CN114391857A - Dual-energy X-ray bone mineral density detection method based on mobile least square algorithm - Google Patents
Dual-energy X-ray bone mineral density detection method based on mobile least square algorithm Download PDFInfo
- Publication number
- CN114391857A CN114391857A CN202210040299.8A CN202210040299A CN114391857A CN 114391857 A CN114391857 A CN 114391857A CN 202210040299 A CN202210040299 A CN 202210040299A CN 114391857 A CN114391857 A CN 114391857A
- Authority
- CN
- China
- Prior art keywords
- energy
- ray
- die body
- low
- photon counting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 59
- 210000000988 bone and bone Anatomy 0.000 title claims abstract description 27
- 229910052500 inorganic mineral Inorganic materials 0.000 title claims abstract description 19
- 239000011707 mineral Substances 0.000 title claims abstract description 19
- 238000001514 detection method Methods 0.000 title claims abstract description 18
- 230000037182 bone density Effects 0.000 claims abstract description 50
- 230000008859 change Effects 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 63
- 238000000034 method Methods 0.000 claims description 52
- 239000011159 matrix material Substances 0.000 claims description 20
- 238000012937 correction Methods 0.000 claims description 14
- 230000003287 optical effect Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 8
- 150000001875 compounds Chemical class 0.000 claims description 6
- 239000004417 polycarbonate Substances 0.000 claims description 6
- 229920000515 polycarbonate Polymers 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 5
- 239000013078 crystal Substances 0.000 claims description 4
- 230000009977 dual effect Effects 0.000 claims description 4
- 229910052698 phosphorus Inorganic materials 0.000 claims description 4
- 230000002441 reversible effect Effects 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 230000002238 attenuated effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000002474 experimental method Methods 0.000 claims description 3
- QWUZMTJBRUASOW-UHFFFAOYSA-N cadmium tellanylidenezinc Chemical compound [Zn].[Cd].[Te] QWUZMTJBRUASOW-UHFFFAOYSA-N 0.000 claims description 2
- 238000004088 simulation Methods 0.000 claims 1
- 230000004807 localization Effects 0.000 abstract description 5
- 230000008569 process Effects 0.000 description 10
- 230000008901 benefit Effects 0.000 description 8
- 210000004872 soft tissue Anatomy 0.000 description 8
- 208000001132 Osteoporosis Diseases 0.000 description 5
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 5
- 238000003745 diagnosis Methods 0.000 description 4
- 238000010521 absorption reaction Methods 0.000 description 3
- 238000001739 density measurement Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- OYPRJOBELJOOCE-UHFFFAOYSA-N Calcium Chemical compound [Ca] OYPRJOBELJOOCE-UHFFFAOYSA-N 0.000 description 2
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 239000011575 calcium Substances 0.000 description 2
- 229910052791 calcium Inorganic materials 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 229910052588 hydroxylapatite Inorganic materials 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- XYJRXVWERLGGKC-UHFFFAOYSA-D pentacalcium;hydroxide;triphosphate Chemical compound [OH-].[Ca+2].[Ca+2].[Ca+2].[Ca+2].[Ca+2].[O-]P([O-])([O-])=O.[O-]P([O-])([O-])=O.[O-]P([O-])([O-])=O XYJRXVWERLGGKC-UHFFFAOYSA-D 0.000 description 2
- 239000011574 phosphorus Substances 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 208000020084 Bone disease Diseases 0.000 description 1
- 206010065687 Bone loss Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 229910052793 cadmium Inorganic materials 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000001054 cortical effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000009643 growth defect Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 210000001694 thigh bone Anatomy 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/42—Arrangements for detecting radiation specially adapted for radiation diagnosis
- A61B6/4208—Arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector
- A61B6/4241—Arrangements for detecting radiation specially adapted for radiation diagnosis characterised by using a particular type of detector using energy resolving detectors, e.g. photon counting
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
- A61B6/505—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications for diagnosis of bone
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5205—Devices using data or image processing specially adapted for radiation diagnosis involving processing of raw data to produce diagnostic data
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5211—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data
- A61B6/5217—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data extracting a diagnostic or physiological parameter from medical diagnostic data
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Radiology & Medical Imaging (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- High Energy & Nuclear Physics (AREA)
- Veterinary Medicine (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Optics & Photonics (AREA)
- Pathology (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Orthopedic Medicine & Surgery (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a dual-energy X-ray bone mineral density detection method based on a mobile least square algorithm, which comprises the following steps: acquiring detection data of a die body under high-energy and low-energy X rays of a photon counting detector, and carrying out preprocessing including denoising; and acquiring incident energy of the high-energy and low-energy X-rays; calculating to obtain high and low energy attenuation values; carrying out grid division according to the thickness of the die body; after a weight function of a moving least square algorithm is determined, in each grid, the radius of an influence domain is divided by taking a target node as a center, in a spatial domain, each high-low energy attenuation value data point is traversed, and a change step length is set in the influence domain; obtaining a final fitting relation by adopting a least square algorithm to obtain a die body thickness-high and low energy attenuation value fitting curved surface; and determining a bone density ROI area to be measured, and calculating to obtain the corresponding die body thickness, wherein the product of the die body thickness and the die body density is the bone density value of the corresponding part. The invention solves the problems of fitting smoothness and localization.
Description
Technical Field
The invention belongs to the technical field of medical instruments, and particularly relates to the field of bone mineral density measurement by applying dual-energy X-ray.
Background
Osteoporosis disease is one of the six most common diseases in the world. However, the disease has no obvious symptoms in the early stage and does not attract attention of people, so that how to prevent and diagnose the osteoporosis disease is very important. Bone density measurement is currently an important method for assessing bone loss and diagnosing osteoporosis. The skeleton of the human body is composed of dense cortical bone and a plurality of spongy cancellous bone in the medullary cavity, the main component of which is hydroxyapatite, mainly existing in calcium and phosphorus elements. Osteoporosis is the loss of calcium and phosphorus elements, which causes the content of bone mineral substances to be reduced, and the main medical measurement of bone density is the density of hydroxyapatite.
The double-energy X-ray absorption method is a method for measuring bone density most effectively by the certification of the world health organization at present, and mainly comprises the steps of generating two X-rays with different energies by an X-ray bulb tube through a switching pulse method or a K-edge filtering method, attenuating the X-rays after passing through a die body of an equivalent human skeleton and soft tissue, receiving the attenuated photon energy by a detector, obtaining the corresponding relation between the bone density and the thickness of the die body through a computer and a related ASIC integrated circuit and an algorithm, and displaying the bone density of an ROI area by a screen. The dual-energy X-ray absorption method has the advantages of high scanning speed, high precision and low radiation, can measure the skeleton range of the whole body such as the vertebra, the thighbone, the lumbar vertebra and the like of a human body, and has higher accuracy and precision.
After the dual-energy X-ray bone density diagnosis system collects the attenuated photon energy through the detector, the X-ray attenuation energy and the thickness data of bones and soft tissues need to be further processed by an algorithm so as to carry out subsequent bone density information analysis and diagnosis. The traditional bone mineral density algorithm mainly comprises a lookup table method and a polynomial fitting method. The lookup table method is to correspond high and low energy attenuation values to different bone thicknesses one by one, and find out relevant thickness values according to a corresponding table by scanning X-ray attenuation energy obtained by a bone and soft tissue die body. The polynomial fitting method is to combine the material decomposition algorithm and the attenuation formula of X-ray to further deduce, use Taylor formula to expand, set the changing step length, and carry out first-order derivation to find the minimum value of the objective function. And obtaining a new objective function according to the step change, continuously performing first-order derivation on the new objective function, and repeating the steps until all the discrete data points are traversed and ended. The relation between the attenuation value of high energy and low energy and the thickness of the object can be obtained.
The commonly used traditional algorithm of the bone density diagnostic system of the dual-energy X-ray, such as a lookup table method, is to establish a database, correspond the high and low energy attenuation values of the collected X-ray with bones and soft tissues with different thicknesses, find the corresponding thickness according to the attenuation values of the X-ray when measuring the bone density, and then perform the linear conversion of the thickness-the bone density, so as to obtain the bone density value of the corresponding part. The method has long running time and needs large data memory. The polynomial fitting method carries out approximate expression through quadratic or cubic polynomial, and the calculation process uses a gradient method or a Newton iteration method for evaluation. The polynomial fitting calculation amount is large, the calculation process is complicated, the requirement on the fitting coefficient is high, otherwise, the fitting surface is difficult to achieve the best and the error is easy to increase.
The dual-energy X-ray bone densitometer based on the photon counting detector is a photon counting detector adopting a tellurium-zinc-cadmium crystal, the photon counting detector can fully utilize X-ray energy spectrum information under ideal conditions, the image quality is effectively improved, material composition information is obtained, and the soft tissue contrast is enhanced, but due to the fact that the manufacturing process is not mature, dead spots of the photon counting detector easily appear on the edge of a module, and non-uniformity exists among pixel units, the mutual influence is easy to occur, and the non-uniformity among the detector pixel units has great influence on the error of a traditional algorithm.
The moving least squares algorithm is most commonly applied in mechanics, remote sensing geography. Based on the traditional bone mineral density algorithm, in order to improve the bone mineral density fitting precision, researchers strengthen the research on system correction and correction die body materials. The pixel nonuniformity of the photon counting detector has great influence on the error of the traditional bone density algorithm, and the moving least square algorithm aims to solve the edge nonuniformity of the photon counting detector and realize better fitting and correction by the special compactness of the edge nonuniformity. The moving least square method has excellent performance in two-dimensional scatter point fitting, bone mineral density measurement relates to three-dimensional data, two independent variables are required to be set in a formula, and a matrix in a numerical calculation process is a nonsingular matrix after a weight function is determined, namely the matrix can be successfully fitted in a three-dimensional space only after being reversible.
Disclosure of Invention
The present invention is directed to solving the above problems of the prior art. A dual-energy X-ray bone mineral density detection method based on a moving least square algorithm is provided. The technical scheme of the invention is as follows:
a dual-energy X-ray bone mineral density detection method based on a moving least square algorithm comprises the following steps:
acquiring detection data of a die body under high-energy and low-energy X rays of a photon counting detector, and carrying out preprocessing including denoising; and acquiring incident energy of the high-energy and low-energy X-rays;
calculating to obtain a high-energy and low-energy attenuation value according to the preprocessed detection data and the incident energy;
performing grid division on discrete high and low energy attenuation value data points according to the thickness of the die body;
after a weight function of the moving least square algorithm is determined, in each grid, the radius of an influence domain is divided by taking a target node as a center, in a spatial domain, traversal is carried out on data points of each high-low energy attenuation value, and a change step length is set in the influence domain;
obtaining a final fitting relation by adopting a moving least square algorithm, and obtaining a die body thickness-high and low energy attenuation value fitting curved surface after all nodes are traversed;
determining a bone density ROI area to be measured, scanning a part to be measured by a photon counting detector to obtain an X-ray attenuation value, calculating to obtain the corresponding die body thickness, and performing linear transformation once again, wherein the product of the die body thickness and the die body density is the bone density value of the corresponding part.
Further, the calculating the attenuation values of high and low energy according to the preprocessed detection data and the incident energy specifically includes:
obtaining X-ray attenuation values I at high and low energy levels by formulas (1) and (2) respectivelyH、ILData collected by detectors respectively representing high-energy and low-energy X-rays under a placed die body, IOH、IOLRepresenting the incident energy of high-and low-energy X-rays, yH、yLRespectively refers to the attenuation degree of the photon energy of the X-ray under the conditions of high energy and low energy;
furthermore, the types of the weight function of the moving least square method comprise a Gaussian weight function, a cubic spline weight function and a double-interpolation weight function, and experiments prove that the fitting effect of the Gaussian weight function is optimal.
Further, an expression of the moving least squares algorithm is shown in (3), where α (x) is a coefficient vector matrix, p (x) is a basis function vector, i represents an actual data point, f (x) is a fitting function, a norm is solved for the formula (3) to obtain a formula (4), w (x) is a weight function, x is a target node, and x is a target nodeiFor other discrete data points, J represents the norm of equation (3), yIFor the actual data point, in order to obtain the minimum value of the actual data point, the formula (4) calculates the partial derivative of alpha (x) to obtain (5);
expanding the matrix A, B in the formula (5), (6) and (7), it can be seen that the two matrices are coefficient matrices formed by basis functions and weight functions;
A(x)=PTW(x)P (6)
B(x)=PTW(x) (7)
yT=[y1,y2,...,yn] (8)
w (x) is a weight function, P represents a basis function, ynRepresenting the actual X-ray attenuation values.
The basis function is a K-order polynomial, the K-order polynomial is expanded as shown in (9), the weight function is a diagonal matrix, the expression of the diagonal matrix is shown in (10), the formula (5) is deformed and is substituted into the formula (3), and finally the standard formula of the moving least square is obtained as (11);
f(x)=pTA-1(x)B(x)y (11)
p is a basis function, pm(xn) The number of x is represented by the nth value of x under the m-order complete polynomial, m represents the order of the basis function P, and n represents the number of x.
Since the data are three-dimensional space fitting, the moving least square method is adopted as shown in formulas (12), (13), wherein x and z represent the thicknesses of the module Al and the polycarbonate high molecular compound respectively, and fLRepresents the fitting value at low energy, fHFor fitting values at high energy, carry in yL、yHThe fitting relation of the die body thickness under high energy and low energy can be respectively obtained
fL(x,z)=pT(x,z)A-1(x,z)B(x,z)yL (12)
fH(x,z)=pT(x,z)A-1(x,z)B(x,z)yH (13)
Further, the detection data of the mold body under the high-energy and low-energy X-ray of the photon counting detector is obtained, and preprocessing including denoising is performed; and acquiring incident energy of the high-energy and low-energy X-rays, specifically comprising:
an optical platform is built for collecting X-ray high and low energy attenuation values of a die body, and the optical platform mainly comprises the following parts: the system comprises an X-ray machine, a photon counting detector, an optical platform, a correction die body, an ASIC integrated circuit and a central computer;
the X-ray machine is used for generating and emitting X-rays, the photon counting detector is used for receiving the X-ray photon energy after attenuation in the bone mineral density measuring system, the optical platform is used for placing the X-ray machine, the correcting die body and the photon counting detector, so that the X-ray machine, the correcting die body and the photon counting detector are positioned on the same horizontal line and have a collimation function, the correcting die body obtains correction parameters in the system, and the corresponding relation between the bone mineral density and the high-low energy attenuation value is established. The ASIC integrated circuit converts the photoelectric information collected by the detector into digital information, and the digital information is transmitted to the central computer through the serial port circuit, so that the central computer can display the bone density value corresponding to the ROI image on a screen;
opening the photon counting detector in an environment with room temperature of 22 ℃ and humidity of 50%, correcting the photon counting detector, debugging working parameters, wherein the photon counting detector must achieve thermal stability before testing, the photon counting detector should be placed under a +5VDC power supply for at least 60 minutes at room temperature of 22 ℃, checking the functional performance of each pixel of the photon counting detector by a correction method including passive scanning and background scanning, and recording the background noise of the photon counting detector;
and (3) opening an X-ray machine, preheating an X-ray bulb tube for two minutes to ensure that photon energy reaches stable output, and then measuring high-low energy incident energy under the X-ray energy of 70KeV and 40KeV respectively.
Furthermore, the photon counting detector is a photon counting detector adopting cadmium zinc telluride crystals.
Furthermore, in each grid, an influence domain is divided by taking the target node as a center, in the influence domain, the target node is only influenced by different weights of adjacent nodes, and other nodes outside the influence domain have no influence on the target node. Traversing each data point of the high-energy and low-energy attenuation values in a spatial domain, and setting a change step length in an influence domain; the change step length is related to the weight function type, a Gaussian weight function is selected, and the change step length is half of the radius of the influence domain. And determining a circular area by taking the target node as a center, wherein the radius of the influence domain is the distance from the target node to the circle, and the radius of the influence domain needs to be set under the condition that the matrix is reversible.
8. Further, a bone density ROI area needing to be measured is determined, the photon counting detector scans a part to be measured to obtain an X-ray attenuation value, corresponding die body thicknesses are obtained through calculation of formulas (12) and (13), linear transformation is performed again to obtain a formula (14), and the product of the die body thickness X and the die body density rho is the bone density (BMD) value of the corresponding ROI area.
ROIBMD=xρ (14)
The invention has the following advantages and beneficial effects:
in a traditional bone density detection system, a photon counting detector is commonly used for receiving photon energy after X-ray attenuation, and compared with a scintillator detector, the photon counting detector has the following advantages: the method is insensitive to the ambient temperature, free of deliquescence, stable in chemical substances, free of a photomultiplier tube and higher in charge collection characteristic and energy resolution, but due to the fact that the preparation process is immature and crystal growth defects exist, nonuniformity among pixels and bad points are prone to appearing on the edge of a detector, the influence of pixels on the edge of the detector by a conventional photon counting algorithm is large, the conventional bone density algorithm has large errors, and the diagnosis result of the bone density value is influenced. The invention adopts a dual-energy X-ray bone mineral density measuring system based on the mobile least square algorithm, the mobile least square algorithm has unique compactness, the distance between a fitting curved surface and a discrete point is smaller by enhancing the influence of adjacent nodes, and the fitting precision is effectively improved. The invention uses the moving least square algorithm to calculate the bone density value on the dual-energy X-ray bone density measuring system based on the photon counting detector, two independent variables are set in the algorithm formula and a weight function is determined, then the matrix in the numerical calculation process is a nonsingular matrix, namely the matrix can be successfully fitted in a three-dimensional space after being reversible, and the formulas (12) and (13) are obtained through improvement, so that the dual-energy X-ray bone density measuring system has the characteristics of self-adaptability and local fitting. In the process of using the code, when different weight functions are called to verify data, operation errors occur, and code debugging is needed. It should be noted that when acquiring high and low energy attenuation values, 10 parts of thickness data should be acquired in each group, and in order to obtain data of a detector in a steady state, an average value of 6 parts of data with the closest data value should be finally selected for final fitting. The algorithm selects a Gaussian weight function, and the radius of an influence domain is flexibly set, so that the function value of the point is only influenced by nodes in the nearby influence domain and not influenced by nodes outside the region, and the influences of nonuniformity among pixels of the detector and edge dead pixels are improved. The method has the advantages that the mobile least square algorithm fits the data relation between the high-energy and low-energy projection values of the X-ray and the thickness, the sample point influence weight and the sample point influence radius are controlled through the weight function, the problems of fitting smoothness and localization are well solved, the method has great advantages in the aspect of data fitting, and the method has the advantages of being strong in scattered point adaptability, having local fitting and optimizing the performance of the detector.
Drawings
FIG. 1 is a schematic diagram of a dual energy X-ray bone density diagnostic system according to a preferred embodiment of the present invention;
FIG. 2 is a flow chart of a bone density system;
FIG. 3 is a flow chart of a moving least squares algorithm;
FIG. 4 is a schematic diagram of a moving least squares spatial surface fit;
FIG. 5 is a comparison of a conventional polynomial with a moving least squares fit.
Detailed Description
The technical solutions in the embodiments of the present invention will be described in detail and clearly with reference to the accompanying drawings. The described embodiments are only some of the embodiments of the present invention.
The technical scheme for solving the technical problems is as follows:
the invention provides a dual-energy X-ray bone density detection system based on a moving least square algorithm, which aims to solve the problem that the nonuniformity of pixels of a photon counting detector has an error influence on the traditional bone density fitting algorithm.
In order to achieve the above purpose, the present invention provides the following technical solutions.
1. An optical platform as shown in figure (1) is set up for collecting the X-ray high and low energy attenuation values of the phantom. The method mainly comprises the following steps: the device comprises an X-ray machine, a photon counting detector, an optical platform, a correction die body, an ASIC integrated circuit, a central computer and the like.
2. The detector is turned on at room temperature (22 ℃) and humidity of 50%, equipment correction is carried out on the detector, working parameters are debugged, and the detector must achieve thermal stability before testing, so the detector should be placed under a +5VDC power supply for at least 60 minutes at room temperature (22 ℃), functional performance of each pixel of the detector is tested through correction methods such as passive scanning and background scanning, and background noise of the detector is recorded.
3. And (3) opening an X-ray machine, preheating an X-ray bulb tube for two minutes to ensure that photon energy reaches stable output, and then measuring high-low energy incident energy under the X-ray energy of 70KeV and 40KeV respectively.
4. In nature, the relative atomic mass of metal Al is similar to that of human skeleton, the relative atomic mass of polycarbonate high molecular compound is similar to that of human soft tissue, and the absorption rate of X-ray by the objects with similar relative atomic numbers can be considered as equivalent, so the Al block and the polycarbonate high molecular compound are commonly used in laboratories to replace the skeleton in human body, and the soft tissue is subjected to die body correction experiments. Al blocks with different thicknesses and polycarbonate high molecular compounds are stacked in a trapezoid shape as shown in a figure (1), and X-ray high-energy and low-energy attenuation values under different thickness combinations can be obtained by scanning once.
5. Because photon energy data directly acquired by the detector is large, data preprocessing is needed, and X-ray attenuation values I under high energy and low energy are respectively obtained through formulas (1) and (2)H、ILRespectively represent high-energy and low-energy XAnd the ray detector under the die body collects data. I isOH、IOLRepresenting the incident energy of high-and low-energy X-rays, yH、yLRespectively refers to the attenuation degree of the photon energy of the X-ray under the conditions of high energy and low energy.
6. The expression of the moving least squares algorithm is shown in (3), where α (x) is a coefficient vector matrix, p (x) is a basis function vector, i represents an actual data point, f (x) is a fitting function, a norm is solved for the formula (3) to obtain a formula (4), w (x) is a weight function, x is a target node, and x is a target nodeiFor other discrete data points, yIFor the actual data point, to solve for its minimum, equation (4) computes the partial derivative for α (x) to obtain (5).
7. Expanding the matrix A, B in equation (5), (6) and (7), it can be seen that the two matrices are coefficient matrices composed of basis functions and weight functions.
A(x)=PTW(x)P (6)
B(x)=PTW(x) (7)
yT=[y1,y2,...,yn] (8)
8. The basis function is a K-order polynomial, the K-order polynomial is expanded as shown in (9), the weight function is a diagonal matrix, the expression of the diagonal matrix is shown in (10), the formula (5) is deformed and is substituted into the formula (3), and finally the standard formula of the moving least square is obtained as (11).
f(x)=pTA-1(x)B(x)y (11)
9. Since the data designed by the system is three-dimensional space fitting, the method of moving least squares is adopted as shown in formulas (12), (13), wherein x and z represent the thicknesses of the module Al and the polycarbonate high molecular compound respectively, and fLRepresents the fitting value at low energy, fHFitting values at high energy. Carry in yL、yHThe fitting relation of the die body thickness under high energy and low energy can be respectively obtained
fL(x,z)=pT(x,z)A-1(x,z)B(x,z)yL (12)
fH(x,z)=pT(x,z)A-1(x,z)B(x,z)yH (13)
7. And finally, scanning the human body, selecting an ROI (region of interest) to obtain corresponding X-ray high-low energy attenuation data, and calculating the thickness of the corresponding bone and soft tissue, wherein the product of the thickness and the module density is the bone density of the corresponding part of the human body.
The step of fitting the phantom thickness-high and low energy attenuation values by using the moving least square algorithm is shown in figure (3), firstly, data collected by a detector are correspondingly preprocessed, background noise and electronic noise are removed, and the high and low energy attenuation values are calculated according to formulas (1) and (2). And (3) performing grid division on discrete data points according to the number of the die body thicknesses, wherein the types of the weight functions commonly used by the mobile least square method comprise a Gaussian weight function, a cubic spline weight function and a double-interpolation weight function. And after determining the weight function, dividing the radius of the influence domain. In the spatial domain, each data point is traversed. As shown in fig. 4, in each mesh, the radius of the influence domain is divided with the target node as the center, and the variation step is set in the influence domain, so that the influence weight of the data point closer to the target node is larger, and conversely, the influence on the target node is smaller as the distance is farther. And the action effect of the data nodes outside the influence domain on the target node is zero, and the final fitting relation is obtained according to the formulas (12) and (13). The radius of an influence domain can be flexibly set by the moving least square algorithm, the problems of fitting smoothness and localization are well solved, and the non-uniform characteristic of the pixels of the photon counting detector is effectively corrected. And after all the nodes are traversed, obtaining a die body thickness-high and low energy attenuation value fitting curved surface. And finally, determining a bone density ROI (region of interest) required to be measured according to personal requirements of a user, scanning the human body by using a detector to obtain an X-ray attenuation value, calculating to obtain the corresponding die body thickness, and performing linear transformation once again, wherein the product of the die body thickness and the die body density is the bone density value of the corresponding part.
The invention discloses a structure of a dual-energy X-ray bone mineral density diagnosis system based on a mobile least square algorithm and a specific flow of the mobile least square algorithm in fitting a die body thickness-high and low energy attenuation value, and the invention has the following beneficial effects:
from a dual-energy X-ray bone density system of a photon counting detector, the fitting of the bone density by the moving least square algorithm is more accurate. In clinical practice, doctors can diagnose osteoporosis diseases more conveniently, and patients can be helped to prevent and treat bone diseases in advance. Compared with the traditional lookup table method, the mobile least square algorithm has smaller requirement on the memory and shorter running time, and effectively improves the performance of the bone density system. Compared with a traditional polynomial fitting algorithm, the weight function and the influence domain are added to the mobile least square algorithm on the basis of a general fitting algorithm, so that the target node is only influenced by the adjacent nodes in the influence domain, the influence weight is smaller along with the distance, the influence domain data point has no weight influence on the target node, and the mobile least square algorithm is more flexible and solves the problem of localization in the fitting process.
When a 4 × 16 channel photon counting detector receives X-ray energy, as can be seen from fig. 5, the error influence of the nonuniformity of the edge pixels of the detector on the conventional algorithm is large, and therefore, the conventional algorithm cannot correct the error influence of the nonuniformity of the detector on the system when fitting the thickness-high and low energy attenuation values. The traditional polynomial fitting method cannot meet the requirement of local processing and causes the problems of difficult model setting and unstable calculation for a large amount of data. The dual-energy X-ray bone density system adopting the moving least square algorithm has obvious superiority in measuring the bone density. Compared with the traditional algorithm, the moving least square algorithm increases a weight function and an influence domain on the mathematical principle, effectively processes multidimensional data, well solves the problems of fitting smoothness and localization, has great advantages in the aspect of data fitting, has the advantages of strong adaptability to scattered points, local fitting or interpolation characteristics, high precision and the like, and is suitable for the detection data fitting of the photon counting detector with poor signal uniformity of a pixel unit and obvious edge dead points.
The invention combines a dual-energy X-ray bone mineral density system with a moving least square algorithm, obviously improves the precision of fitting bone mineral density, and obviously improves the pixel nonuniformity of a photon counting detector. Has practical feasibility and wide application value.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above examples are to be construed as merely illustrative and not limitative of the remainder of the disclosure. After reading the description of the invention, the skilled person can make various changes or modifications to the invention, and these equivalent changes and modifications also fall into the scope of the invention defined by the claims.
Claims (8)
1. A dual-energy X-ray bone mineral density detection method based on a mobile least square algorithm is characterized by comprising the following steps:
acquiring detection data of a die body under high-energy and low-energy X rays of a photon counting detector, and carrying out preprocessing including denoising; and acquiring incident energy of the high-energy and low-energy X-rays;
calculating to obtain a high-energy and low-energy attenuation value according to the preprocessed detection data and the incident energy;
performing grid division on discrete high and low energy attenuation value data points according to the thickness of the die body;
after a weight function of the moving least square algorithm is determined, in each grid, the radius of an influence domain is divided by taking a target node as a center, in a spatial domain, traversal is carried out on data points of each high-low energy attenuation value, and a change step length is set in the influence domain;
obtaining a final fitting relation by adopting a moving least square algorithm, and obtaining a die body thickness-high and low energy attenuation value fitting curved surface after all nodes are traversed;
determining a bone density ROI area to be measured, acquiring an X-ray attenuation value of a part to be measured by a photon counting detector, calculating to obtain the corresponding die body thickness, and performing linear transformation once again, wherein the product of the die body thickness and the die body density is the bone density value of the corresponding part.
2. The method for detecting bone density of dual-energy X-ray based on mobile least square algorithm according to claim 1, wherein the calculating of the attenuation values of high and low energy according to the preprocessed detection data and the incident energy specifically comprises:
obtaining X-ray attenuation values I at high and low energy levels by formulas (1) and (2) respectivelyH、ILData collected by detectors respectively representing high-energy and low-energy X-rays under a placed die body, IOH、IOLRespectively representing high-energy and low-energy X-raysIncident energy of the line, yH、yLRespectively refers to the attenuation degree of the photon energy of the X-ray under the conditions of high energy and low energy;
3. the method for detecting bone mineral density by dual-energy X-ray based on mobile least square algorithm of claim 1, wherein the weight function of the mobile least square method is a Gaussian weight function, a cubic spline weight function and a dual interpolation weight function, and simulation fitting and actual test experiments prove that the fitting effect of the Gaussian weight function is optimal.
4. The method for detecting bone density of dual energy X-ray based on mobile least square algorithm according to claim 1, 2 or 3, wherein the expression of the mobile least square algorithm is shown in (3), where α (X) is coefficient vector matrix, P (X) is basis function vector, i represents actual data point, f (X) is fitting function, solving norm of (3) to obtain formula (4), w (X) is weight function, X is target node, X is weight functioniFor other discrete data points, J represents the norm of equation (3), yIFor the actual data point, in order to obtain the minimum value of the actual data point, the formula (4) calculates the partial derivative of alpha (x) to obtain (5);
expanding the matrix A, B in the formula (5), (6) and (7), it can be seen that the two matrices are coefficient matrices formed by basis functions and weight functions;
A(x)=PTW(x)P (6)
B(x)=PTW(x) (7)
yT=[y1,y2,...,yn] (8)
w (x) is a weight function, P represents a basis function, ynRepresenting the actual X-ray attenuation values.
The basis function is a K-order polynomial, the K-order polynomial is expanded as shown in (9), the weight function is a diagonal matrix, the expression of the diagonal matrix is shown in (10), the formula (5) is deformed and is substituted into the formula (3), and finally the standard formula of the moving least square is obtained as (11);
f(x)=pTA-1(x)B(x)y (11)
p is a basis function, pm(xn) The nth value of x under the m-order complete polynomial, m represents the highest order of the basis function P, and n represents the number of x;
since the data are three-dimensional space fitting, the moving least square method is adopted as shown in formulas (12), (13), wherein x and z represent the thicknesses of the module Al and the polycarbonate high molecular compound respectively, and fLRepresents the fitting value at low energy, fHFor fitting values at high energy, carry in yL、yHThe fitting relation of the die body thickness under high energy and low energy can be respectively obtained
fL(x,z)=pT(x,z)A-1(x,z)B(x,z)yL (12)
fH(x,z)=pT(x,z)A-1(x,z)B(x,z)yH (13)。
5. The method for detecting bone density of dual-energy X-ray based on mobile least square algorithm as claimed in claim 4, wherein the detection data of the phantom under the high-energy and low-energy X-ray of the photon counting detector is obtained and preprocessed to remove noise; and acquiring incident energy of the high-energy and low-energy X-rays, specifically comprising:
an optical platform is built for collecting X-ray high and low energy attenuation values of a die body, and the optical platform mainly comprises the following parts: the system comprises an X-ray machine, a photon counting detector, an optical platform, a correction die body, an ASIC integrated circuit and a central computer; the X-ray machine is used for generating and emitting X-rays, the photon counting detector is used for receiving the attenuated X-ray photon energy in the bone density measuring system, and the optical platform is used for placing the X-ray machine, the correcting die body and the photon counting detector, so that the X-ray machine, the correcting die body and the photon counting detector are positioned on the same horizontal line and have a collimation function. The correction die body obtains correction parameters in the system, the corresponding relation between the bone density and the high and low energy attenuation values is established, the ASIC integrated circuit converts photoelectric information collected by the detector into digital information, and the digital information is transmitted to the central computer through the serial port circuit, so that the screen display of the bone density value corresponding to the ROI image is realized;
opening the photon counting detector in an environment with room temperature of 22 ℃ and humidity of 50%, correcting the photon counting detector, debugging working parameters, wherein the photon counting detector must achieve thermal stability before testing, the photon counting detector should be placed under a +5VDC power supply for at least 60 minutes at room temperature of 22 ℃, checking the functional performance of each pixel of the photon counting detector by a correction method including passive scanning and background scanning, and recording the background noise of the photon counting detector;
and (3) opening an X-ray machine, preheating an X-ray bulb tube for two minutes to ensure that photon energy reaches stable output, and then measuring high-low energy incident energy under the X-ray energy of 70KeV and 40KeV respectively.
6. The method for detecting bone density of dual-energy X-ray based on moving least square algorithm of claim 5, wherein the photon counting detector is a photon counting detector using cadmium zinc telluride crystal.
7. The method for detecting bone density of dual energy X-ray based on moving least square algorithm as claimed in claim 1, wherein in each grid, an influence domain is divided by taking a target node as a center, in the influence domain, the target node is only influenced by different weights of adjacent nodes, and other nodes outside the influence domain have no influence on the target node; traversing each data point of the high-energy and low-energy attenuation values in a spatial domain, and setting a change step length in an influence domain; the change step length is related to the weight function type, a Gaussian weight function is selected, and the change step length is half of the radius of the influence domain. And determining a circular area by taking the target node as a center, wherein the radius of the influence domain is the distance from the target node to the circle, and the radius of the influence domain needs to be set under the condition that the matrix is reversible.
8. The method for detecting the bone density of the dual-energy X-ray based on the moving least square algorithm as claimed in claim 1, wherein the ROI area of the bone density to be measured is determined, the photon counting detector scans the part to be measured to obtain the X-ray attenuation value, the corresponding die body thickness is obtained by the calculation of the formulas (12) and (13), the linear transformation is performed again to obtain the formula (14), and the product of the die body thickness X and the die body density p is the bone density (BMD) value of the part corresponding to the ROI.
ROIBMD=xρ (14)。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210040299.8A CN114391857A (en) | 2022-01-14 | 2022-01-14 | Dual-energy X-ray bone mineral density detection method based on mobile least square algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210040299.8A CN114391857A (en) | 2022-01-14 | 2022-01-14 | Dual-energy X-ray bone mineral density detection method based on mobile least square algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114391857A true CN114391857A (en) | 2022-04-26 |
Family
ID=81230203
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210040299.8A Pending CN114391857A (en) | 2022-01-14 | 2022-01-14 | Dual-energy X-ray bone mineral density detection method based on mobile least square algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114391857A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116595291A (en) * | 2023-07-18 | 2023-08-15 | 武汉怡特环保科技有限公司 | Multi-exponential fitting method and device for fluorescence attenuation curve and computer equipment |
CN117257333A (en) * | 2023-11-17 | 2023-12-22 | 深圳翱翔锐影科技有限公司 | True dual-energy X-ray bone densitometer based on semiconductor detector |
-
2022
- 2022-01-14 CN CN202210040299.8A patent/CN114391857A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116595291A (en) * | 2023-07-18 | 2023-08-15 | 武汉怡特环保科技有限公司 | Multi-exponential fitting method and device for fluorescence attenuation curve and computer equipment |
CN116595291B (en) * | 2023-07-18 | 2023-09-26 | 武汉怡特环保科技有限公司 | Multi-exponential fitting method and device for fluorescence attenuation curve and computer equipment |
CN117257333A (en) * | 2023-11-17 | 2023-12-22 | 深圳翱翔锐影科技有限公司 | True dual-energy X-ray bone densitometer based on semiconductor detector |
CN117257333B (en) * | 2023-11-17 | 2024-02-20 | 深圳翱翔锐影科技有限公司 | True dual-energy X-ray bone densitometer based on semiconductor detector |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Feldkamp et al. | The direct examination of three‐dimensional bone architecture in vitro by computed tomography | |
JP5834129B2 (en) | Enhanced low-contrast detection capability for radiographic imaging systems | |
JP6590381B2 (en) | X-ray apparatus, data processing apparatus and data processing method | |
US6904118B2 (en) | Method and apparatus for generating a density map using dual-energy CT | |
US4149081A (en) | Removal of spectral artifacts and utilization of spectral effects in computerized tomography | |
CN1880949B (en) | Method for calculating absorber-specific weighting coefficients and method for improving a contrast-to-noise ratio | |
EP2243020B1 (en) | System and method for quantitative imaging of chemical composition to decompose more than two materials | |
CN114391857A (en) | Dual-energy X-ray bone mineral density detection method based on mobile least square algorithm | |
EP2243021B1 (en) | System and method for quantitative imaging of chemical composition to decompose multiple materials | |
CN106618620A (en) | Osteosclerosis artifact correction coefficient calculation method and device | |
CN104411245B (en) | To the dynamic modeling of the defect of photon counting detector | |
CN104939848B (en) | The generation of monochrome image | |
JPH05502610A (en) | Methods to improve the dynamic range of imaging devices | |
US7086780B2 (en) | Methods for spectrally calibrating CT imaging apparatus detectors | |
CN110175957A (en) | Multipotency amount CT sill substance decomposition method | |
HUE035719T2 (en) | New imaging modality using penetrating radiations | |
CN101384218A (en) | X-ray beam calibration for bone mineral density assessment using mammography system | |
US7050530B2 (en) | Method and apparatus for the spatially-resolved determination of the element concentrations in objects to be examined | |
CN108720858A (en) | A kind of novel bone density computational methods and device using high-order curved surface fitting | |
CN110123349B (en) | Bone mineral density measuring method and device | |
US20050002484A1 (en) | Method and apparatus for correcting bone induced spectral artifacts | |
Seeram et al. | Computed tomography: Physical principles, instrumentation, and quality control | |
CN111089869B (en) | Multi-energy detector X-ray phase contrast imaging method and system, storage medium and equipment | |
US6943549B2 (en) | Method for processing magnetic resonance imaging image information and magnetic resonance imaging system | |
CN110731788A (en) | method for rapidly decomposing base substance based on dual-energy CT scanner |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |