the following defects: (1) The correction procedure is cumbersome, and it needs to be modeled, scanned and calculated separately for different scanning objects; (2) The validity of the correction model depends on the scanning conditions. If the conditions change, the modeling scan and calculation need to be re-executed; (3) The correction model is affected by noise, and the signal-to-noise ratio of the corrected image decreases; (4) The corrected image must be scanned by a single material composition.

Contents of the invention

The technical problem to be solved by the present invention is to provide a CT beam hardening correction method based on the original projection sinogram, which belongs to the beam of the third generation of transmission X-ray CT, aiming at the above-mentioned shortcomings existing in the current beam hardening correction method. Hardening correction method, this method has high correction quality, does not need modeling scanning and calculation in advance, is not limited by scanning objects and scanning conditions, and the correction procedure is simple and real-time, and does not require specific correction hardware.

The technical solution of the present invention: the CT beam hardening correction method based on the original projected sinogram, is characterized in that it comprises the following steps:

(1) Perform three generations of CT scans to obtain the original projection sinograms arranged in a matrix;

(2) Perform dark current and inconsistency correction of the original projected sinogram;

(3) search for the minimum value of the pixel projection value of each row of the projected sinogram;

(4) determine a group of positive real number correction coefficients less than 1;

(5) Subtract the product of the corresponding row minimum value obtained in step (3) and the corresponding row correction coefficient obtained in step (4) from the projection value of each row of pixels of the projected sinogram, and complete all row projection beam hardening corrections;

(6) Send the corrected projected sinogram into the CT reconstruction dynamic library.

The principle of the present invention is based on the following three physical facts and two mathematical theorems:

Physical fact 1: When beam hardening occurs, at the i-th projection angle, the intensity I ₂ (i, j) of the ray collected by the detector j after being attenuated by the object is greater than the ideal intensity I ₁ (i, j). The (I ₀ (i)/I ₂ (i,j)) used for reconstruction will be smaller than the true (I ₀ (i)/I ₁ (i,j)). Wherein, I ₀ (i) represents the i-th projection angle without attenuating ray intensity.

Physical fact 2: When beam hardening occurs, the longer the beam passes through the object, the more severe the hardening, and the more the radiation intensity received by the detector deviates from the ideal intensity. It can be expressed mathematically as:

I ₂₁ (i, j)/I ₂₂ (i, j)>I ₁₁ (i, j)/I ₁₂ (i, j)

I₁₁(i，j)为穿过物体路径较短时，探测器接受的理想射线强度；

I ₁₁ (i, j) is the ideal ray intensity received by the detector when the path through the object is short;

I ₂₁ (i, j) is the ideal ray intensity received by the detector when the path through the object is long;

I ₁₂ (i, j) is the actual ray intensity received by the detector when the path through the object is short;

I ₂₂ (i, j) is the actual ray intensity received by the detector when the path through the object is long;

Physical fact 3: The same object section has the same amount of attenuation of rays under different scanning angles. It can be expressed mathematically as:

$\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; I</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

I(i，j)为第i个扫描角度，探测器j接受的射线强度；

I(i, j) is the i-th scanning angle, the ray intensity received by detector j;

N is the number of detectors; M is the number of scanning angle divisions.

Mathematical theorem 1: Suppose A, B, C∈R ⁺ , and A>B>(C+1)>1, then there is the following relationship: ((AC)/(BC))>(A/B).

Proof: By assuming AC>BC, therefore, -AC<-BC, therefore, AB-AC<AB-BC,

So A(B-C)<B(A-C), therefore, A/B<(A-C)/(B-C);

The conclusion is proven.

Mathematical theorem 2: Suppose A, B, C, D∈R+, and A>B>(C+1)>1, B<D, then there is the following relationship: ((A-C)/(B-C))/(A/ B)>((A-C)/(D-C))/(A/D).

Proof: By assuming that DC>BC, so, -DC<-BC, so, BD-BC>BD-CD,

So AB(D-C)>AD(B-C), therefore, there is B/A(B-C)>D/A(D-C)

So, ((A-C)/(B-C))/(A/B)>((A-C)/(D-C))/(A/D);

The conclusion is proven.

Analyzing the above physical facts and mathematical theorems, it can be seen that A, B, and D in the above theorems are regarded as I ₀ (i), I ₂₂ (i, j), and I ₁₂ (i, j) in the actual projection, and C is regarded as A correction value related to the original projection, then the mathematical theorem 1, 2 just corrects the hardening phenomenon caused by the physical fact 1, 2. Therefore, subtracting a value related to the original projection from the actual hardened projection can better approximate the ideal projection value. According to this understanding, propose the correction method of the present invention as follows:

I _C (i, j) = I _O (i, j) - K (i) × min (I _O (i, j))

Wherein, I _O (i, j) represents the ray intensity signal collected by the jth detector under the ith scanning angle before hardening correction;

I _C (i, j) represents the ray intensity signal collected by the jth detector at the ith scanning angle after hardening correction;

K(i) represents the hardening correction coefficient at the i-th scanning angle; min() represents the minimum value for the one-dimensional projection sequence.

Determine K(i) according to physical fact 3, as follows:

$\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; C</span>}<span\; class="notranslate">\; \&times;</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; o</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; max</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; S</span>}<span\; class="notranslate">\; )</span>$ $\mathrm{<span\; class="notranslate">\; S</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; o</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

Wherein, I _O (i, j) represents the ray intensity signal collected by the jth detector under the ith scanning angle before hardening correction;

max() means to take the maximum value for a one-dimensional array;

C: constant between [0.5 0.8], selected according to experience.

The advantages of the present invention compared with prior art are as follows:

(1) Since the present invention extracts hardening information from the original projection of each scanning angle, and determines the respective correction coefficients of each scanning angle, the correction model has higher validity and high correction quality;

(2) The correction process only involves the addition and subtraction of the original projection, so the correction speed is higher and the correction procedure is real-time;

(3) It does not need to carry out modeling scanning and calculation in advance, it is not limited by the scanning object and scanning conditions, and does not require specific calibration hardware, so the calibration is simpler;

(4) The correction does not need to be modeled in advance, so it is less affected by projection noise, and the correction signal-to-noise ratio is higher.

Description of drawings

Fig. 1 is the flowchart of hardening correction method of the present invention;

Fig. 2 is the CT image reconstructed using the projected sinogram without beam hardening correction and the filtered back-projection reconstruction algorithm when the scanned object is composed of only one material (aluminum);

Fig. 3 is when the scanning object is only made of a kind of material (aluminum), adopts the projection sinogram after the correction of the present invention and filter back-projection reconstruction algorithm, reconstructs the CT image that obtains;

Fig. 4 is the CT image obtained by reconstructing the projected sinogram and the filtered back-projection reconstruction algorithm corrected by the fitting ray hardening correction method of the comparative literature in the background technology;

Fig. 5 is the comparison curve of the reconstructed gray value of the corresponding row of the reconstructed image in Fig. 3 and Fig. 4;

Figure 6 shows the projection sinogram and filtered back projection reconstruction without beam hardening correction when the scanned object is composed of three substances (aluminum, pure water with a density of 1.000g/ ^cm3 , and saline with a density of 1.005g/ ^cm3 ) Algorithm, reconstructing the obtained CT image;

Fig. 7 shows that when the scanned object is composed of three substances (aluminum, pure water with a density of 1.000g/ ^cm3 , and salt water with a density of 1.005g/ ^cm3 ), the corrected projected sinogram and filtered back-projection reconstruction algorithm of the present invention are used. The reconstructed CT image;

Fig. 8 is a comparison curve of the reconstructed gray value of corresponding lines of the reconstructed images in Fig. 3 and Fig. 4 .

Detailed ways

As shown in Figure 1, the specific implementation steps of the present invention are as follows:

(1) Place the object to be scanned on the rotating inspection table of the third-generation CT scanning system to ensure that the object is covered by the fan beam at any scanning angle;

(2) Transilluminate the object with the collimated fan-beam rays. At the same time, the detection platform rotates continuously at a constant speed. Ray projection of the object to obtain multiple sets of one-dimensional signals;

(3) When the inspection table rotates 360 degrees, the detector stops sampling, and the inspection table and the radiation source stop at the same time, that is, a third-generation CT scan is completed;

(4) Stacking multiple sets of one-dimensional signals obtained in step (2) and arranging them into a matrix to form an original CT projection sinogram. Among them, a column of the matrix represents the data collected by the same detection channel within 360 degrees, and a row of the matrix represents the data collected by all detection channels at a certain scanning angle;

(5) When there is no ray, according to the method of steps (2), (3) and (4), a dark field projection map is formed, and the dark field projection map is averaged by columns to obtain a one-dimensional dark field projection array D;

(6) Remove the object from the scanning table to ensure that there is no object between the ray source and the line array detector;

(7) According to the method of steps (2), (3) and (4), a bright field projection map is formed, and the bright field projection map is averaged by columns to obtain a one-dimensional bright field projection array L;

(8) Subtract D from L to complete the bright field dark current correction to obtain L1;

(9) Calculate the average value of L1, and divide each value of L1 by the average value to obtain a one-dimensional array U;

(10) Subtracting D from each line of data in the projection image obtained in step (4) to complete the dark current correction;

(11) Multiply the data of each line of the projection map obtained in step (10) by U to complete the inconsistency correction.

(12) the minimum value of each row of pixel projection value of the projected sinogram that search step (11) obtains, they are stored in one-dimensional array A;

(13) The sum of all pixel projection values of each row of the projected sinogram obtained in the calculation step (11) is stored in a one-dimensional array B;

(14) Divide each value of array B by the maximum value of array B, and store the result in array B;

(15) Array B is multiplied with the corresponding data of array A, and the result is stored in array A;

(16) Subtract the i-th row of the projection sinogram (i is a positive integer from 1 to N, and N is the total number of rows of the projection sinogram) from all pixel projection values of the array A (i is a positive integer from 1 to N) Integer, N is the total number of array A data) data;

(17) Repeat above-mentioned steps (16), until all pixel projection values of the last line of projected sinogram subtract the last data of array A, form the projected sinogram after correction;

(18) Send the projection sinogram above into the CT reconstruction dynamic library.

Figures 2-5 show specific examples of corrections where the object to be detected consists of only one substance (aluminum). Fig. 2 is the CT image reconstructed using the projected sinogram and the filtered back-projection reconstruction algorithm without beam hardening correction; Fig. 3 is the projected sinogram and the filtered back-projection reconstruction algorithm using the beam hardening correction of the present invention, reconstructed to obtain CT image; Fig. 4 is the projection sinogram and filtered back-projection reconstruction algorithm corrected by the fitting ray hardening correction method of the comparative literature in the background technology, and the reconstructed CT image is obtained; Fig. 5 is the corresponding row of Fig. 3 and Fig. 4 Reconstructed grayscale curve comparison. The dotted line in Figure 5 is the gray value of the pixel in the 256th row in Figure 2; the thick solid line is the gray value of the pixel in the 256th row in Figure 3; the thin solid line is the gray value of the pixel in the 256th row in Figure 4. Comparing the three curves shows that the correction effect of the present invention is the best. Calculated, the signal-to-noise ratios of Fig. 2, Fig. 3 and Fig. 4 are respectively: 24.9, 38.6 and 12.4, indicating that the present invention has the highest signal-to-noise ratio after correction.

Fig. 6-Fig. 8 have provided the concrete correction example that the detected object is made of three substances (aluminum, pure water with a density of 1.000g/cm3, and salt water with a density of 1.005g/cm3). Fig. 6 is a CT image reconstructed using a projected sinogram without beam hardening correction and a filtered back-projection reconstruction algorithm; Fig. 7 is a reconstructed CT image obtained using a beam-hardening corrected projected sinogram and a filtered back-projection reconstruction algorithm of the present invention CT image; Figure 8 is a comparison of the reconstructed gray-scale curves of the corresponding lines in Figure 6 and Figure 7. The dotted line in Fig. 8 is the gray value of the pixel in the 256th row in Fig. 6; the thick solid line is the gray value of the pixel in the 256th row in Fig. 7 . Comparing the two curves, it can be seen that the present invention has an obvious correction effect on beam hardening for objects composed of multiple substances.

Claims (4)

1, based on the CT beam hardening correction method of original projection sinogram, it is characterized in that comprising the following steps:

(1) carries out three generations's CT scan, obtain to be arranged in the original projection sinogram of matrix;

(2) dark current and the discordance of carrying out original projection sinogram proofreaied and correct;

(3) minima of each row pixel projection value in the projection sinogram that after overcorrect, obtains of search step (2);

(4) determine one group less than 1 arithmetic number correction coefficient;

(5) projection value of each row pixel of the projection sinogram that step (2) is obtained after overcorrect deducts the product of the corresponding row correction coefficient that corresponding row minima that step (3) obtains and step (4) obtain, and finishes all row projection pencil sclerosis corrections;

(6) projection sinogram is sent into CT and is rebuild dynamic base after the correction that step (5) is obtained.

2, the CT beam hardening correction method based on original projection sinogram according to claim 1, it is characterized in that: described step is carried out three generations's CT scan in (1), and the step that acquisition is arranged in the original projection sinogram of matrix is:

(a) scanned object is positioned over three generations's CT scan system rotation inspection platform, guarantees that object is covered by fan-beam under arbitrary scanning angle;

(b) with the fan-beam ray that forms through collimation object is implemented transillumination, simultaneously, the inspection platform rotates at the uniform velocity continuously, the detector that is made of a plurality of detection channels that are arranged in linear array, cross the ray projection of object with the transmission of fixed sample speed continuous acquisition, obtain many group one-dimensional signals;

(c) when the inspection platform revolved three-sixth turn, detector stopped sampling, and inspection platform and radiographic source stop simultaneously, promptly finish three generations's CT scan one time;

(d) many groups one-dimensional signal that step (b) is obtained piles up, and is arranged in matrix, forms the original CT projection sinogram, wherein, the data that on behalf of same detection channels, the string of matrix gather in 360 degree, the delegation of matrix represents under a certain scanning angle, the data that all detection channels are gathered.

3, the CT beam hardening correction method based on original projection sinogram according to claim 1 is characterized in that: the dark current and the gauged step of discordance of carrying out original projection sinogram in the described step (2) are:

(a) scanned object is positioned over three generations's CT scan system rotation inspection platform, guarantees that object is covered by fan-beam under arbitrary scanning angle; During no ray, the inspection platform rotate at the uniform velocity continuously, by the detector that a plurality of detection channels that are arranged in linear array constitute, with the continuous acquisition projection under no ray situation of fixed sample speed, organizes one-dimensional signals obtain more; When the inspection platform revolved three-sixth turn, detector stopped sampling, and the inspection platform stops; The many groups one-dimensional signal that obtains is piled up, be arranged in matrix, form the details in a play not acted out on stage, but told through dialogues projection, wherein, the data that on behalf of same detection channels, the string of matrix gather in 360 degree, the delegation of matrix represents under a certain scanning angle, the data that all detection channels are gathered; The details in a play not acted out on stage, but told through dialogues projection is averaged by row, obtain one dimension details in a play not acted out on stage, but told through dialogues projection array D;

(b) object is removed from scan table, guaranteed no any object between radiographic source and linear array detector;

(c) with the fan-beam ray that forms through collimation detector is implemented irradiation, simultaneously, the inspection platform rotates at the uniform velocity continuously, and the detector by a plurality of detection channels that are arranged in linear array constitute with fixed sample speed continuous acquisition ray projection, obtains many group one-dimensional signals; When the inspection platform revolved three-sixth turn, detector stopped sampling, and inspection platform and radiographic source stop simultaneously; The many groups one-dimensional signal that obtains is piled up, be arranged in matrix, form the bright field projection, wherein, the data that on behalf of same detection channels, the string of matrix gather in 360 degree, the delegation of matrix represents under a certain scanning angle, the data that all detection channels are gathered; The bright field projection is averaged by row, obtain one dimension bright field projection array L;

(d) L is deducted D, finish the bright field dark current correction, obtain L1;

(e) ask the meansigma methods of L1,, obtain one-dimension array U with meansigma methods each value divided by L1;

(f) the projection each row of data that step (1) is obtained deducts D, finishes dark current correction;

(g) each row of data of projection multiply by U behind the dark current correction that step (f) is obtained, and finishes discordance and proofreaies and correct.

4, the CT beam hardening correction method based on original projection sinogram according to claim 1 is characterized in that, the step of beam hardening correction is in described step (3), (4) and (5):

(a) there is one-dimension array A in the minima of each row pixel projection value of the projection sinogram obtained of search step (2) with them;

(b) each all pixel projection value of row of the projection sinogram obtained of calculation procedure (2) and, there is one-dimension array B in they;

(c) there is array B divided by array B maximum in each value of array B with the result;

(d) array B and array A corresponding data multiply each other, and there is array A in the result;

(e) capable all the pixel projection values of i with projection sinogram deduct i data of array A, and wherein, i is the positive integer from 1 to N, and N is the total line number of projection sinogram;

(f) repeating step (e) deducts last data of array A up to all pixel projection values of last column of projection sinogram, forms to proofread and correct the back projection sinogram.

Images