Clinical image-based cancellous bone mechanical property prediction method and system

Abstract

The invention provides a clinical image-based cancellous bone mechanical property prediction method and system, and the method comprises the steps: carrying out the medical CT scanning of a plurality of prepared cancellous bone cubic samples, and obtaining the medical CT image data; carrying out segmentation processing on the medical CT image data, extracting a sample area and calculating an average HU value of a sample CT image; performing a uniaxial compression experiment on the sample, accurately measuring strain by using a digital image correlation method, and further obtaining a stress-strain curve to calculate an elastic modulus and compression strength; according to the average HU value and the corresponding mechanical index data, a linear relation model of the average HU value of the cancellous bone CT image and the elastic modulus and the compression strength of the cancellous bone is established, and then a personalized treatment scheme is formulated according to the characteristics of the cancellous bone of the patient. The method has high universality and accuracy, and the elastic modulus and the compression strength of the cancellous bone sample can be accurately measured.

Description

Technical Field

[0001] The present invention relates to the technical field of cancellous bone mechanical property prediction, and specifically to a cancellous bone mechanical property prediction method and system based on clinical images, and more particularly to a clinical image processing method, a mechanical property testing method, a prediction model modeling method, and a model application method. Background Art

[0002] Osteoporosis is one of the main problems affecting the health of the middle-aged and elderly. Its most serious complication, osteoporotic fractures, also has a high disability and mortality rate, which brings a serious burden to society and families. In the clinical treatment of osteoporotic vertebral compression fractures, a unified treatment plan is adopted clinically, which does not take into account the significant differences in bone mechanical properties among patients of different ages, genders, and osteoporosis degrees. Problems such as no or poor pain relief and recurrent fractures of adjacent vertebrae often occur. Therefore, it is necessary to develop personalized treatment plans for different patients, and to establish the relationship between medical imaging and bone mechanical properties to assist in treatment.

[0003] Medical CT is accurate, safe and efficient, and clinical CT examination plays an indispensable role in modern medicine. In order to solve the problems of clinical diagnosis and personalized treatment, it is urgent to establish a consistent clinical image-mechanical testing process and create a mapping relationship model that can accurately predict the mechanical properties of cancellous bone based on clinical images. Therefore, the relationship between the average HU value measured by medical CT and the mechanical properties of cancellous bone is established, and a method for predicting the mechanical properties of cancellous bone based on the average HU value of clinical CT images is created. Cancellous bone is an anisotropic porous material. It is difficult to obtain accurate results by measuring its elastic modulus, compression strength and other mechanical properties using traditional methods. In order to establish an accurate prediction model that can be applied clinically, it is necessary to develop a method that can more accurately measure the strain during the compression test of cancellous bone.

[0004] Patent document CN106447787A discloses a method for determining the relationship between bone CT value and elastic modulus and a test loading device thereof. The method determines the accurate relationship between the CT value and the bone elastic modulus without destroying the bone tissue, and establishes a loading device (caliber 10 cm) with a large test range that can be used in micro-CT to ensure that it can be inserted into large joints such as the femoral head to ensure the integrity of the sample; the elastic modulus of the bone tissue corresponding to the CT scan voxel is obtained through the finite element inversion calculation method.

[0005] However, patent document CN106447787A mainly obtains high-resolution imaging information through micro-CT and performs finite element simulation, which is not suitable for clinical prediction of the mechanical properties of cancellous bone of patients. Summary of the invention

[0006] In view of the defects in the prior art, the object of the present invention is to provide a method and system for predicting the mechanical properties of cancellous bone based on clinical images.

[0007] A method for predicting mechanical properties of cancellous bone based on clinical images provided by the present invention comprises:

[0008] Data acquisition steps: performing medical CT scanning on a number of prepared cancellous bone cube samples to obtain medical CT image data;

[0009] Image processing step: segmenting the medical CT image data, extracting the sample area, and storing the HU of each voxel of the three-dimensional medical CT image in a three-dimensional digital matrix, and then calculating the average HU value of the sample CT image;

[0010] Mechanical testing steps: a uniaxial compression test is performed on the sample, and the strain is accurately measured using a digital image correlation method to obtain a stress-strain curve, and the elastic modulus and compression strength are calculated based on the stress-strain curve;

[0011] Mapping relationship model step: According to the average HU value and the corresponding mechanical index data, a linear relationship model between the average HU value of the cancellous bone CT image and the elastic modulus and compressive strength of the cancellous bone is established, and multiple HU value-mechanical property models established for different age groups and different genders are stored. The linear relationship model is used to predict the elastic modulus and compressive strength of the patient's cancellous bone, and then a personalized treatment plan is formulated according to the characteristics of the patient's cancellous bone.

[0012] Preferably, the plurality of cancellous bone cube samples are taken from the central area of ​​the vertebral cancellous bone, with a size of 15 mm×15 mm×15 mm, and the axial direction of the sample is parallel to the axial direction of the vertebral segment;

[0013] The extraction of the cancellous bone cube sample adopts a precision machining method;

[0014] The medical CT scanning scans the cancellous bone cube sample by using a medical CT device.

[0015] Preferably, the uniaxial compression test is carried out on a material testing machine, and a camera is used to record the speckle changes on the sample surface in real time and convert them into strain changes over time. The compressor directly outputs a force-time curve, and the stress changes over time are calculated according to the cross-sectional area of ​​the tested sample, and a stress-strain curve of the sample compression process is drawn;

[0016] The calculation of the elastic modulus includes determining the elastic range, that is, the portion of the stress-strain curve that presents a linear relationship. In the elastic range, stress and strain satisfy Hooke's law σ=Eε. The slope of the linear segment calculated by linear fitting is the elastic modulus E.

[0017] The calculation of the compressive strength includes finding the maximum stress point in the stress-strain curve. The maximum stress point corresponds to the maximum stress that the material can withstand before failure, which is the compressive strength σ c =max(σ).

[0018] Preferably, the mapping relationship model step includes: importing the average HU value of different samples and the elastic modulus and compression strength corresponding to each sample into data statistical software, performing linear regression analysis respectively, establishing a linear relationship model and calculating the determination coefficient to verify the accuracy of the model. The linear relationship model is as follows:

[0019] E=a·HU+b

[0020] S=c·HU+d

[0021] Where, E is the elastic modulus, S is the compressive strength, HU is the average HU value of the sample, and a, b, c, and d are fitting parameters.

[0022] Preferably, the calculation formula of the determination coefficient is:

[0023]

[0024] Where y represents the elastic modulus E or the compressive strength S, y i is the observed value, The predicted values ​​of the fitted model, is the mean of the observations.

[0025] A cancellous bone mechanical property prediction system based on clinical images provided by the present invention comprises:

[0026] Data acquisition module: performing medical CT scanning on a number of prepared cancellous bone cube samples to obtain medical CT image data;

[0027] Image processing module: segmenting the medical CT image data, extracting the sample area, and storing the HU of each voxel of the three-dimensional medical CT image in a three-dimensional digital matrix, and then calculating the average HU value of the sample CT image;

[0028] Mechanical testing module: Perform uniaxial compression test on the sample, use digital image correlation method to accurately measure strain, and then obtain stress-strain curve, and calculate elastic modulus and compression strength according to the stress-strain curve;

[0029] Mapping relationship model module: Based on the average HU value and the corresponding mechanical index data, a linear relationship model between the average HU value of the cancellous bone CT image and the elastic modulus and compressive strength of the cancellous bone is established, and multiple HU value-mechanical property models established for different age groups and genders are stored. The linear relationship model is used to predict the elastic modulus and compressive strength of the patient's cancellous bone, and then a personalized treatment plan is formulated based on the patient's cancellous bone characteristics.

[0030] Preferably, the plurality of cancellous bone cube samples are taken from the central area of ​​the vertebral cancellous bone, with a size of 15 mm×15 mm×15 mm, and the axial direction of the sample is parallel to the axial direction of the vertebral segment;

[0031] The extraction of the cancellous bone cube specimen adopts a precision machining system;

[0032] The medical CT scanning scans the cancellous bone cube sample by using a medical CT device.

[0033] Preferably, the uniaxial compression test is carried out on a material testing machine, and a camera is used to record the speckle changes on the sample surface in real time and convert them into strain changes over time. The compressor directly outputs a force-time curve, and the stress changes over time are calculated according to the cross-sectional area of ​​the tested sample, and a stress-strain curve of the sample compression process is drawn;

[0034] The calculation of the elastic modulus includes determining the elastic range, that is, the portion of the stress-strain curve that presents a linear relationship. In the elastic range, stress and strain satisfy Hooke's law σ=Eε. The slope of the linear segment calculated by linear fitting is the elastic modulus E.

[0035] The calculation of the compressive strength includes finding the maximum stress point in the stress-strain curve. The maximum stress point corresponds to the maximum stress that the material can withstand before failure, which is the compressive strength σ c =max(σ).

[0036] Preferably, the mapping relationship model module includes: importing the average HU value of different samples and the elastic modulus and compression strength corresponding to each sample into data statistical software, performing linear regression analysis respectively, establishing a linear relationship model and calculating the determination coefficient to verify the accuracy of the model. The linear relationship model is as follows:

[0037] E=a·HU+b

[0038] S=c·HU+d

[0039] Where, E is the elastic modulus, S is the compressive strength, HU is the average HU value of the sample, and a, b, c, and d are fitting parameters.

[0040] Preferably, the calculation formula of the determination coefficient is:

[0041]

[0042] Where y represents the elastic modulus E or the compressive strength S, y i is the observed value, The predicted values ​​of the fitted model, is the mean of the observations.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The prediction model of the present invention can directly output the predicted mechanical properties based on the three-dimensional measurement results of medical CT images, which has high universality and accuracy, and provides an important reference for the diagnosis and personalized treatment of osteoporosis. Combined with the digital image correlation method, the uniaxial compression test of the sample can accurately measure the elastic modulus and compression strength of the cancellous bone sample.

[0045] 2. Based on the differences between individuals of different ages and genders, the present invention matches the corresponding HU values ​​with the mechanical property models, and can establish multiple HU-mechanical property models according to cancellous bone samples of different ages and genders, making the clinical prediction results more accurate and individualized.

[0046] 3. The imaging CT images of the present invention are from medical CT equipment, which has the characteristics of low resolution, short scanning time, and small radiation dose. It can be used for clinical diagnosis and treatment of CT equipment. Therefore, the established prediction model is suitable for clinical prediction of the mechanical properties of cancellous bone in patients.

[0047] 4. The present invention uses a large sample size of data for verification, so that the established model has robustness and wide applicability. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0049] Figure 1 It is a schematic diagram of the working method of the present invention;

[0050] Figure 2 It is a schematic diagram of the sample surface in the present invention;

[0051] Figure 3 is the segmented medical CT image in the present invention;

[0052] Figure 4 is the voxel number-HU value distribution diagram in the present invention;

[0053] Figure 5 In the present invention, a uniaxial compression test is performed on the sample;

[0054] Figure 6 This is the linear stage fitting result of stress-strain curve of sample No. 02;

[0055] Figure 7 It is the elastic modulus-average HU value fitting diagram;

[0056] Figure 8 It is the compressive strength-average HU value fitting diagram. DETAILED DESCRIPTION

[0057] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0058] According to a method for predicting mechanical properties of cancellous bone based on clinical images provided by the present invention, Figure 1 As shown, including:

[0059] Data collection step: Perform medical CT scanning on the prepared cancellous bone cube samples to obtain medical CT image data. The cancellous bone cube samples are taken from the central area of ​​the vertebral cancellous bone, with a size of 15mm×15mm×15mm, and the axial direction of the sample is parallel to the axial direction of the vertebral segment. The extraction of the cancellous bone cube samples adopts a precision machining method to ensure the roughness of the sample surface and the accuracy of the sample size. The sample surface is as follows: Figure 2 As shown, the sample is always in a frozen state of -40°C during non-test period. The medical CT scanning is performed by scanning the cancellous bone cube sample using a medical CT device.

[0060] Image processing step: segment the medical CT image data, extract the sample area, and store the HU of each voxel of the three-dimensional medical CT image in a three-dimensional digital matrix, and then calculate the average HU value of the sample CT image. Specifically, the CT image size is reduced by three-dimensional image processing software for manual segmentation, the sample area is accurately extracted, and the non-sample area of ​​the image is deleted to improve the calculation efficiency. Then, the three-dimensional CT image of the segmented sample is converted into a three-dimensional digital matrix, the HU value of each voxel is stored in the corresponding digital matrix, and the average HU value is calculated using the voxel weighted average method.

[0061] Mechanical testing steps: Perform a uniaxial compression test on the sample, use the digital image correlation method to accurately measure the strain, and then obtain the stress-strain curve, and calculate the elastic modulus and compressive strength based on the stress-strain curve. The uniaxial compression test is carried out on a material testing machine, and a camera is used to record the speckle changes on the sample surface in real time and convert them into strain changes over time. The compressor directly outputs the force-time curve, and the stress changes over time are calculated based on the cross-sectional area of ​​the tested sample. The stress-strain curve of the sample compression process is drawn to ensure the accuracy of the elastic modulus and compressive strength calculations. To calculate the elastic modulus, first determine the elastic range, that is, the part of the stress-strain curve that shows a linear relationship. Within the elastic range, stress and strain satisfy Hooke's law σ=Eε. The slope of the linear segment calculated by linear fitting is the elastic modulus E. To calculate the compressive strength, find the maximum stress point in the stress-strain curve. This point corresponds to the maximum stress that the material can withstand before failure, which is the compressive strength σ c =nax(σ).

[0062] Mapping relationship model step: According to the average HU value and the corresponding mechanical index data, a linear relationship model between the average HU value of the cancellous bone CT image and the elastic modulus and compressive strength of the cancellous bone is established, and multiple HU value-mechanical performance models established for different age groups and different genders are stored. The linear relationship model is used to predict the elastic modulus and compressive strength of the cancellous bone of the patient, and then a personalized treatment plan is formulated according to the characteristics of the cancellous bone of the patient. The mapping relationship model step includes: importing the average HU value of different samples and their elastic modulus and compressive strength into the data statistics software, performing linear regression analysis respectively, establishing a linear relationship model and calculating the determination coefficient to verify the accuracy of the model. The linear relationship model is as follows:

[0063] E=a·HU+b

[0064] S=c·HU+d

[0065] Where, E is the elastic modulus, S is the compressive strength, HU is the average HU value of the sample, and a, b, c, and d are fitting parameters.

[0066] The linear regression analysis process imports the average HU value of the CT image of the sample and the measured elastic modulus and compression strength into the data statistics software, applies the least squares method to perform data fitting analysis, obtains the linear fitting equation, and calculates the determination coefficient to evaluate the accuracy and reliability of the fitting model. The calculation formula of the determination coefficient is:

[0067]

[0068] Where y represents the elastic modulus E or the compressive strength S, y i is the observed value, The predicted values ​​of the fitted model, is the average value of the observed values. In the process of establishing the linear relationship model, the elastic modulus and compressive strength of a large number of cancellous bone specimens are predicted by using the established linear fitting model through data verification of a large sample size, and experimental measurement verification is performed to ensure the robustness and wide applicability of the model.

[0069] Based on clinical 3D CT detection, the 3D image obtained by medical CT scanning is first segmented, the cortical bone area is deleted, the cancellous bone area is extracted and divided into different areas, and the image is converted into a 3D digital matrix using numerical software. The voxel weighted average method is used to calculate the average HU value of different areas of the cancellous bone of the patient, and the linear relationship model is used to predict the elastic modulus and compression strength of the cancellous bone of the patient. Multiple HU value-mechanical property models established for different age groups, different genders and different health conditions are stored to meet the prediction needs of cancellous bone mechanical properties of different populations and realize personalized medical treatment.

[0070] The present invention is not only applicable to the evaluation of the elastic modulus and compressive strength of cancellous bone, but can also be further used for the construction of finite element numerical models, the diagnosis of osteoporosis, the assessment of fracture risk and the formulation of personalized treatment plans. The prediction method described in the control module can be adjusted according to individual data of different ages, genders and health conditions to meet the cancellous bone prediction needs of different populations and realize personalized medicine.

[0071] The present invention aims to provide a standard and accurate prediction method of the mechanical properties of cancellous bone based on medical images in response to the current lack of personalized treatment plans for patients with different degrees of osteoporosis in the clinic, and to provide a practical tool for personalized treatment of osteoporosis.

[0072] Furthermore, four bovine lumbar vertebrae were selected as samples and combined with the attached Figure 2 To Attachment Figure 8 The method for predicting the mechanical properties of cancellous bone based on clinical images of the present invention is specifically described as follows:

[0073] First, four bovine lumbar vertebrae were selected. The average age of the cattle was 2.5 years old and they had no bone diseases. Regularly shaped 15mm*15mm*15mm cubic specimens were extracted from the central part of the vertebral segment by machining and numbered. When cutting the specimens, the axial direction of the specimens was parallel to the distribution direction of the trabeculae and this direction was marked, such as Figure 2 As shown, the prepared sample was placed in a -40°C environment for cryopreservation. The cancellous bone cube sample was scanned using a Philips IQon spectral CT instrument to obtain a medical CT image of the sample.

[0074] Then, the three-dimensional CT image was imported into Avizo software, and the three-dimensional CT image was manually segmented to accurately extract the sample area, delete the non-sample area of ​​the image, and reduce the image size, such as Figure 3 shown.

[0075] Next, the processed three-dimensional CT image is imported into the MATLAB numerical software, the three-dimensional image is converted into a digital matrix, and the HU value of each voxel is stored in the digital matrix. According to the three-dimensional digital matrix, the average HU value of the sample CT image is calculated by the voxel weighted average method, as shown in FIG. Figure 4 As shown, the abscissa represents the HU value of the voxel, and the ordinate represents the voxel number corresponding to the HU value.

[0076] Next, the samples were subjected to compression tests using a compression testing machine, and a camera was used to record the changes in the speckle displacement on the sample surface over time at a speed of 4 frames / s. Figure 5 As shown in the figure, the change of line length is measured in the image correlation processing software, and then combined with the pressure-displacement curve recorded by the compression tester to obtain the stress-strain curve, as shown in Figure 6 The linear stage is fitted linearly to obtain the corresponding elastic modulus of the sample, and the compressive strength of the sample can be obtained from the load peak.

[0077] Finally, the average HU values ​​of different samples and their elastic modulus and compressive strength data were imported into the function plotting software to draw the elastic modulus-HU average value scatter plot and strength-HU average value scatter plot, respectively. Figure 7 and Figure 8 As shown, linear regression analysis was performed to establish a linear relationship model and calculate the determination coefficient to verify the accuracy of the model:

[0078] E=1.6069·HU+187.1,R 2 =0.602

[0079] S=0.04·HU-4.8464,R 2 =0.630

[0080] Where E is the elastic modulus, S is the compressive strength, HU is the average HU value of the sample, R 2 The vertebral specimens of different ages and genders were distinguished, and multiple HU value-mechanical property models were established. The linear relationship model was used to predict the elastic modulus and compressive strength of the cancellous bone of the patient, and then a personalized treatment plan was formulated according to the characteristics of the cancellous bone of the patient.

[0081] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0082] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for predicting the mechanical properties of cancellous bone based on clinical images, characterized in that: include: Data acquisition steps: performing medical CT scanning on a number of prepared cancellous bone cube samples to obtain medical CT image data; Image processing step: segmenting the medical CT image data, extracting the sample area, and storing the HU of each voxel of the three-dimensional medical CT image in a three-dimensional digital matrix, and then calculating the average HU value of the sample CT image; Mechanical testing steps: a uniaxial compression test is performed on the sample, and the strain is accurately measured using a digital image correlation method to obtain a stress-strain curve, and the elastic modulus and compression strength are calculated based on the stress-strain curve; Mapping relationship model step: According to the average HU value and the corresponding mechanical index data, a linear relationship model between the average HU value of the cancellous bone CT image and the elastic modulus and compressive strength of the cancellous bone is established, and multiple HU value-mechanical property models established for different age groups and different genders are stored. The linear relationship model is used to predict the elastic modulus and compressive strength of the patient's cancellous bone, and then a personalized treatment plan is formulated according to the characteristics of the patient's cancellous bone.

2. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 1, characterized in that: The plurality of cancellous bone cube samples are taken from the central area of ​​the vertebral cancellous bone; The extraction of the cancellous bone cube sample adopts a precision machining method; The medical CT scanning scans the cancellous bone cube sample by using a medical CT device.

3. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 2, characterized in that: The dimensions of the plurality of cancellous bone cubic samples are 15 mm×15 mm×15 mm, and the axial direction of the samples is parallel to the axial direction of the vertebral segment.

4. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 2, characterized in that: The samples were kept in a frozen state at -40°C during non-test periods.

5. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 1, characterized in that: The uniaxial compression test is carried out on a material testing machine, and a camera is used to record the speckle changes on the sample surface in real time and convert them into changes in strain over time. The compressor directly outputs the force-time curve, and the stress change over time is calculated based on the cross-sectional area of ​​the tested sample, and the stress-strain curve of the sample compression process is plotted.

6. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 1, characterized in that: The calculation of the elastic modulus includes determining the elastic range, that is, the portion of the stress-strain curve that presents a linear relationship. In the elastic range, stress and strain satisfy Hooke's law σ=Eε. The slope of the linear segment calculated by linear fitting is the elastic modulus E.

7. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 1, characterized in that: The calculation of the compressive strength includes finding the maximum stress point in the stress-strain curve. The maximum stress point corresponds to the maximum stress that the material can withstand before failure, which is the compressive strength σ c =max(σ).

8. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 1, characterized in that: The mapping relationship model step includes: importing the average HU value of different samples and the elastic modulus and compression strength corresponding to each sample into data statistical software, performing linear regression analysis respectively, establishing a linear relationship model and calculating the determination coefficient to verify the accuracy of the model. The linear relationship model is as follows: E=a·HU+b S=c·HU+d Where, E is the elastic modulus, S is the compressive strength, HU is the average HU value of the sample, and a, b, c, and d are fitting parameters.

9. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 8, characterized in that: The calculation formula of the determination coefficient is: Where y represents the elastic modulus E or the compressive strength S, y i is the observed value, The predicted values ​​of the fitted model, is the mean of the observations.

10. The method for predicting mechanical properties of cancellous bone based on clinical images according to claim 1, characterized in that: In the image processing step, a voxel weighted average method is used to calculate the average HU value.

Images