Human body lower limb bone surface reconstruction method and system based on A-type ultrasonic equipment

Abstract

The invention discloses a human body lower limb bone surface reconstruction method and system based on A-type ultrasonic equipment, and the method comprises the following steps: carrying out the signal receiving and transmitting of the A-type ultrasonic equipment, and obtaining ultrasonic data; processing the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object; calibrating the A-type ultrasonic probe, converting one-dimensional depth information into spatial three-dimensional coordinate points, and collecting three-dimensional coordinates of bone surface points of the lower limb of a human body; establishing a lower limb bone model database, and calculating an SSM deformable general model; reconstructing the bone surface of the lower limb of the human body; the system comprises an A-type ultrasonic equipment sub-module, an A-type ultrasonic equipment calibration sub-module, a statistical shape model construction sub-module, a deformable lower limb bone surface reconstruction sub-module and a visual interaction sub-module. The bone surface echo signal can be automatically detected and recognized, the bone surface echo signal is converted into the three-dimensional space point cloud through calibration, operation is easy and convenient, cost is low, extra radiation is avoided, and precision is high.

Description

Technical Field

[0001] The invention belongs to the technical field of surface reconstruction, and in particular relates to a method and system for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment. Background Art

[0002] Total knee arthroplasty (TKA) is the main means of treating serious knee diseases, relieving knee pain, and reconstructing knee function. With the rapid development of computer image processing, tracking and positioning technologies, computer-assisted orthopedic surgery systems have been proven to assist doctors in completing precise surgical treatments and have become an advantageous solution for knee surgery. The results depend on accurate preoperative planning and intraoperative techniques. The main purpose of preoperative planning is to diagnose patients, design joint implants, and plan surgical plans, which is the premise and basis for successful surgery.

[0003] Lower limb bone surface reconstruction technology can not only help doctors achieve accurate preoperative diagnosis and provide important reference for joint implant design, but also serve as an input source for computer-assisted orthopedic surgery. Through registration technology, it can be integrated with intraoperative bone information to guide the surgical process. A large number of studies have shown that for complex cases, simulation operations on preoperative three-dimensional models are necessary, and virtual models can be operated multiple times to verify the feasibility of surgical plans. Therefore, it is necessary to obtain the patient's specific anatomical structure, perform preoperative diagnosis, design personalized implants, and ultimately plan a reasonable and efficient surgical plan.

[0004] X-ray-based diagnosis is the mainstream today, but this method exposes patients and medical staff to radiation, increases inherent risks, and X-rays can only display the joint space and bone anatomy in a single plane. All visual assessments are performed in two-dimensional space, which cannot provide intuitive guidance to doctors. For this reason, CT and MRI technologies have been introduced into computer-assisted knee surgery systems and have become a key step in preoperative planning. Although these methods can generate highly accurate models, CT and MRI have inherent disadvantages, namely, both are expensive (especially MRI), and CT exposes patients to ionizing radiation. These limitations have prompted people to study new methods for lower limb bone surface reconstruction.

[0005] Compared with the above methods, B-mode ultrasound provides a green and non-invasive method to digitize the patient's bone surface, and uses its advantages of non-invasiveness and real-time measurement to improve diagnostic efficiency. In order to convert two-dimensional B-mode ultrasound images into three-dimensional intuitive display, researchers introduced the statistical shape model (SSM) into the human lower limb bone surface reconstruction system. SSM is a parameter-driven deformable general model. It is registered with the two-dimensional B-mode ultrasound image collected before surgery. By changing the model parameters to complete the deformation, the patient's lower limb bone surface reconstruction can be completed, thereby replacing the image data reconstructed by the computer-assisted navigation system before surgery. This method only needs to establish an SSM database before surgery to carry out surgical planning for different patients. However, due to the limitations of ultrasound imaging mechanism, ultrasound two-dimensional images have low resolution, high noise, and poor image quality. It is impossible to accurately obtain the digital point cloud of the bone surface. In addition, the additional image segmentation link prolongs the doctor's learning curve and is inefficient. In addition, the B-ultrasound equipment is expensive and the probe is large, which cannot accurately locate the bone surface. The above problems lead to the poor accuracy of the lower limb bone surface reconstruction system based on B-ultrasound images, and it is impossible to give doctors correct guidance.

[0006] In addition, after obtaining the bone surface template point cloud data, how to drive the SSM to fit it is an important factor in determining the reconstruction and tracking accuracy. Bruce et al. established an objective function based on Euclidean distance, manually selected feature points on the SSM surface, established the correspondence between the SSM and the template point cloud, and drove the SSM deformation by minimizing the distance between the corresponding points. However, the surgical operation area of ​​TKA is quite limited. The manual marking method is not only inefficient, but also difficult to fully obtain surface feature points. On this basis, Accuracy of femur reconstruction from sparse geometric data using astatistical shape model introduced additional anatomical parameters in the objective function, such as the trochanter to epicondyle length, epicondyle width, etc. Although it improved the accuracy of reconstruction, it did not fundamentally solve the problem.

[0007] In summary, the existing human lower limb bone surface reconstruction system is difficult to meet the requirements of preoperative planning. Summary of the invention

[0008] The present invention is proposed to solve the problems existing in the prior art, and its purpose is to provide a method and system for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment.

[0009] The technical solution of the present invention is: a method for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment, comprising the following steps:

[0010] Ⅰ. Send and receive type A ultrasonic equipment signals to obtain ultrasonic data;

[0011] Ⅱ. Process the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object;

[0012] III. Calibrate the A-type ultrasonic probe, convert the one-dimensional depth information into three-dimensional coordinate points in space, and collect the three-dimensional coordinates of the surface points of the lower limb bones of the human body;

[0013] IV. Establish a lower limb bone model database and calculate the SSM deformable general model;

[0014] Ⅴ. Reconstruct the bone surface of human lower limbs.

[0015] Furthermore, step I performs the transmission and reception of type A ultrasonic equipment signals to obtain ultrasonic data. The specific process is as follows:

[0016] First, the A-type ultrasound equipment probe adopts a four-element form, which generates a 7.5MHz pulse signal under the control of the main control chip;

[0017] Then, the pulse signal passes through the power amplifier to drive the transducer to emit pulsed ultrasonic waves;

[0018] After that, it is transmitted to the analog-to-digital conversion chip for data conversion after bandpass filtering;

[0019] Finally, the converted ultrasonic data is stored in the main control chip.

[0020] Furthermore, step II processes the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object. The specific process is as follows:

[0021] First, the converted ultrasonic data carries the biological tissue information of the human body;

[0022] Then, the ultrasonic data is subjected to Gaussian filtering, Hilbert transform to extract data envelope, binarization, and pulse width recognition in sequence;

[0023] Finally, the first five envelope bands of each data set are extracted according to intensity, and the average distance value is calculated, which is recorded as d.

[0024] Furthermore, step III calibrates the A-type ultrasonic probe, converts the one-dimensional depth information into three-dimensional coordinate points in space, and collects the three-dimensional coordinates of the surface points of the lower limb bones of the human body. The specific process is as follows:

[0025] First, immerse a regular aluminum ball in a water tank, and obtain the center coordinates C of the aluminum ball under the optical tracking system;

[0026] Then, the tracking frame DRO was fixed on the A-type ultrasonic probe, and the A-type ultrasonic probe was clamped by the mechanical arm to measure the surface points of the aluminum ball 30 times, and each measurement result was recorded as (0,0,d i ),i=1,...,30;

[0027] At this time, the coordinates of the i-th spherical point in the world coordinate system are expressed as:

[0028]

[0029] In the formula, It is the attitude description matrix of the tracking frame DRO coordinate system relative to the world coordinate system. It is the attitude description matrix of the A-type ultrasound probe coordinate system relative to the tracking frame DRO coordinate system;

[0030] After that, The solution is iteratively solved by minimizing the following equation:

[0031]

[0032] Finally, the A-type ultrasound device is used to detect the three-dimensional coordinates of the surface points of the lower limb bones of the human body, and it is recorded as the template point cloud:

[0033]

[0034] Where D i It is the depth from the skin to the bone surface.

[0035] Furthermore, in step IV, a lower limb bone model database is established to calculate the SSM deformable general model. The specific process is as follows:

[0036] The medical image processing software Mimics was used to segment and reconstruct multiple human lower limb CT bone model data, export them as STL models, align all models, and perform principal component analysis to calculate the SSM deformable general model:

[0037]

[0038] In the formula, represents the average model, n represents the number of principal components retained after principal component analysis, and p i represents the eigenvector of the model after principal component analysis, b i Represents the deformation parameter of the model. By changing this value, a new shape S can be generated. new ,λ i Represents the eigenvalues ​​of the model after principal component analysis.

[0039] Furthermore, step V performs the reconstruction of the surface of the lower limb bone of the human body, and the specific process is as follows:

[0040] First, the template point cloud s collected in step III i The average model obtained in step IV Registration, let b = (b1, b2, ..., b n ), the relationship between each template point cloud and the corresponding point of the average model is expressed as the following formula, which is minimized to solve the deformation parameters b, t, r and ω:

[0041]

[0042] In the formula, S fitted (t,r,ω,b) represents the translation t, rotation r, isotropic scaling ω and application of parameterized deformation b to the deformable general model, a is the penalty coefficient of the Mahalanobis distance, and m is the Mahalanobis distance;

[0043] Finally, after determining various parameters, S fitted It is a lower limb bone surface reconstruction model, which is displayed on a computer screen for doctors to interactively operate and view.

[0044] A system for human lower limb bone surface reconstruction method based on A-type ultrasonic equipment, comprising the following modules:

[0045] A-type ultrasound equipment submodule collects tissue depth information from the skin to the surface of the lower limb bones, completes the extraction and processing of ultrasound data in the host computer, and displays and stores the obtained one-dimensional ultrasound depth image;

[0046] A type ultrasound device calibration submodule converts the one-dimensional ultrasound depth information collected by the optical tracking device into three-dimensional point cloud spatial information, and defines the point cloud spatial information as a template point cloud for subsequent reconstruction of the lower limb bone surface deformation;

[0047] The statistical shape model construction submodule is used to generate the SSM deformable general model, extract the lower limb bone area from the CT data of the lower limb bone anatomical parts and perform three-dimensional reconstruction, and export the reconstruction results as an STL network model; the statistical shape model construction submodule also performs multiple three-dimensional reconstructions according to the model capacity of the training set, and aligns and reduces the dimension of all training models;

[0048] The deformable lower limb bone surface reconstruction submodule aligns the template point cloud collected by the A-type ultrasound device with the SSM deformable general model, and drives the SSM model to deform, completes the human lower limb bone surface reconstruction, and generates a patient-specific bone model;

[0049] The visualization interaction submodule builds a three-dimensional virtual scene and displays necessary information such as the A-ultrasound acquisition probe and the reconstructed lower limb bone surface on the computer screen.

[0050] Furthermore, the A-type ultrasonic equipment submodule includes a transducer, a drive control circuit, an echo sampling circuit, and a host computer.

[0051] The beneficial effects of the present invention are as follows:

[0052] The present invention introduces a type A ultrasonic device into the lower limb bone surface reconstruction system, automatically detects and identifies bone surface echo signals, and converts them into three-dimensional space point clouds through calibration. The operation is simple, the cost is low, and additional radiation is avoided.

[0053] The present invention combines the SSM deformable general model to complete the surface reconstruction of the human lower limb bone, and introduces an improved deformation driving algorithm to reduce the error caused by the picking point noise with high precision.

[0054] The present invention provides a visual interactive interface, provides doctors with accurate three-dimensional bone surface information, and improves diagnostic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is the overall flow chart of the present invention;

[0056] Figure 2 It is the overall framework diagram of the present invention;

[0057] Figure 3 It is a sub-flow chart of automatic identification of type A ultrasonic echo signal in the present invention;

[0058] Figure 4 It is a schematic diagram of type A ultrasonic calibration in the present invention;

[0059] Figure 5 It is a sub-flow chart of human lower limb bone surface reconstruction in the present invention;

[0060] Figure 6 It is the visual interactive interface of the present invention. DETAILED DESCRIPTION

[0061] Hereinafter, the present invention will be described in detail with reference to the accompanying drawings and embodiments:

[0062] like Figures 1 to 6 As shown, a method for reconstructing the surface of human lower limb bones based on an A-type ultrasonic device comprises the following steps:

[0063] Ⅰ. Send and receive type A ultrasonic equipment signals to obtain ultrasonic data;

[0064] Ⅱ. Process the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object;

[0065] III. Calibrate the A-type ultrasonic probe, convert the one-dimensional depth information into three-dimensional coordinate points in space, and collect the three-dimensional coordinates of the surface points of the lower limb bones of the human body;

[0066] IV. Establish a lower limb bone model database and calculate the SSM deformable general model;

[0067] Ⅴ. Reconstruct the bone surface of human lower limbs.

[0068] Step I performs A-type ultrasonic equipment signal transmission and reception to obtain ultrasonic data. The specific process is as follows:

[0069] First, the A-type ultrasound equipment probe adopts a four-element form, which generates a 7.5MHz pulse signal under the control of the main control chip;

[0070] Then, the pulse signal passes through the power amplifier to drive the transducer to emit pulsed ultrasonic waves;

[0071] After that, it is transmitted to the analog-to-digital conversion chip for data conversion after bandpass filtering;

[0072] Finally, the converted ultrasonic data is stored in the main control chip.

[0073] Step II processes the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object. The specific process is as follows:

[0074] First, the converted ultrasonic data carries the biological tissue information of the human body;

[0075] Then, the ultrasonic data is subjected to Gaussian filtering, Hilbert transform to extract data envelope, binarization, and pulse width recognition in sequence;

[0076] Finally, the first five envelope bands of each data set are extracted according to intensity, and the average distance value is calculated, which is recorded as d.

[0077] Step III calibrates the A-type ultrasonic probe, converts the one-dimensional depth information into three-dimensional coordinate points in space, and collects the three-dimensional coordinates of the surface points of the lower limb bones of the human body. The specific process is as follows:

[0078] First, immerse a regular aluminum ball in a water tank, and obtain the center coordinates C of the aluminum ball under the optical tracking system;

[0079] Then, the tracking frame DRO was fixed on the A-type ultrasonic probe, and the A-type ultrasonic probe was clamped by the mechanical arm to measure the surface points of the aluminum ball 30 times, and each measurement result was recorded as (0,0,d i ),i=1,...,30;

[0080] At this time, the coordinates of the i-th spherical point in the world coordinate system are expressed as:

[0081]

[0082] In the formula, It is the attitude description matrix of the tracking frame DRO coordinate system relative to the world coordinate system. It is the attitude description matrix of the A-type ultrasound probe coordinate system relative to the tracking frame DRO coordinate system;

[0083] After that, The solution is iteratively solved by minimizing the following equation:

[0084]

[0085] Finally, the A-type ultrasound device is used to detect the three-dimensional coordinates of the surface points of the lower limb bones of the human body, and it is recorded as the template point cloud:

[0086]

[0087] Where D i It is the depth from the skin to the bone surface.

[0088] Step IV: Establish a lower limb bone model database and calculate the SSM deformable general model. The specific process is as follows:

[0089] The medical image processing software Mimics was used to segment and reconstruct multiple human lower limb CT bone model data, export them as STL models, align all models, and perform principal component analysis to calculate the SSM deformable general model:

[0090]

[0091] In the formula, represents the average model, n represents the number of principal components retained after principal component analysis, and p i represents the eigenvector of the model after principal component analysis, b i Represents the deformation parameter of the model. By changing this value, a new shape S can be generated. new ,λ i Represents the eigenvalues ​​of the model after principal component analysis.

[0092] Step V is to reconstruct the surface of the lower limb bones of the human body. The specific process is as follows:

[0093] First, the template point cloud s collected in step III i The average model obtained in step IV Registration, let b = (b1, b2, ..., b n ), the relationship between each template point cloud and the corresponding point of the average model is expressed as the following formula, which is minimized to solve the deformation parameters b, t, r and ω:

[0094]

[0095] In the formula, S fitted(t,r,ω,b) represents the translation t, rotation r, isotropic scaling ω and application of parameterized deformation b to the deformable general model, a is the penalty coefficient of the Mahalanobis distance, and m is the Mahalanobis distance;

[0096] Finally, after determining various parameters, S fitted It is a lower limb bone surface reconstruction model, which is displayed on a computer screen for doctors to interactively operate and view.

[0097] A system for human lower limb bone surface reconstruction method based on A-type ultrasonic equipment, comprising the following modules:

[0098] A-type ultrasound equipment submodule collects tissue depth information from the skin to the surface of the lower limb bones, completes the extraction and processing of ultrasound data in the host computer, and displays and stores the obtained one-dimensional ultrasound depth image;

[0099] A type ultrasound device calibration submodule converts the one-dimensional ultrasound depth information collected by the optical tracking device into three-dimensional point cloud spatial information, and defines the point cloud spatial information as a template point cloud for subsequent reconstruction of the lower limb bone surface deformation;

[0100] The statistical shape model construction submodule is used to generate the SSM deformable general model, extract the lower limb bone area from the CT data of the lower limb bone anatomical parts and perform three-dimensional reconstruction, and export the reconstruction results as an STL network model; the statistical shape model construction submodule also performs multiple three-dimensional reconstructions according to the model capacity of the training set, and aligns and reduces the dimension of all training models;

[0101] The deformable lower limb bone surface reconstruction submodule aligns the template point cloud collected by the A-type ultrasound device with the SSM deformable general model, and drives the SSM model to deform, completes the human lower limb bone surface reconstruction, and generates a patient-specific bone model;

[0102] The visualization interaction submodule builds a three-dimensional virtual scene and displays necessary information such as the A-ultrasound acquisition probe and the reconstructed lower limb bone surface on the computer screen.

[0103] The A-type ultrasonic equipment submodule includes a transducer, a drive control circuit, an echo sampling circuit, and a host computer.

[0104] Embodiment 1

[0105] A method for reconstructing the surface of human lower limb bones based on an A-type ultrasonic device comprises the following steps:

[0106] Ⅰ. Perform A-type ultrasonic equipment signal transmission and reception to obtain ultrasonic data, as shown in the attached Figure 2 shown.

[0107] Under the action of the main control chip, a 7.5MHz pulse signal is generated, which is driven by the power amplifier to emit pulsed ultrasonic waves. The ultrasonic transducer adopts a four-element structure, with elements 1 and 2 emitting ultrasonic waves and elements 3 and 4 receiving ultrasonic waves. It is then band-pass filtered and transmitted to the digital-to-analog conversion chip for data conversion. Finally, the converted ultrasonic data is stored in the main control chip.

[0108] At the same time, the external power supply supplies power to the main control chip and the power amplifier, and the host computer controls the main control chip.

[0109] Ⅱ. Process the ultrasonic data to obtain the distance from the ultrasonic probe to the object being detected, as shown in the attached figure. Figure 3 shown.

[0110] The ultrasonic data stored in the main control chip is extracted, and the distance d from the ultrasonic probe to the detected object is obtained through the bone surface automatic recognition algorithm.

[0111] Specifically, the automatic bone surface distance recognition algorithm is as follows:

[0112] Step 2a. Extract the ultrasonic echo data in the main control chip and record it as D1, which carries distance information.

[0113] Step 2b. Perform Gaussian filtering on the ultrasonic echo data, denoted as D2:

[0114]

[0115] In the formula, the standard normal distribution μ=0, σ=1, which suppresses noise and smoothes the signal.

[0116] Step 2c. Continue to perform Hilbert transform on D2 to extract the data envelope, recorded as D3:

[0117]

[0118] Where n is the position index of the output signal, indicating the value of the resulting convolution signal at the nth position, and m is the intermediate variable for summation, indicating all position indexes of the traversed signal.

[0119] Step 2d. Binarize D3 using a dynamic threshold to obtain D4, then perform pulse width identification on D4 to retain the envelope band with the same width range as the transmitted wave, and filter out other waveforms;

[0120] Step 2e. Calculate the distance d:

[0121]

[0122] Where K represents the compensation coefficient, v is the sound velocity in the medium to be measured, and P end is the peak position of the echo, Pstart is the peak position of the transmitted wave, and f is the sampling frequency.

[0123] III. Calibrate the A-type ultrasonic probe, convert the one-dimensional depth information into three-dimensional coordinate points in space, and collect the three-dimensional coordinates of the surface points of the lower limb bones of the human body, as shown in the attached Figure 4 shown.

[0124] The A-type ultrasonic probe is calibrated to convert the one-dimensional depth information d into three-dimensional spatial coordinate points.

[0125] Specifically, the A-type ultrasonic probe is calibrated to obtain the spatial three-dimensional point cloud coordinates, as follows:

[0126] Step 3a. Immerse an aluminum ball with a known radius R in a water tank, and obtain the coordinates C of the center of the aluminum ball under an optical tracking system;

[0127] Step 3b. Fix the tracking frame DRO on the A-type ultrasound probe and record the posture description matrix of the DRO coordinate system relative to the world coordinate system

[0128] Step 3c. Use the robot arm to hold the A-type ultrasonic probe to measure the surface of the aluminum ball 30 times, and record each measurement result as (0, 0, d i ),i=1,…,30.

[0129] At this time, the coordinates of the i-th spherical point in the world coordinate system are expressed as:

[0130]

[0131] In the formula, It is the posture description matrix of the A-type ultrasound probe coordinate system relative to the tracking frame DRO coordinate system. It is necessary to iteratively solve by minimizing the following formula:

[0132]

[0133] Where, distance(C, World q i ) is the distance between the center of the aluminum ball and the i-th point on the surface of the ball in the world coordinate system; unknown parameter The matrix form of is:

[0134]

[0135] Where α, β, and γ are the rotation angles of the A-type ultrasound probe coordinate system around the x, y, and z axes of the DRO coordinate system, respectively. x , T y , T zThey are respectively the translation transformation of the A-type ultrasound probe coordinate system along the x, y, and z axes of the DRO coordinate system.

[0136] Step 3d. Save the calculated This transformation can be used for each acquisition while ensuring that the relative position of the DRO and the A-mode ultrasound is fixed.

[0137] Step 3e. Collect the three-dimensional coordinates of the surface points of the lower limb bones of the human body for subsequent deformation and reconstruction, and record them as the template point cloud s i :

[0138]

[0139] IV. Establish a lower limb bone model database and calculate the SSM deformable general model, as shown in the attached Figure 5 shown.

[0140] Specifically, the present invention adopts an SSM construction algorithm based on a coherent point drift algorithm, which is as follows:

[0141] Step 4a. Use the medical image processing software Mimics to segment and reconstruct multiple human lower limb CT bone model data, export them into STL models, and construct an SSM bone model dataset.

[0142] Step 4b. Smooth the bone models in the training set and uniformly set the number of vertices of the STL model to M.

[0143] Step 4c. Extract the model vertex coordinates of the bone model in the training set and express them in the form of a point set:

[0144]

[0145] Where N is the number of samples in the training set, and x, y, and z represent the three values ​​of the three-dimensional coordinate points.

[0146] Step 4d. Select any model in the training set as the reference model, and use the coherent point drift algorithm to align the remaining N-1 models with it to obtain the correspondence between the points in the models.

[0147] Step 4e. Based on the correspondence of the points, perform a Prospectus analysis on the original model and align all the training models.

[0148] Step 4f. Calculate the average model:

[0149]

[0150] Calculate the covariance matrix:

[0151]

[0152] The SSM deformable general model is finally obtained through principal component analysis:

[0153]

[0154] In the formula, n represents the number of principal components retained after principal component analysis, and p i represents the eigenvector of the covariance matrix, b i Represents the deformation parameter of the model. By changing this value, a new shape S can be generated. new ,λ i represents the eigenvalues ​​of the covariance matrix.

[0155] Ⅴ. Reconstruct the bone surface of human lower limbs.

[0156] The template point cloud s obtained from the patient's bone surface in step III i , and the average model generated in step IV Perform registration, let b = (b1, b2, ..., b n ), the relationship between each template point cloud and the corresponding point of the average model is expressed as the following formula, which is minimized to solve the deformation parameters b, t, r and ω:

[0157]

[0158] In the formula, S fitted (t,r,ω,b) represents the translation t, rotation r, isotropic scaling ω and application of parameterized deformation b to the deformable general model, a is the penalty coefficient of the Mahalanobis distance, and m is the Mahalanobis distance.

[0159] After determining the parameters, S fitted It is the lower limb bone surface reconstruction model.

[0160] Figure 6 The following is a visualization function panel that integrates the important functions of each module. The visualization window is used to display the spatial position of the A-type ultrasound probe and the bone model in real time; the connection status display bar is used to display the status of the tracking frame; the operation bar can realize the module function by clicking. A three-dimensional virtual scene is built to render the reconstructed lower limb bone surface and the tracked A-type ultrasound probe in real time on the display device, providing an intuitive reference for doctors.

[0161] The present invention can reconstruct the bone surface of the patient's lower limbs with high precision before surgery. The A-type ultrasonic device is introduced into the system, and the corresponding bone surface echo automatic recognition algorithm is designed, which reduces the difficulty of acquisition, avoids additional radiation, and reduces costs. At the same time, combined with the SSM deformable general model, the deformation-driven algorithm is improved, the error caused by the pickup point noise is reduced, and the accurate and efficient reconstruction of the bone surface of the human lower limb is completed. A visual interactive interface is provided to provide doctors with accurate three-dimensional bone surface information, improve diagnostic accuracy and efficiency, and the present invention can meet the needs of preoperative surgical planning.

[0162] The above description of the present invention is merely illustrative and not restrictive, so the embodiments of the present invention are not limited to the above specific embodiments. If a person skilled in the art is inspired by the above, without departing from the scope of protection of the present invention and the claims, other changes or modifications are made, which belong to the protection scope of the present invention.

Claims

1. A method for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment, characterized in that: The following steps are involved: Ⅰ. Send and receive type A ultrasonic equipment signals to obtain ultrasonic data; Ⅱ. Process the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object; III. Calibrate the A-type ultrasonic probe, convert the one-dimensional depth information into three-dimensional coordinate points in space, and collect the three-dimensional coordinates of the surface points of the lower limb bones of the human body; IV. Establish a lower limb bone model database and calculate the SSM deformable general model; Ⅴ. Reconstruct the bone surface of human lower limbs.

2. A method for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment according to claim 1, characterized in that: Step I performs A-type ultrasonic equipment signal transmission and reception to obtain ultrasonic data. The specific process is as follows: First, the A-type ultrasound equipment probe adopts a four-element form, which generates a 7.5MHz pulse signal under the control of the main control chip; Then, the pulse signal passes through the power amplifier to drive the transducer to emit pulsed ultrasonic waves; After that, it is transmitted to the analog-to-digital conversion chip for data conversion after bandpass filtering; Finally, the converted ultrasonic data is stored in the main control chip.

3. The method for reconstructing the surface of human lower limb bones based on type A ultrasonic equipment according to claim 1, characterized in that: Step II processes the obtained ultrasonic data to obtain the distance from the ultrasonic probe to the detected object. The specific process is as follows: First, the converted ultrasonic data carries the biological tissue information of the human body; Then, the ultrasonic data is subjected to Gaussian filtering, Hilbert transform to extract data envelope, binarization, and pulse width recognition in sequence; Finally, the first five envelope bands of each data set are extracted according to intensity, and the average distance value is calculated, which is recorded as d.

4. The method for reconstructing the surface of human lower limb bones based on type A ultrasonic equipment according to claim 1, characterized in that: Step III calibrates the A-type ultrasonic probe, converts the one-dimensional depth information into three-dimensional coordinate points in space, and collects the three-dimensional coordinates of the surface points of the lower limb bones of the human body. The specific process is as follows: First, immerse a regular aluminum ball in a water tank, and obtain the center coordinates C of the aluminum ball under the optical tracking system; Then, the tracking frame DRO was fixed on the A-type ultrasonic probe, and the A-type ultrasonic probe was clamped by the mechanical arm to measure the surface points of the aluminum ball 30 times, and each measurement result was recorded as (0,0,d i ),i=1,...,30; At this time, the coordinates of the i-th spherical point in the world coordinate system are expressed as: In the formula, It is the attitude description matrix of the tracking frame DRO coordinate system relative to the world coordinate system. It is the attitude description matrix of the A-type ultrasound probe coordinate system relative to the tracking frame DRO coordinate system; After that, The solution is iteratively solved by minimizing the following equation: Finally, the A-type ultrasound device is used to detect the three-dimensional coordinates of the surface points of the lower limb bones of the human body, and it is recorded as the template point cloud: Where D i It is the depth from the skin to the bone surface.

5. The method for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment according to claim 1, characterized in that: Step IV: Establish a lower limb bone model database and calculate the SSM deformable general model. The specific process is as follows: The medical image processing software Mimics was used to segment and reconstruct multiple human lower limb CT bone model data, export them as STL models, align all models, and perform principal component analysis to calculate the SSM deformable general model: In the formula, represents the average model, n represents the number of principal components retained after principal component analysis, and p i represents the eigenvector of the model after principal component analysis, b i Represents the deformation parameter of the model. By changing this value, a new shape S can be generated. new ,λ i Represents the eigenvalues ​​of the model after principal component analysis.

6. The method for reconstructing the surface of human lower limb bones based on type A ultrasonic equipment according to claim 1, characterized in that: Step V is to reconstruct the surface of the lower limb bones of the human body. The specific process is as follows: First, the template point cloud s collected in step III i Align with the average model S obtained in step IV, let b = (b1, b2, ..., b n ), the relationship between each template point cloud and the corresponding point of the average model is expressed as the following formula, which is minimized to solve the deformation parameters b, t, r and ω: In the formula, S fitted (t,r,ω,b) represents the translation t, rotation r, isotropic scaling ω and application of parameterized deformation b to the deformable general model, a is the penalty coefficient of the Mahalanobis distance, and m is the Mahalanobis distance; Finally, after determining various parameters, S fitted It is a lower limb bone surface reconstruction model, which is displayed on a computer screen for doctors to interactively operate and view.

7. A system for human lower limb bone surface reconstruction method based on A-type ultrasonic equipment, characterized in that: Includes the following modules: A-type ultrasound equipment submodule collects tissue depth information from the skin to the surface of the lower limb bones, completes the extraction and processing of ultrasound data in the host computer, and displays and stores the obtained one-dimensional ultrasound depth image; A type ultrasound device calibration submodule converts the one-dimensional ultrasound depth information collected by the optical tracking device into three-dimensional point cloud spatial information, and defines the point cloud spatial information as a template point cloud for subsequent reconstruction of the lower limb bone surface deformation; The statistical shape model construction submodule is used to generate the SSM deformable general model, extract the lower limb bone area from the CT data of the lower limb bone anatomical parts and perform three-dimensional reconstruction, and export the reconstruction results as an STL network model; the statistical shape model construction submodule also performs multiple three-dimensional reconstructions according to the model capacity of the training set, and aligns and reduces the dimension of all training models; The deformable lower limb bone surface reconstruction submodule aligns the template point cloud collected by the A-type ultrasound device with the SSM deformable general model, and drives the SSM model to deform, completes the human lower limb bone surface reconstruction, and generates a patient-specific bone model; The visualization interaction submodule builds a three-dimensional virtual scene and displays necessary information such as the A-ultrasound acquisition probe and the reconstructed lower limb bone surface on the computer screen.

8. The system of the method for reconstructing the surface of human lower limb bones based on A-type ultrasonic equipment according to claim 7, characterized in that: The A-type ultrasonic equipment submodule includes a transducer, a drive control circuit, an echo sampling circuit, and a host computer.

Images