An ultrasonic robotic inspection quality adaptive optimization method

Abstract

The application discloses a kind of ultrasonic robot examination quality self-adaptive optimization method, it is related to the field of ultrasonic robot, including: obtaining the anatomical structure feature information of the part to be checked, based on anatomical feature information setting ultrasonic robot ultrasonic probe initial operating frequency and initial focusing depth;Control the ultrasonic probe of ultrasonic robot and execute scanning operation according to initial operating frequency and initial focusing depth, real-time acquisition ultrasonic image and extract the quality quantization index of image, and synchronously obtain the current probe posture data of ultrasonic robot in scanning process;The application sets up the initial operating parameter of ultrasonic probe by obtaining the anatomical structure feature of the part to be checked, improves initial scanning pertinence to reduce invalid operation, real-time acquisition ultrasonic image quality quantization index and probe posture data, can master scanning state in time, combines anatomical feature and sets dynamic quality threshold, so that image quality determination is more in line with actual anatomical situation to avoid the deviation of the determination of uniform threshold.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic robot technology, specifically to an adaptive optimization method for ultrasonic robot inspection quality. Background Technology

[0002] Ultrasound robots are an innovative technology in the medical field, integrating ultrasound imaging with the precise control capabilities of robots. They have been widely used in diagnosis and treatment in various scenarios. They can accurately locate lesions in interventional treatments, such as guiding puncture biopsy and tumor ablation, reducing human error. They can provide high-definition ultrasound images in real time during surgery to assist doctors in judging anatomical structures and improve surgical safety. They can also be used in intensive care and other scenarios to realize automated bedside ultrasound examinations, reduce the workload of medical staff, and have the advantages of precision, efficiency, and minimally invasiveness.

[0003] In the process of outputting ultrasound images, the quality of existing ultrasound robots is affected by various external factors and interference from the ultrasound robot itself, resulting in a deterioration in ultrasound image quality.

[0004] To address this, we propose an adaptive optimization method for the quality of ultrasound robot inspection. Summary of the Invention

[0005] In view of the above-mentioned shortcomings of the existing technology, the present invention provides an adaptive optimization method for ultrasound robot inspection quality, which can effectively solve the problems of the existing technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions;

[0007] This invention discloses an adaptive optimization method for the quality of ultrasound robot examination, comprising:

[0008] The system acquires anatomical structural features of the area to be examined, sets the initial operating frequency and initial focusing depth of the ultrasound probe of the ultrasound robot based on the anatomical features, controls the ultrasound probe of the ultrasound robot to perform scanning operations according to the initial operating frequency and initial focusing depth, acquires ultrasound images in real time and extracts the quality quantification indicators of the images, and simultaneously acquires the current probe posture data of the ultrasound robot during the scanning process.

[0009] Based on the anatomical features of the area to be examined, a dynamic quality threshold matching these features is set. The real-time extracted quality quantification index is compared with the dynamic quality threshold to determine whether the current ultrasound image quality is acceptable. If the current ultrasound image quality is deemed unacceptable, the working frequency, focusing depth, and scanning angle of the ultrasound robot are adjusted collaboratively based on the deviation between the quality quantification index and the dynamic quality threshold, combined with the acquired current probe posture data. The ultrasound robot is then controlled to perform a second scan according to the adjusted parameters, acquiring the ultrasound image from the second scan and re-extracting the quality quantification index. The newly extracted quality quantification index is compared with the dynamic quality threshold again to verify whether the image quality is acceptable. If the image quality of the second scan is still unacceptable, the parameter adjustment and scanning verification process is repeated until the image quality is acceptable. If the image quality is still unacceptable after a preset number of scans, the process jumps to the anatomical feature information acquisition stage of the area to be examined to reset and execute.

[0010] Furthermore, the operation of acquiring the anatomical structural feature information of the area to be examined includes:

[0011] Collect medical imaging anatomical data of the area to be examined and real-time palpation pressure data of the ultrasound robot;

[0012] Three-dimensional reconstruction of medical imaging anatomical data is performed to extract organ contours, tissue layer boundaries, blood vessel course characteristics, and tissue density distribution of the area to be examined.

[0013] Spatial mapping of real-time palpation pressure data was performed to obtain tissue elasticity characteristics of different regions of the site to be examined.

[0014] The data of organ contour, tissue layer boundaries, blood vessel course characteristics, tissue density distribution and tissue elasticity characteristics are fused to construct the anatomical structure feature vector of the site to be examined.

[0015] Calculate the initial operating frequency of the ultrasound probe based on the anatomical structure feature vector. ;

[0016] and initial focus depth ;

[0017] In the formula: For the influence coefficients of tissue elasticity and blood vessel distribution density on frequency; The elasticity of the tissue at the site to be examined is expressed in kPa. The density of blood vessels in the area to be examined is expressed in units of vessels / mm². The basic operating frequency corresponding to the part to be inspected; These are the influence coefficients of average tissue thickness and tissue density on depth of focus. This represents the average thickness of tissue layers in the anatomical feature vector; The tissue density of the area to be examined is expressed in g / cm³. The baseline focusing depth corresponding to the area to be inspected;

[0018] The initial operating frequency was calculated. With initial focus depth Then, it is verified whether it is within the range of the ultrasonic probe's hardware operating parameters. If it is outside the range, it is corrected according to the probe's maximum / minimum parameter limits to finally determine the initial operating parameters used for scanning.

[0019] Furthermore, the extraction logic for the ultrasound image quality quantification index is as follows:

[0020] Signal-to-noise ratio, contrast, and edge sharpness are used as quantitative indicators of ultrasound image quality, and a comprehensive quality index is constructed based on this. ;

[0021] In the formula: , , These are the weighting coefficients for signal-to-noise ratio, contrast, and edge sharpness, and their sum is 1. Signal-to-noise ratio; For contrast; For edge sharpness, the calculation formula is: ,in This represents the total number of pixels at the edge of the target anatomical structure in the ultrasound image. This represents the grayscale gradient value of the i-th edge pixel calculated by the edge detection operator.

[0022] Furthermore, the aforementioned , , The value of follows:

[0023] The diagnostic demand coefficient F for the area to be examined is preset according to the purpose of the examination;

[0024] The purposes of the examinations include structural imaging, blood flow imaging, and lesion screening, and:

[0025] When the purpose of the inspection is structural imaging, set ;

[0026] When the purpose of the examination is blood flow imaging, set ;

[0027] When the purpose of the examination is lesion screening, set ;

[0028] but , This represents the i-th weight coefficient. ∈{ , , }

[0029] Furthermore, when setting a dynamic quality threshold that matches the anatomical features of the area to be examined, the following applies:

[0030] Key anatomical parameters strongly correlated with ultrasound imaging quality are extracted from the anatomical structure feature vector, including tissue density, blood vessel distribution density, and organ thickness of the area to be examined.

[0031] A calculation model for dynamic quality thresholds is constructed based on key anatomical parameters. The model expression is as follows:

[0032] ;

[0033] In the formula: For dynamic quality thresholds; The threshold influence coefficients corresponding to tissue density, blood vessel distribution density, and organ thickness; The tissue density of the area to be examined; Blood vessel distribution density; For organ thickness; This represents the baseline quality threshold corresponding to the area to be inspected.

[0034] Furthermore, when the current ultrasound image quality is deemed unqualified, the deviation between the quality quantification index and the dynamic quality threshold is calculated. This deviation is then combined with the acquired current probe posture data to collaboratively adjust the ultrasound probe's operating frequency, focusing depth, and the ultrasound robot's probe scanning angle. ;

[0035] based on Based on the current probe orientation data, construct a collaborative adjustment model for the ultrasonic probe's operating parameters:

[0036] ;

[0037] In the formula: This is the adjustment amount for the operating frequency; This is the operating frequency adjustment factor; The probe's elevation angle; To focus on the depth of adjustment; To focus on depth adjustment coefficient; The contact pressure between the probe and the area to be inspected; This is the standard contact pressure between the probe and the area to be inspected. This is the amount of adjustment for the probe scanning angle; This is the scanning angle adjustment factor; This is the roll angle of the probe.

[0038] Furthermore, when the ultrasonic robot acquires the current probe posture data during the scanning process, it collects the contact pressure data between the probe and the part to be inspected through a preset six-dimensional force sensor at the end of the ultrasonic robot, and collects the position coordinates (X,Y,Z) and attitude angles of the probe through the robot joint encoder, wherein the attitude angles include pitch angle, roll angle and yaw angle.

[0039] The acquired position coordinates and attitude angles are simultaneously filtered in real time; the contact pressure data is smoothed in the time domain.

[0040] Using the robot system's reference clock as a time reference, the filtered position coordinates, attitude angle data, and smoothed contact pressure data are timestamped. At the same time, using the center point of the probe end as a spatial reference, the contact pressure data is mapped to the spatial coordinate system of this reference to complete spatiotemporal alignment and obtain the complete attitude dataset of the current probe.

[0041] Furthermore, the preset number of scans is adapted to... , Indicates the preset number of scans. Indicates the frequency adjustment factor. This represents the initial deviation between the quality quantification index and the dynamic quality threshold during the first scan. This represents the complexity coefficient of the anatomical features of the area to be examined. Indicates the number of baseline scans. Customized by the user. ≥3;

[0042] When the preset number of scans is reached and the image quality is still unsatisfactory, a second acquisition of anatomical structure feature information is triggered:

[0043] Multimodal anatomical data of the area to be examined are reacquired to correct the initially constructed anatomical feature vector. Based on the corrected anatomical feature vector, the initial operating frequency and initial focusing depth of the ultrasound probe are reset to restart the ultrasound robot scanning process.

[0044] Furthermore, after coordinating the adjustment of the ultrasonic probe's operating frequency, focusing depth, and the ultrasonic robot's probe scanning angle, the propagation time of the ultrasonic beam in the tissue to be examined is simultaneously calculated based on the adjusted operating frequency and focusing depth before performing the second scan. ,in Indicates the adjusted focus depth. This indicates the speed at which ultrasound propagates through the tissue being examined.

[0045] Further based on the time of transmission Adjust the probe scanning trigger timing of the ultrasonic robot to synchronize the transmission and reception of ultrasonic signals.

[0046] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects:

[0047] This invention provides an adaptive optimization method for ultrasound robot examination quality. During execution, the method sets initial working parameters of the ultrasound probe by acquiring anatomical features of the area to be examined, improving the targeting of the initial scan and reducing invalid operations. Real-time acquisition of ultrasound image quality quantification indicators and probe posture data allows for timely monitoring of the scanning status. Dynamic quality thresholds are set based on anatomical features, making image quality judgment more closely aligned with actual anatomical conditions to avoid bias from uniform thresholds. When image quality is unsatisfactory, the working frequency, focusing depth, and scanning angle are adjusted collaboratively based on the deviation value and probe posture to precisely optimize scanning parameters. Secondary scanning verification and repeated adjustments continuously improve image quality. If the preset number of scans fails to meet the requirements, the acquisition of anatomical features and the initial parameter settings are reset to address quality issues caused by anatomical changes. Synchronous adjustment of the scanning trigger sequence ensures ultrasound signal synchronization, thereby improving the overall quality and efficiency of ultrasound examination and reducing reliance on manual intervention. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0049] Figure 1 This is a flowchart illustrating an adaptive optimization method for the quality of ultrasound robot examinations. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0051] The present invention will be further described below with reference to embodiments.

[0052] Example:

[0053] This embodiment provides an adaptive optimization method for the quality of ultrasound robot examination, such as... Figure 1As shown, it includes:

[0054] Obtain anatomical structural feature information of the area to be examined, and set the initial working frequency and initial focusing depth of the ultrasound probe of the ultrasound robot based on the anatomical feature information;

[0055] The operations for obtaining anatomical structural feature information of the area to be examined include:

[0056] Collect medical imaging anatomical data and real-time palpation pressure data of the area to be examined; perform three-dimensional reconstruction of the medical imaging anatomical data to extract organ contours, tissue layer boundaries, blood vessel course characteristics and tissue density distribution of the area to be examined;

[0057] Spatial mapping of real-time palpation pressure data was performed to obtain tissue elasticity characteristics of different regions of the site to be examined.

[0058] The data of organ contour, tissue layer boundaries, blood vessel course characteristics, tissue density distribution and tissue elasticity characteristics are fused to construct the anatomical structure feature vector of the site to be examined.

[0059] In this embodiment, during the construction of the anatomical structure feature vector, the parameters required for the formula are explicitly extracted and calculated in the following manner:

[0060] First, after 3D reconstruction of the medical imaging anatomical data, the organ contours, tissue layer boundaries, and vascular features of the examined area are extracted using image segmentation algorithms. Specifically, the extraction of tissue layer boundaries employs an edge detection method based on gray-level gradients to identify the interfaces between different tissue layers and calculate the thickness value of each tissue layer. The average thickness of all tissue layers is obtained by averaging the thickness values. The extraction of vascular features uses vascular enhancement filtering and skeleton extraction algorithms to count the number of vascular branches per unit volume, thereby obtaining the vascular distribution density. Simultaneously, based on the CT values ​​or MRI signal intensities of different tissues in medical images, combined with a tissue type database, the tissue density of the examined area is calculated. Second, after spatial mapping, the real-time palpation pressure data is correlated with the contact pressure and tissue deformation at different locations using an elasticity model to construct an elastic distribution map of the examined area. This allows for the extraction of tissue elasticity features in each region, whose numerical values ​​characterize the tissue's resistance to deformation. The extracted parameters, along with the geometric dimensions of the organ contours, together constitute the anatomical structure feature vector.

[0061] Calculate the initial operating frequency of the ultrasound probe based on the anatomical structure feature vector. ;

[0062] The above formula calculates the initial frequency by combining tissue elasticity, vascular density, and the baseline operating frequency. The lower the tissue elasticity, the higher the initial frequency. Larger sizes compensate for resolution, and denser blood vessels also contribute to higher resolution. The larger the blood vessels, the clearer the imaging of blood vessels, and the personalized frequency initialization of different tissues is achieved by superimposing the base frequency;

[0063] Initial focus depth ;

[0064] The above formula calculates the initial value based on the average tissue thickness, tissue density, and basic focusing depth. The thicker the tissue... Larger sizes are needed to accommodate longer transmission distances, and higher densities are also required. The larger the scale, the better to compensate for differences in transmission, and combined with the basic depth, the more accurately the focus falls on the area to be inspected;

[0065] In the formula: For the influence coefficients of tissue elasticity and blood vessel distribution density on frequency; The elasticity of the tissue in the area to be examined is expressed in kPa and is calculated from real-time palpation pressure data obtained by the ultrasound robot. The density of blood vessels in the area to be examined is expressed in units of vessels / mm². 2 ; The basic operating frequency corresponding to the part to be inspected; These are the influence coefficients of average tissue thickness and tissue density on depth of focus. This represents the average thickness of tissue layers in the anatomical feature vector; The tissue density of the area to be examined is expressed in g / cm³. 3 ; The baseline focusing depth corresponding to the area to be inspected;

[0066] The initial operating frequency was calculated. With initial focus depth Then, check whether it is within the range of the ultrasonic probe's hardware operating parameters. If it is outside the range, correct it according to the probe's maximum / minimum parameter limits to finally determine the initial operating parameters used for scanning.

[0067] in, The value range is 5~50MHz·kPa. The value range is 0.05~0.3MHz·mm 2 / indivual, The value range is 0.1~0.6 mm / mm. The value range is 2~15 mm·cm 3 / g, the lower the tissue elasticity The higher the value, the denser the blood vessel distribution. The larger the value, the greater the influence of the average tissue thickness of the area to be examined on the depth of focus. The higher the value, the higher the tissue density. The larger the value;

[0068] The ultrasonic probe of the ultrasonic robot is controlled to perform scanning operations at the initial working frequency and initial focusing depth, and ultrasonic images are acquired in real time and the quality quantification index of the images is extracted. The current probe posture data of the ultrasonic robot during the scanning process is also acquired simultaneously.

[0069] The extraction logic for ultrasound image quality quantification indicators is as follows:

[0070] Signal-to-noise ratio, contrast, and edge sharpness are used as quantitative indicators of ultrasound image quality, and a comprehensive quality index is constructed based on this. ;

[0071] In the formula: , , These are the weighting coefficients for signal-to-noise ratio, contrast, and edge sharpness, and their sum is 1. Signal-to-noise ratio; For contrast; For edge sharpness, , This represents the total number of pixels at the edge of the target anatomical structure in the ultrasound image. This represents the grayscale gradient value of the i-th edge pixel calculated by the edge detection operator;

[0072] The above formula integrates signal-to-noise ratio, contrast ratio, and edge sharpness, and adapts the weighting coefficients to the inspection purpose, thereby achieving multi-dimensional and targeted quality quantification.

[0073] in , , All values ​​are greater than zero, and when the tissue noise interference in the area to be examined is strong, noise reduction should be prioritized to ensure the basic image recognizability. The larger the value, the lower the value. The smaller the value, the better when the inspection objective is to clearly distinguish the target organization from the surrounding background and to enhance the grayscale difference between the target and the background. The larger the value, the lower the value. The smaller the value, the better when the examination requires focusing on the morphological integrity of the organ and the regularity of its structural outline. The larger the value, the lower the value. The smaller the value;

[0074] , , The value of or follows:

[0075] The diagnostic demand coefficient F for the area to be examined is preset according to the purpose of the examination;

[0076] The purposes of the examinations include structural imaging, blood flow imaging, and lesion screening, and:

[0077] When the purpose of the examination is structural imaging, the diagnostic requirements for edge sharpness are higher, and the settings are... ;

[0078] When the purpose of the examination is blood flow imaging, the diagnostic requirements for signal-to-noise ratio are higher, and the setting... ;

[0079] When the purpose of the examination is lesion screening, the diagnostic requirement for contrast is higher, and the setting... ;

[0080] but , This represents the i-th weight coefficient. ∈{ , , };

[0081] Based on the anatomical features of the area to be examined, a dynamic quality threshold matching the anatomical features is set. The real-time extracted quality quantification index is compared with the dynamic quality threshold to determine whether the current ultrasound image quality is qualified.

[0082] When setting a dynamic quality threshold that matches the anatomical features of the area to be examined, the following applies:

[0083] Key anatomical parameters strongly correlated with ultrasound imaging quality are extracted from the anatomical structure feature vector, including tissue density, blood vessel distribution density, and organ thickness of the area to be examined.

[0084] A calculation model for dynamic quality thresholds is constructed based on key anatomical parameters. The model expression is as follows:

[0085] ;

[0086] In the formula: For dynamic quality thresholds; The threshold influence coefficients corresponding to tissue density, blood vessel distribution density, and organ thickness; The tissue density of the area to be examined; Blood vessel distribution density; For organ thickness; The baseline quality threshold corresponding to the area to be inspected;

[0087] The above formula combines tissue density, blood vessel density, organ thickness and basic threshold to calculate dynamic standards. The higher the density, the larger a is; the denser the blood vessels, the larger b is; the thicker the organ, the larger c is, so that the quality threshold can be adapted to different tissue characteristics.

[0088] in, All are greater than zero. The value of 'a' is larger when the tissue density of the area to be examined is higher (e.g., dense connective tissue around bones), and smaller when the tissue density is lower (e.g., superficial adipose tissue). The value of 'b' is larger when the blood vessel density of the area to be examined is higher (e.g., highly vascularized organs such as the liver and kidneys), and smaller when the blood vessel density is lower (e.g., skeletal muscle and tendon tissue). The value of 'c' is larger when the organ thickness of the area to be examined is thicker (e.g., the ventricular wall of the heart and the body of the pancreas), and smaller when the organ thickness is thinner (e.g., the thyroid gland and the mammary gland).

[0089] If the current ultrasound image quality is determined to be unqualified, the working frequency, focusing depth, and scanning angle of the ultrasound robot are adjusted in a coordinated manner based on the deviation between the quality quantification index and the dynamic quality threshold, combined with the acquired current probe posture data.

[0090] When determining that the current ultrasound image quality is unacceptable, the deviation between the quality quantification index and the dynamic quality threshold is calculated. This deviation is then combined with the acquired current probe posture data to collaboratively adjust the ultrasound probe's operating frequency, depth of focus, and the ultrasound robot's probe scanning angle. ;

[0091] based on Based on the current probe orientation data, construct a collaborative adjustment model for the ultrasonic probe's operating parameters:

[0092] ;

[0093] In the formula: This is the adjustment amount for the operating frequency; This is the operating frequency adjustment factor; The probe's elevation angle; To focus on the depth of adjustment; To focus on depth adjustment coefficient; The contact pressure between the probe and the area to be inspected; This is the standard contact pressure between the probe and the area to be inspected. This is the amount of adjustment for the probe scanning angle; This is the scanning angle adjustment factor; The roll angle of the probe;

[0094] in, , , The range of values ​​is defined by the user and follows the following rules:

[0095] When the area to be examined is superficial tissue and the diagnosis requires high resolution. The larger the value, the more likely the area to be examined is deep tissue. The smaller the value, the smaller the pressure fluctuation between the probe and the area to be examined, and the greater the elasticity of the tissue in the area to be examined. The larger the value, the greater the fluctuation in contact pressure and the smaller the tissue elasticity. The smaller the value, the lower the anatomical complexity of the area to be examined and the larger the initial roll angle deviation of the probe. The larger the value, the higher the complexity of the anatomical structure. The smaller the value;

[0096] During the scanning process, the ultrasonic robot acquires the current probe posture data by using a preset six-dimensional force sensor at the end of the ultrasonic robot to collect the contact pressure data between the probe and the area to be inspected, and by using the robot joint encoder to collect the probe's position coordinates (X, Y, Z) and attitude angles, including pitch angle, roll angle, and yaw angle.

[0097] The acquired position coordinates and attitude angles are simultaneously filtered in real time; the contact pressure data is smoothed in the time domain.

[0098] Using the robot system's reference clock as a time reference, the filtered position coordinates, attitude angle data and smoothed contact pressure data are timestamped. At the same time, the contact pressure data is mapped to the spatial coordinate system of the probe end as a spatial reference, thus completing the spatiotemporal alignment and obtaining the complete attitude dataset of the current probe.

[0099] Among them, the algorithms used for filtering position coordinates and attitude angles and smoothing time thresholds of stress relief data are any of the existing algorithms adapted to the current scenario.

[0100] The ultrasound robot is controlled to perform a second scan according to the adjusted parameters. The ultrasound images of the second scan are acquired and the quality quantification index is re-extracted. The newly extracted quality quantification index is compared with the dynamic quality threshold again to verify whether the image quality is qualified.

[0101] If the image quality is still not up to standard after the second scan, the parameter adjustment and scanning verification process will be repeated until the image quality is up to standard. If the image quality is still not up to standard after the preset number of scans, the process will jump to the anatomical structure feature information acquisition stage of the area to be examined and reset.

[0102] Preset scan count adaptation , Indicates the preset number of scans. Indicates the frequency adjustment factor. This represents the initial deviation between the quality quantification index and the dynamic quality threshold during the first scan. This represents the complexity coefficient of the anatomical features of the area to be examined. Indicates the number of baseline scans. Customized by the user. ≥3;

[0103] The above formula, combining initial quality deviation, anatomical complexity, and baseline scan count (≥3), determines the maximum number of scans. Adapt deviations and complexities as needed, avoid insufficient or excessive attempts, stop when the target is met, and restart the acquisition of anatomical features if the target is not met, thereby forming a control closed-loop optimization;

[0104] In addition, it should be explained When the result of the calculation is not an integer, round it up.

[0105] in, The initial value range is 1 to 10. The larger the initial deviation value or the smaller the anatomical feature complexity coefficient, the more likely the value will change. The larger the value, the smaller the initial deviation value or the larger the anatomical feature complexity coefficient. The smaller the value; The initial value range is (0.5, 2.0). When the area to be examined has overlapping organs, densely interwoven blood vessels, or blurred tissue layers, the value range is adjusted accordingly. The higher the value, the better when the area to be examined is a single organ, has sparse blood vessels, or has clearly defined tissue layers. The smaller the value;

[0106] When the preset number of scans is reached and the image quality is still unsatisfactory, a second acquisition of anatomical structure feature information is triggered:

[0107] Multimodal anatomical data of the area to be examined are reacquired to correct the initially constructed anatomical feature vector. For example, the anatomical position shift caused by respiratory movements and changes in body position is supplemented. Based on the corrected anatomical feature vector, the initial working frequency and initial focusing depth of the ultrasound probe are reset to restart the ultrasound robot scanning process.

[0108] After coordinating the adjustment of the ultrasonic probe's operating frequency, depth of focus, and the ultrasonic robot's probe scanning angle, the propagation time of the ultrasonic beam in the tissue to be examined is calculated synchronously based on the adjusted operating frequency and depth of focus before performing the second scan. ,in Indicates the adjusted focus depth. This indicates the speed at which ultrasound propagates through the tissue being examined.

[0109] Further based on the time of transmission Adjust the probe scanning trigger timing of the ultrasonic robot to synchronize the transmission and reception of ultrasonic signals.

[0110] In this embodiment, under the ultrasound robot examination scenario, the initial parameters of the probe can be accurately set according to the anatomical characteristics of the area to be examined, the image quality can be judged in real time, and the working parameters and scanning angle can be flexibly adjusted when the quality is unsatisfactory. The process can be repeatedly verified until the quality is satisfactory, and anatomical position deviations can also be corrected. This can effectively improve the clarity and accuracy of ultrasound images, reduce repetitive operations, improve examination efficiency, adapt to the needs of different areas, ensure a stable and smooth examination process, and provide reliable image support for subsequent diagnosis.

[0111] Referring to the methods in the above embodiments, the following are two application examples of the methods described in the above embodiments:

[0112] Example 1:

[0113] When screening patients for liver lesions, the first step is to acquire anatomical structural features of the area to be examined: CT medical imaging data of the liver and real-time palpation pressure data from an ultrasound robot are collected. Three-dimensional reconstruction is performed on the CT data to extract the liver organ contour, liver tissue layer boundaries, and intrahepatic blood vessel course characteristics. Spatial mapping is performed on the palpation pressure data to obtain the tissue elasticity characteristics of different regions of the liver (the final liver tissue elasticity is determined to be 15 kPa). The above organ contour, tissue layer boundaries, blood vessel course, and tissue elasticity characteristics are fused to construct a liver anatomical structural feature vector.

[0114] Based on this feature vector, the initial operating parameters of the ultrasound probe were set as follows: the influence coefficient of tissue elasticity on frequency was set to 30 MHz·kPa, and the influence coefficient of blood vessel distribution density on frequency was set to 0.2 MHz·mm. 2 / (liver blood vessel density is 0.8 vessels / mm) 2 The baseline operating frequency for the liver is 5MHz, and the calculated initial operating frequency of the ultrasound probe is 7.2MHz. The influence coefficient of average tissue thickness on the focusing depth is taken as 0.2mm / mm (the average tissue thickness of the liver is 8mm), and the influence coefficient of tissue density on the focusing depth is taken as 2mm·cm. 3 / g (liver tissue density is 1.05g / cm³) 3 The baseline focusing depth for the liver is 6 mm, and the calculated initial focusing depth is 9.5 mm. Verification confirmed that both the initial operating frequency and initial focusing depth are within the range of the ultrasound probe's hardware operating parameters (frequency 2-12 MHz, depth 2-15 mm), thus confirming this set of parameters as the initial scanning operating parameters.

[0115] The ultrasound probe of the controlled ultrasound robot performs scanning at a working frequency of 7.2MHz and a focusing depth of 9.5mm, acquiring liver ultrasound images in real time and extracting quality quantification indicators. Simultaneously, probe posture data is acquired: the contact pressure between the probe and the liver is acquired through a six-dimensional force sensor at the robot's end effector (current contact pressure is 12N), and the probe position coordinates and posture angles (pitch angle 5°, roll angle 3°) are acquired through a joint encoder. The position coordinates and posture angles are filtered in real time, the contact pressure data is smoothed in the time domain, and the processed data is spatiotemporally aligned to form a complete current probe posture dataset.

[0116] Since the purpose of this examination is to screen for liver lesions, it is necessary to prioritize enhancing the grayscale difference between the target tissue and the background. The diagnostic requirement coefficients are set as follows: signal-to-noise ratio requirement coefficient is 2, contrast requirement coefficient is 4, and edge sharpness requirement coefficient is 1. According to the weighting calculation rules, the signal-to-noise ratio weight is 0.29, the contrast weight is 0.57, and the edge sharpness weight is 0.14. Based on this, the overall quality index of the initial ultrasound scan is calculated to be 65.

[0117] A dynamic quality threshold was set based on the anatomical characteristics of the liver: extracted tissue density (1.05 g / cm³). 3 ), blood vessel distribution density (0.8 vessels / mm) 2 Organ thickness (10mm) was used as a key anatomical parameter; the threshold influence coefficients for tissue density, blood vessel distribution density, and organ thickness were set to 5, 5, and 2, respectively. The baseline quality threshold for the liver was 60, and the dynamic quality threshold was calculated to be 89. The overall quality index (65) of the first scan was compared with the dynamic quality threshold (89), and the current ultrasound image quality was determined to be unqualified, with a quality deviation of 24.

[0118] Based on the ΔQ and current probe posture data, the following parameters were adjusted collaboratively: The operating frequency adjustment coefficient was set to 0.02, resulting in an adjustment of approximately 0.47 MHz, and the adjusted operating frequency was 7.67 MHz; the focusing depth adjustment coefficient was set to 0.03, with a standard contact pressure of 10 N between the probe and the liver, resulting in an adjustment of approximately 0.72 mm, and the adjusted focusing depth was 10.22 mm; the scanning angle adjustment coefficient was set to 0.01, resulting in an adjustment of approximately 0.012°, and the probe scanning angle was synchronously corrected after adjustment. Simultaneously, based on the adjusted focusing depth of 10.22 mm and the ultrasound propagation velocity in liver tissue (1570 m / s), the propagation time of the ultrasound beam in the liver tissue was calculated to be approximately 1.3 × 10⁻⁶ m / s. -5 The ultrasound robot's probe scanning trigger timing is adjusted based on the specified time (seconds) to ensure that ultrasound signal transmission and reception are synchronized.

[0119] The ultrasound robot was controlled to perform a second scan according to the adjusted parameters. Images from the second scan were acquired, and quality quantification indicators were re-extracted. The calculated overall quality index for the second scan was 82, still less than the dynamic quality threshold of 89, thus the quality was deemed unacceptable. The parameter adjustment and scan verification process was repeated. Before the third scan, the adjustment coefficients were fine-tuned. The final overall quality index for the third scan was 91, greater than the dynamic quality threshold of 89, thus the ultrasound image quality was deemed acceptable, and the ultrasound robot examination for liver lesion screening was completed.

[0120] Example 2:

[0121] When performing breast ultrasound examinations on patients, the first step is to acquire anatomical structural features of the area to be examined: mammogram anatomical data and real-time palpation pressure data from the ultrasound robot are collected. Three-dimensional reconstruction is performed on the mammogram data to extract the organ contours, tissue layer boundaries, and vascular course features of the breast. Spatial mapping is performed on the palpation pressure data to obtain the tissue elasticity features of different regions of the breast (the tissue elasticity was ultimately determined to be 20 kPa). All of the above feature data are fused to construct the anatomical structural feature vector of the breast.

[0122] Based on this feature vector, the initial operating parameters of the ultrasound probe were set: combined with the elasticity of breast tissue (20 kPa) and the blood vessel density (0.08 vessels / mm²). 2 The baseline operating frequency for the breast is 5 MHz, and the influence coefficients of tissue elasticity on frequency are 10 MHz·kPa and vascular distribution density on frequency are 0.1 MHz·mm. 2 / each, the initial operating frequency was calculated to be 5.51MHz; combined with the average thickness of breast tissue layers of 8mm and tissue density of 1.05g / cm³. 3 The baseline focusing depth for the breast is 10 mm, and the influence coefficients of average tissue thickness on focusing depth are 0.3 mm / mm and tissue density on focusing depth are 8 mm·cm. 3 The initial focusing depth was calculated to be 20.8 mm. This initial parameter was verified to be within the operating range of the ultrasound probe hardware and was therefore determined as the initial operating parameter for the scan.

[0123] The ultrasound probe of the controlled ultrasound robot performs a breast scan according to initial parameters, acquiring ultrasound images in real time and extracting quality quantification indicators. Simultaneously, probe posture data is acquired. Since the purpose of this examination is breast lesion screening, image contrast must be prioritized. The diagnostic requirement coefficients for signal-to-noise ratio (SNR), contrast, and edge sharpness are set to 2, 3, and 1, respectively, with calculated weighting coefficients of 0.33, 0.5, and 0.17. The measured SNR, contrast, and edge sharpness of the current ultrasound image are 25, 1.8, and 0.6, respectively, resulting in a calculated comprehensive quality index of 9.25. Furthermore, based on a breast tissue density of 1.05 g / cm³...3 Blood vessel density: 0.08 vessels / mm² 2 With an organ thickness of 8 mm and a baseline quality threshold of 10 for the breast, the influence coefficients for tissue density, vascular density, and organ thickness were selected as 0.5, 0.3, and 0.4 respectively, resulting in a calculated dynamic quality threshold of 13.75. Comparison revealed that the overall quality index of 9.25 was lower than the dynamic quality threshold of 13.75, indicating that the current ultrasound image quality was substandard.

[0124] The deviation between the calculated quality quantification index and the dynamic quality threshold was 4.5. Combined with real-time probe attitude data (tilt angle 15°, probe-breast contact pressure 2.5N, standard contact pressure 2N, roll angle 10°), the following parameters were adjusted: Due to the superficial nature of breast tissue and the high resolution requirements for diagnosis, a working frequency adjustment coefficient of 3 was selected; the small fluctuation in probe-breast contact pressure led to a focusing depth adjustment coefficient of 2; and the moderate complexity of breast anatomy resulted in a scanning angle adjustment coefficient of 1.5. The calculated working frequency adjustment was 13.04MHz, resulting in an adjusted working frequency of 18.55MHz; the focusing depth adjustment was 11.25mm, resulting in an adjusted focusing depth of 32.05mm; and the scanning angle adjustment was 1.17°, thus determining the adjusted scanning angle. Meanwhile, based on the adjusted focusing depth of 32.05 mm and the ultrasound propagation speed in breast tissue of 1540 m / s, the propagation time of the ultrasound beam in breast tissue was calculated to be 41.6 μs. Accordingly, the probe scanning triggering sequence of the ultrasound robot was adjusted to ensure that the transmission and reception of ultrasound signals were synchronized.

[0125] The ultrasound robot is controlled to perform a second scan according to the adjusted parameters. Ultrasound images are reacquired, and the overall quality index is calculated to be 14.2. This index is higher than the dynamic quality threshold of 13.75, indicating that the image quality is acceptable, and the breast ultrasound examination is completed. If the image quality of the second scan is still unacceptable, the parameter adjustment and scan verification process is repeated. The preset number of scans for this examination is 26. If the image quality is still unacceptable after 26 scans, multimodal breast anatomical data is reacquired (e.g., supplementing features indicating breast positional shifts due to respiratory movements), the initially constructed anatomical feature vector is corrected, and the initial operating frequency and focusing depth of the ultrasound probe are reset based on the corrected vector. The ultrasound robot's scanning process is then restarted.

[0126] In summary, the methods described in the above embodiments, during execution, improve the initial scanning targeting by acquiring the anatomical structural features of the area to be examined and setting the initial working parameters of the ultrasound probe, thereby reducing invalid operations. Real-time acquisition of ultrasound image quality quantification indicators and probe posture data allows for timely monitoring of the scanning status. Combining anatomical features to set dynamic quality thresholds makes image quality judgment more consistent with the actual anatomical situation, avoiding the judgment deviation of uniform thresholds. When the image quality is unqualified, the working frequency, focusing depth, and scanning angle are adjusted collaboratively based on the deviation value and probe posture to accurately optimize the scanning parameters. The secondary scanning verification and repeated adjustment process can continuously improve image quality. If the preset number of times fails to meet the requirements, the acquisition of anatomical features and the initial parameter settings are reset to address quality problems caused by anatomical changes. The scanning trigger sequence is adjusted synchronously to ensure ultrasound signal synchronization, thereby improving the overall quality and efficiency of ultrasound examination and reducing reliance on manual intervention.

[0127] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive optimization method for the quality of ultrasound robot examination, characterized in that, include: Obtain anatomical structural feature information of the area to be examined, and set the initial working frequency and initial focusing depth of the ultrasound probe of the ultrasound robot based on the anatomical feature information; The ultrasonic probe of the ultrasonic robot is controlled to perform scanning operations at the initial working frequency and initial focusing depth, acquire ultrasonic images in real time and extract the quality quantification indicators of the images, and simultaneously acquire the current probe posture data of the ultrasonic robot during the scanning process; when acquiring the current probe posture data of the ultrasonic robot during the scanning process, the contact pressure data between the probe and the part to be examined is collected by a preset six-dimensional force sensor at the end of the ultrasonic robot, and the position coordinates (X,Y,Z) and attitude angles of the probe are collected by the robot joint encoder, wherein the attitude angles include pitch angle, roll angle and yaw angle; The acquired position coordinates and attitude angles are simultaneously filtered in real time; the contact pressure data is smoothed in the time domain. Using the robot system's reference clock as a time reference, the filtered position coordinates and attitude angle data are timestamped with the smoothed contact pressure data. Simultaneously, using the center point of the probe's end effector as a spatial reference, the contact pressure data is mapped to the spatial coordinate system of this reference, completing spatiotemporal alignment and obtaining the complete attitude dataset of the current probe. Based on the anatomical features of the area to be examined, a dynamic quality threshold matching the anatomical features is set. The real-time extracted quality quantification index is compared with the dynamic quality threshold to determine whether the current ultrasound image quality is qualified. When determining that the current ultrasound image quality is unacceptable, the deviation between the quality quantification index and the dynamic quality threshold is calculated. This deviation is then combined with the acquired current probe posture data to collaboratively adjust the ultrasound probe's operating frequency, depth of focus, and the ultrasound robot's probe scanning angle. ; For dynamic quality thresholds, It is a comprehensive quality index; based on Based on the current probe orientation data, construct a collaborative adjustment model for the ultrasonic probe's operating parameters: ； In the formula: This is the adjustment amount for the operating frequency; This is the operating frequency adjustment factor; The probe's elevation angle; To focus on the depth of adjustment; To focus on depth adjustment coefficient; The contact pressure between the probe and the area to be inspected; This is the standard contact pressure between the probe and the area to be inspected. This is the amount of adjustment for the probe scanning angle; This is the scanning angle adjustment factor; The roll angle of the probe; If the current ultrasound image quality is determined to be unqualified, the working frequency, focusing depth, and scanning angle of the ultrasound robot are adjusted in a coordinated manner based on the deviation between the quality quantification index and the dynamic quality threshold, combined with the acquired current probe posture data. The ultrasound robot is controlled to perform a second scan according to the adjusted parameters. The ultrasound images of the second scan are acquired and the quality quantification index is re-extracted. The newly extracted quality quantification index is compared with the dynamic quality threshold again to verify whether the image quality is qualified. If the image quality is still not up to standard after the second scan, the parameter adjustment and scanning verification process will be repeated until the image quality is up to standard. If the image quality is still not up to standard after the preset number of scans, the process will jump to the stage of obtaining anatomical structural feature information of the area to be examined and reset the execution.

2. The adaptive optimization method for ultrasound robot examination quality according to claim 1, characterized in that, The operation of acquiring anatomical structural feature information of the area to be examined includes: Collect medical imaging anatomical data of the area to be examined and real-time palpation pressure data of the ultrasound robot; Three-dimensional reconstruction of medical imaging anatomical data is performed to extract organ contours, tissue layer boundaries, blood vessel course characteristics, and tissue density distribution of the area to be examined. Spatial mapping of real-time palpation pressure data was performed to obtain tissue elasticity characteristics of different regions of the site to be examined. The data of organ contour, tissue layer boundaries, blood vessel course characteristics, tissue density distribution and tissue elasticity characteristics are fused to construct the anatomical structure feature vector of the site to be examined. Calculate the initial operating frequency of the ultrasound probe based on the anatomical structure feature vector. ; and initial focus depth ; In the formula: For the influence coefficients of tissue elasticity and blood vessel distribution density on frequency; The elasticity of the tissue at the site to be examined is expressed in kPa. The density of blood vessels in the area to be examined is expressed in units of vessels / mm². The basic operating frequency corresponding to the part to be inspected; These are the influence coefficients of average tissue thickness and tissue density on depth of focus. This represents the average thickness of tissue layers in the anatomical feature vector; The tissue density of the area to be examined is expressed in g / cm³. The baseline focusing depth corresponding to the area to be inspected; The initial operating frequency was calculated. With initial focus depth Then, it is verified whether it is within the range of the ultrasonic probe's hardware operating parameters. If it is outside the range, it is corrected according to the probe's maximum / minimum parameter limits to finally determine the initial operating parameters used for scanning.

3. The adaptive optimization method for ultrasound robot examination quality according to claim 1, characterized in that, The extraction logic for the ultrasound image quality quantification index is as follows: Signal-to-noise ratio, contrast, and edge sharpness are used as quantitative indicators of ultrasound image quality, and a comprehensive quality index is constructed based on this. ; In the formula: , , These are the weighting coefficients for signal-to-noise ratio, contrast, and edge sharpness, and their sum is 1. Signal-to-noise ratio; For contrast; For edge sharpness, the calculation formula is: ,in This represents the total number of pixels at the edge of the target anatomical structure in the ultrasound image. This represents the grayscale gradient value of the i-th edge pixel calculated by the edge detection operator.

4. The adaptive optimization method for ultrasound robot examination quality according to claim 3, characterized in that, The , , The value of follows: The diagnostic demand coefficient F for the area to be examined is preset according to the purpose of the examination; The purposes of the examinations include structural imaging, blood flow imaging, and lesion screening, and: When the purpose of the inspection is structural imaging, set ; When the purpose of the examination is blood flow imaging, set ; When the purpose of the examination is lesion screening, set ; but , This represents the i-th weight coefficient. ∈{ , , } 5. The adaptive optimization method for ultrasound robot examination quality according to claim 1, characterized in that, When setting a dynamic quality threshold that matches the anatomical features of the area to be examined, the following applies: Key anatomical parameters strongly correlated with ultrasound imaging quality are extracted from the anatomical structure feature vector, including tissue density, blood vessel distribution density, and organ thickness of the area to be examined. A calculation model for dynamic quality thresholds is constructed based on key anatomical parameters. The model expression is as follows: ; In the formula: For dynamic quality thresholds; The threshold influence coefficients corresponding to tissue density, blood vessel distribution density, and organ thickness; The tissue density of the area to be examined; Blood vessel distribution density; For organ thickness; This represents the baseline quality threshold corresponding to the area to be inspected.

6. The adaptive optimization method for ultrasound robot examination quality according to claim 1, characterized in that, The preset number of scans is adapted to , Indicates the preset number of scans. Indicates the frequency adjustment factor. This represents the initial deviation between the quality quantification index and the dynamic quality threshold during the first scan. This represents the complexity coefficient of the anatomical features of the area to be examined. Indicates the number of baseline scans. Customized by the user. ≥3; When the preset number of scans is reached and the image quality is still unsatisfactory, a second acquisition of anatomical structure feature information is triggered: Multimodal anatomical data of the area to be examined are reacquired to correct the initially constructed anatomical feature vector. Based on the corrected anatomical feature vector, the initial operating frequency and initial focusing depth of the ultrasound probe are reset to restart the ultrasound robot scanning process.

7. The adaptive optimization method for ultrasound robot inspection quality according to claim 1, characterized in that, After coordinating the adjustment of the ultrasonic probe's operating frequency, depth of focus, and the ultrasonic robot's probe scanning angle, the propagation time of the ultrasonic beam in the tissue to be examined is calculated synchronously based on the adjusted operating frequency and depth of focus before performing the second scan. ,in Indicates the adjusted focus depth. This indicates the speed at which ultrasound propagates through the tissue being examined. Based on the time of transmission Adjust the probe scanning trigger timing of the ultrasonic robot to synchronize the transmission and reception of ultrasonic signals.

Images