A transperineal pelvic floor color ultrasound examination system and method of use thereof

Abstract

The application discloses a transperineal pelvic floor color ultrasound examination system and a use method thereof, relates to the technical field of color ultrasound examination, and the image optimization module is used for automatically adjusting gain, contrast, filtering and tissue harmonic imaging mode, the three-dimensional imaging module is used for capturing the change of the pelvic floor muscle in the resting and moving states, the data analysis module is combined with the optimized ultrasonic image data and muscle dynamic information, intelligently evaluates the image quality, automatically calculates the bladder neck activity, the urethral rotation angle and the pelvic floor muscle thickness in the high-quality image, and provides an abnormal annotation prompt; the user interaction module realizes the visual display of the ultrasonic image and the analysis result, supports the doctor in inputting patient information, improves operation convenience and diagnostic accuracy; the overall system optimizes the visualization ability of the anatomical structure, reduces the interference of artifacts, improves the evaluation accuracy of the pelvic floor dysfunction, provides more reliable data support for the personalized treatment scheme, and finally improves the clinical diagnosis and treatment efficiency.

Description

Technical Field

[0001] This invention relates to the field of color Doppler ultrasound examination technology, specifically to a transperineal pelvic floor color Doppler ultrasound examination system and its usage method. Background Technology

[0002] The transperineal pelvic floor ultrasound system is a system that uses transperineal ultrasound technology to visualize the pelvic floor structures. This system typically includes an ultrasound probe, imaging equipment, and analysis software. It is used to assess the anatomical structure and functional status of the pelvic floor muscles, urethra, bladder, rectum, and related tissues, playing a particularly important role in the diagnosis and treatment evaluation of conditions such as postpartum recovery, stress urinary incontinence, and pelvic organ prolapse. The procedure usually involves the patient assuming an appropriate examination position, the operator placing the ultrasound probe in the perineum, acquiring dynamic data of the pelvic floor tissues through real-time imaging, and combining the results with software analysis to assist in clinical decision-making.

[0003] In terms of existing technologies, traditional methods for assessing pelvic floor function mainly include digital rectal examination, pelvic floor electromyography (EMG), and magnetic resonance imaging (MRI). However, these methods each have their shortcomings in terms of resolution, real-time dynamic monitoring, and ease of operation. For example, digital rectal examination relies on the doctor's subjective judgment, EMG can quantify muscle function but lacks intuitive images, and while MRI can provide high-resolution anatomical images, it is costly and difficult to perform dynamic assessments. In contrast, transperineal pelvic floor ultrasound can provide high-definition, real-time dynamic images at a non-invasive and low-cost level, and therefore has been widely used and developed in clinical practice.

[0004] The existing technology has the following shortcomings: Ultrasound imaging relies on the propagation and reflection of sound waves between different tissues. However, the pelvic floor region contains various tissues (such as muscles, ligaments, connective tissue, and fat), and interfaces with different densities and acoustic impedances can produce artifacts, affecting image quality. For example, gas in the bladder and rectum may cause sound attenuation or reflection, making some anatomical structures difficult to distinguish, thus interfering with the assessment of key areas such as the urethra, bladder neck, and levator ani muscle. Furthermore, if artifacts lead to misdiagnosis, it may cause clinicians to misjudge the degree of pelvic floor dysfunction (such as urinary incontinence or organ prolapse), affecting the formulation of treatment plans. Summary of the Invention

[0005] The purpose of this invention is to provide a transperineal pelvic floor color Doppler ultrasound examination system and its usage method to address the shortcomings of the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a transperineal pelvic floor color Doppler ultrasound examination system, comprising an ultrasound probe module, an imaging module, an image optimization module, a three-dimensional imaging module, a data analysis module, and a user interaction module; The ultrasound probe module is used to transmit and receive ultrasound signals and acquire echo signals from the pelvic floor tissues. The imaging module, connected to the ultrasound probe module, is used to receive echo signals and generate real-time ultrasound images; The image optimization module is used to automatically adjust ultrasound imaging parameters, including gain, contrast, filtering, and tissue harmonic imaging mode, to reduce artifacts caused by the complex structure of the pelvic floor tissue. The 3D imaging module is used for 3D reconstruction of pelvic floor tissues, capturing data on changes in pelvic floor muscles in static and dynamic states through 4D imaging technology. The data analysis module is used to combine ultrasound image data with artifact reduction and pelvic floor muscle change data to evaluate the quality of real-time ultrasound images. For high-quality real-time ultrasound images, it automatically calculates key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, and provides abnormal annotation prompts. The user interaction module displays real-time ultrasound images and the analysis results from the data analysis module, and supports doctors in inputting patient information.

[0007] Preferably, in the data analysis module, a tissue boundary stability index is generated by measuring whether the pelvic floor tissue boundaries are clear and stable in ultrasound images. The method for obtaining the tissue boundary stability index is as follows: Preprocess ultrasound images; if the input image is color, convert it to grayscale. The image is smoothed using a Gaussian kernel, and the expression is calculated as follows: ;in: It is the grayscale value of the original ultrasound image. It is a two-dimensional Gaussian filter kernel, and * denotes the convolution operation. It is the standard deviation of the Gaussian filter. This is a smoothed grayscale image; Calculate the horizontal direction and vertical direction The gradient is obtained using a Sobel filter: , ; Calculate the gradient magnitude: ; Calculate the gradient direction: ; Along the gradient direction To refine the process, only local maxima are retained. High thresholds Th and Tl are set to filter edges. Mark as a strong edge point; if Mark a weak edge point as a weak edge point. If it is connected to a strong edge point, keep it; otherwise, discard it. If so, discard it; Calculate the local contrast of pixels within the boundary region The expression is: Where: N is the number of boundary points (|E|). It is the average gray value of all boundary points; calculate the gradient direction consistency value of the boundary points. E represents the set of edge pixels in the image, expressed as: ;in: It is the average gradient direction at all boundary points; The tissue boundary stability index is obtained by weighted average summation of the boundary contrast change rate and gradient consistency value.

[0008] Preferably, in the data analysis module, a dynamic deformation index of the pelvic floor muscles is generated by measuring the consistency and repeatability of pelvic floor muscle deformation under different physiological states. The method for obtaining the dynamic deformation index of the pelvic floor muscles is as follows: Preprocess ultrasound image sequences; if the ultrasound images are in RGB format, convert them to grayscale images. Calculate the horizontal direction and vertical direction The gradient is obtained using a Sobel filter: , ; Calculate the time gradient: ; The fundamental equation for optical flow is as follows: (This is used to calculate the velocity of each pixel.) Where: u is the displacement velocity in the x-direction, and v is the displacement velocity in the y-direction. , It is the spatial gradient. It is the time gradient; If we assume that the velocity of a local pixel block remains constant within the window size W, then we have: ; Solve for u and v to obtain the displacement vector of each pixel; Calculate the average motion velocity of all pixels in the pelvic floor muscle region using the following expression: ;in: The dynamic deformation index (PFDC) of the pelvic floor muscles represents the local deformation velocity, where u and v are the horizontal and vertical movement velocities, respectively. The calculation expression is as follows: Where M is the total number of pixels in the pelvic floor muscle region. It is the deformation speed of each pixel.

[0009] Preferably, in the data analysis module, the tissue boundary stability index and the pelvic floor muscle dynamic deformation index are converted into a comprehensive feature vector. The comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses the prediction of the quality analysis value label of the real-time ultrasound image for each set of comprehensive feature vectors as the prediction objective, and minimizes the sum of prediction errors of the quality analysis value labels of all real-time ultrasound images as the training objective. The machine learning model is trained until the sum of prediction errors converges and the model training stops. The quality analysis value of the real-time ultrasound image is determined based on the model output. The machine learning model is a multinomial regression model.

[0010] Preferably, in the data analysis module, the quality analysis value of the acquired real-time ultrasound image is compared with a pre-set quality analysis value reference threshold based on historical data. If the quality analysis value of the real-time ultrasound image is greater than or equal to the pre-set quality analysis value reference threshold, it indicates that the quality of the acquired real-time ultrasound image is high, and it is classified as a high-quality real-time ultrasound image. For high-quality real-time ultrasound images, key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, are automatically calculated, and abnormality annotation prompts are provided. If the quality analysis value of the real-time ultrasound image is less than the pre-set quality analysis value reference threshold, it indicates that the quality of the acquired real-time ultrasound image is low, and it is classified as a low-quality real-time ultrasound image, requiring re-acquiring the image or adjusting the ultrasound parameters.

[0011] This invention also provides a method for using transperineal pelvic floor ultrasound examination, including: S1: The patient lies supine or on their side, relaxes their pelvic floor muscles, and prepares for the examination. S2: The operator applies coupling gel to the ultrasound probe and places it on the patient's perineum, then adjusts the probe angle to obtain the best imaging view. S3: The ultrasound images of the pelvic floor tissues are displayed in real time by the imaging device and enhanced by the image optimization module to reduce artifact interference. S4: Enable the 3D imaging module to acquire 3D reconstructed images of the pelvic floor muscles and analyze key anatomical structures; S5: By combining ultrasound image data with reduced artifacts and pelvic floor muscle change data, the quality of real-time ultrasound images is evaluated, pelvic floor function parameters are calculated, and abnormal indications are provided in conjunction with patient symptoms. S6: Based on ultrasound images and analysis results, doctors determine whether the patient has pelvic floor dysfunction and develop corresponding treatment or rehabilitation plans.

[0012] Preferably, in S2, the operator applies a coupling gel to the ultrasound probe, places it on the patient's perineum, and adjusts the probe angle to obtain the optimal imaging view, specifically including: Let the current angle of the ultrasound probe be... The dynamic adjustment of the probe angle is then expressed as: ;in: The current angle of the ultrasound probe. The adjusted ultrasound probe angle is represented by Q, which is the real-time ultrasound image quality analysis value calculated by the data analysis module. This represents the gradient of the image quality analysis value with respect to the probe angle, i.e., the rate of quality change at the current angle. The learning rate is used to control the adjustment range. Taking the partial derivative with respect to the probe angle, we get: ;in: , The weighting coefficient is set to... ; The influence of probe angle on tissue boundary stability is considered. If the boundary is blurred, the partial derivative of this term is small, and the probe angle needs to be adjusted. To assess the impact of probe angle on the dynamic deformation of the pelvic floor muscles, if the pelvic floor muscle contour is abnormally deformed, this partial derivative is small, and the probe angle needs to be adjusted.

[0013] Preferably, in S2, in order to achieve dynamic adjustment, an iterative optimization strategy is adopted. The iterative optimization strategy includes: setting an initial probe angle. Between 30° and 60°; calculated using current imagery. ; Calculate the rate of change of mass with respect to angle, and update the probe angle: If the change in image quality is less than the set threshold ϵ, stop adjusting; otherwise, continue iterating.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: 1. This invention achieves accurate assessment of pelvic floor function through the collaborative optimization of ultrasound probe modules, imaging modules, image optimization modules, 3D imaging modules, data analysis modules, and user interaction modules. The system employs technologies such as automatic artifact suppression, 3D / 4D dynamic imaging, intelligent image quality analysis, and parameter optimization to overcome problems in existing ultrasound examinations, such as low image quality and large diagnostic errors caused by tissue complexity, sound attenuation, and reflection. In particular, the data analysis module introduces the Tissue Boundary Stability Index (TBSI) and the Pelvic Floor Muscle Dynamic Deformation Index (PFDC), combined with a machine learning multinomial regression model, to automatically assess image quality and dynamically adjust the probe angle based on the analysis results. This ensures image clarity, improves the measurement accuracy of key anatomical parameters such as bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, and provides abnormality annotations to assist doctors in accurate diagnosis.

[0015] 2. This invention achieves automatic screening of high-quality ultrasound images through intelligent image quality optimization and dynamic probe adjustment, and ensures accurate calculation of key pelvic floor functional parameters, significantly improving the reliability and ease of operation of ultrasound examinations. Doctors can intuitively view real-time images, data analysis results, and abnormality annotation information through the user interaction module, and input patient medical history to achieve personalized diagnosis. This system not only reduces the risk of misdiagnosis due to artifacts or improper ultrasound angles, but also improves the screening and follow-up efficiency of pelvic floor diseases (such as stress urinary incontinence, pelvic organ prolapse, etc.), providing clinicians with a more accurate, intelligent, and efficient pelvic floor ultrasound examination solution. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a system module diagram of the present invention. Detailed Implementation

[0018] 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 embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] For examples, please refer to Figure 1 As shown in this embodiment, a transperineal pelvic floor color Doppler ultrasound examination system includes an ultrasound probe module, an imaging module, an image optimization module, a three-dimensional imaging module, a data analysis module, and a user interaction module. The ultrasound probe module is used to transmit and receive ultrasound signals and acquire echo signals from the pelvic floor tissues. The imaging module, connected to the ultrasound probe module, is used to receive echo signals and generate real-time ultrasound images; The image optimization module is used to automatically adjust ultrasound imaging parameters, including gain, contrast, filtering, and tissue harmonic imaging mode, to reduce artifacts caused by the complex structure of the pelvic floor tissue. The 3D imaging module is used for 3D reconstruction of pelvic floor tissues, capturing data on changes in pelvic floor muscles in static and dynamic states through 4D imaging technology. The data analysis module is used to combine ultrasound image data with artifact reduction and pelvic floor muscle change data to evaluate the quality of real-time ultrasound images. For high-quality real-time ultrasound images, it automatically calculates key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, and provides abnormal annotation prompts. The user interaction module displays real-time ultrasound images and the analysis results from the data analysis module, and supports doctors in inputting patient information.

[0020] The ultrasound probe module is the core component of a transperineal pelvic floor ultrasound examination system. It is primarily used to transmit and receive ultrasound signals and transmit the echo signals to the imaging module to generate ultrasound images of the pelvic floor tissues. This module typically includes the following key parts: Transducer array: Constructed of piezoelectric materials (such as PZT or PMN-PT), it can convert between electrical signals and ultrasonic waves. Transducer arrays can adopt linear array, convex array, or phased array structures to meet different pelvic floor examination needs.

[0021] Operating frequency: Typically in the range of 5MHz to 12MHz, which can be adjusted according to the imaging depth requirements. Higher frequencies (>8MHz) provide higher resolution and are suitable for superficial tissues (such as the urethra and bladder neck), while lower frequencies (<7MHz) can penetrate deeper tissues (such as the levator ani muscle and rectum).

[0022] Dynamic focus adjustment: The ultrasound probe supports electronic focusing and dynamic receiver focusing technology to optimize imaging clarity at different depths and reduce blurring caused by focus mismatch.

[0023] Signal transmission and noise suppression: The probe integrates a low-noise amplifier (LNA), which can improve the detection sensitivity of weak echo signals and reduce the impact of electromagnetic interference on ultrasound signals.

[0024] The imaging module is the core computing module of the transperineal pelvic floor ultrasound system. It is responsible for receiving the echo signals acquired by the ultrasound probe and converting them into real-time ultrasound images. Its main functions and technical features include: Signal reception and preprocessing: Employing multi-channel high dynamic range (HDR) signal reception, it effectively captures the contrast between strong echoes (such as bones) and weak echoes (such as soft tissue), enhancing image detail. It features adaptive gain control (TGC), which automatically adjusts the signal gain at different depths to prevent images from being too dark or overexposed.

[0025] Real-time image processing: Tissue harmonic imaging (THI): Utilizes high-order harmonic signals to improve resolution, reduce artifacts, and enhance the clarity of pelvic floor tissue boundaries. Adaptive filtering (Speckle Reduction Algorithm, SRA): Removes speckle noise, enhances image contrast, and makes the layers of pelvic floor muscles more clearly visible. Spatial composite imaging: Acquires ultrasound signals from multiple angles and fuses them into a single image, improving the visualization of key pelvic floor structures (such as urethral rotation angle and bladder neck mobility). Automatic tissue segmentation: Based on artificial intelligence (AI) algorithms, it can automatically identify pelvic floor muscles, urethra, bladder, and other tissues, and annotate key anatomical parameters such as pelvic floor muscle thickness and urethral rotation angle.

[0026] Dynamic image storage and analysis: Equipped with real-time 4D ultrasound imaging, it can record dynamic changes in the pelvic floor during rest, Valsalva maneuvers, coughing, etc., for the assessment of stress urinary incontinence and pelvic organ prolapse. A built-in image storage and playback system allows for post-examination analysis and supports DICOM format export for remote consultations or post-operative follow-up.

[0027] The image optimization module in the transperineal pelvic floor color Doppler ultrasound examination system automatically adjusts the ultrasound imaging parameters to improve image quality, reduce artifact interference caused by the complex structure of the pelvic floor tissues (such as muscles, ligaments, connective tissue, fat, etc.), and ensure clear visualization of key anatomical structures.

[0028] Because the pelvic floor region contains different types of tissues (such as hyperechoic ligaments and hypoechoic muscles), improper gain adjustment can lead to images that are too bright (high gain) or too dark (low gain), affecting the observation of key anatomical structures (such as the urethra, bladder neck, and levator ani muscle). This module can automatically adjust the gain based on the distribution of echo intensity (TGC - Time Gain Compensation) to ensure uniform brightness across tissues at different depths. In the near field (near the probe), the gain is reduced to avoid signal overexposure; in the far field (deep tissues), the gain is increased to enhance signal intensity and ensure clear visibility of deep pelvic floor structures (such as the puborectalis muscle and levator ani muscle).

[0029] Low echo contrast in pelvic floor tissues, especially at the boundaries between pelvic floor muscles, can make it difficult to distinguish different structures, such as the blurred boundary between the urethra and bladder wall. Dynamic Range Optimization (DMO) automatically adjusts the grayscale levels of the image to highlight details in low-contrast areas, improving the ability to distinguish muscles, ligaments, and connective tissues. Tissue-Specific Contrast Enhancement uses AI to identify key pelvic floor structures and automatically adjusts local contrast, making target areas clearer.

[0030] Traditional ultrasound imaging can be affected by artifacts, such as reflection artifacts caused by gas in the bladder and rectum, which can blur some tissue structures. Tissue harmonic imaging (THI) uses the second harmonic signal of ultrasound waves, rather than the fundamental wave, for imaging. This concentrates high-frequency signals in soft tissue areas, reduces near-field artifacts, and improves image clarity. This module can automatically select the optimal harmonic frequency based on the patient's pelvic floor anatomy to optimize image quality.

[0031] Pelvic floor ultrasound images are often affected by speckle noise, resulting in granular spots and reduced resolution of muscles, ligaments, and other tissues. Multi-Scale Adaptive Filtering (MAD) is employed to smooth the images, removing noise while preserving tissue boundary details. Combined with AI algorithms, region-specific noise reduction is performed on different pelvic floor tissue areas to avoid image distortion caused by over-smoothing.

[0032] The three-dimensional imaging module is an important component of the transperineal pelvic floor ultrasound examination system. It is mainly used to reconstruct the pelvic floor tissues in three dimensions and capture the changes of the pelvic floor muscles in static and dynamic states through 4D imaging technology (i.e., real-time three-dimensional ultrasound) to improve the accuracy of pelvic floor function assessment.

[0033] Traditional two-dimensional ultrasound can only provide images from a single cross section, making it difficult to fully display the three-dimensional structure of the pelvic floor muscles, such as the overall shape of the levator ani muscle and the spatial relationship between the urethra and the bladder neck. This affects doctors' accurate assessment of pelvic floor dysfunction (such as urinary incontinence and pelvic organ prolapse).

[0034] By employing 3D ultrasound imaging technology, data is acquired from multiple angles using an ultrasound probe and processed by computer to construct high-resolution 3D images of the pelvic floor. Combined with volume rendering, the visualization of different tissues (such as pelvic floor muscles, ligaments, urethra, rectum, etc.) is enhanced, enabling doctors to observe the pelvic floor anatomy from multiple perspectives.

[0035] The diagnosis of pelvic floor dysfunction relies not only on static structures but also on observing the movement of the pelvic floor muscles under different conditions. For example, the dynamic changes in the bladder neck, urethra, and rectum during Valsalva maneuvers (increasing abdominal pressure, such as coughing or straining) or sphincter contractions are key indicators for assessing urinary incontinence and pelvic organ prolapse. However, traditional two-dimensional ultrasound can only provide single-frame images, making it difficult to intuitively reflect the dynamic movement patterns of tissues.

[0036] By employing 4D ultrasound technology (real-time 3D ultrasound), continuous 3D image data can be acquired, and the movement changes of pelvic floor tissues under different physiological states can be displayed in real time. Time Sequence Analysis: Records the movement trajectory of pelvic floor muscles at different time points, quantifies the dynamic changes of the urethra, bladder neck, and levator ani muscles, and provides objective functional assessment data.

[0037] The data analysis module is used to evaluate the quality of real-time ultrasound images by combining ultrasound image data with reduced artifacts and pelvic floor muscle change data. For high-quality real-time ultrasound images, it automatically calculates key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, and provides abnormal annotation prompts.

[0038] Ultrasound signals in the pelvic floor region are significantly affected by tissue heterogeneity. Artifacts or blurred boundaries may exist in certain areas (such as around the bladder), affecting the clarity of tissue boundaries. Traditional ultrasound imaging mainly relies on the doctor's subjective judgment of whether tissue boundaries are clear. By measuring whether the boundaries of pelvic floor tissues (such as the urethra, bladder neck, and puborectalis muscle) are clear and stable in ultrasound images, a tissue boundary stability index is generated to automatically assess boundary quality. The method for obtaining the organizational boundary stability index is as follows: Preprocess ultrasound images; if the input image is color (RGB), convert it to grayscale. The image is smoothed using a Gaussian kernel to reduce speckle noise. The expression for this is: ;in: It is the grayscale value of the original ultrasound image. It is a two-dimensional Gaussian filter kernel, and * denotes the convolution operation. It is the standard deviation of the Gaussian filter (typical value) =1.0 to 2.0). This is a smoothed grayscale image.

[0039] Calculate the horizontal direction and vertical direction The gradient is obtained using a Sobel filter: , ; Calculate the gradient magnitude: ; Calculate the gradient direction: ; Along the gradient direction Refine the data, retaining only local maxima, using a dual thresholding process: set a high threshold Th and a low threshold Tl, and filter edges: if... Mark as a strong edge point; if Mark a weak edge point as a weak edge point. If it is connected to a strong edge point, keep it; otherwise, discard it. If so, discard it; Calculate the local contrast of pixels within the boundary region The expression is: Where: N is the number of boundary points (|E|). It is the average gray value of all boundary points; calculate the gradient direction consistency value of the boundary points. E represents the set of edge pixels in the image, expressed as: ;in: It is the average gradient direction at all boundary points.

[0040] The tissue boundary stability index is obtained by weighted average summation of boundary contrast change rate and gradient consistency value. The higher the tissue boundary stability index, the clearer and more stable the boundary, and the better the ultrasound image quality.

[0041] Dynamic changes in the pelvic floor muscles are an important indicator for functional assessment. However, the echo signals of the muscles are easily affected by artifacts and motion artifacts, leading to measurement errors. Traditional assessment methods rely solely on physicians observing changes in muscle morphology. By measuring the consistency and repeatability of pelvic floor muscle deformation under different physiological states (such as resting, Valsalva maneuver, coughing, etc.), a pelvic floor muscle dynamic deformation index is generated to assess the quality of real-time ultrasound images. The method for obtaining the pelvic floor muscle dynamic deformation index is as follows: Preprocess ultrasound image sequences; if the ultrasound images are in RGB format, convert them to grayscale images. (Two adjacent ultrasound images). Acquire spatial and temporal gradient information to calculate pixel motion. Calculate the horizontal direction. and vertical direction The gradient is obtained using a Sobel filter: , Calculate the temporal gradient (inter-frame pixel value change): .

[0042] Calculate the motion velocity of each pixel (the motion vector of the pelvic floor muscles), the basic equation of optical flow: Where: u is the displacement velocity in the x-direction (horizontal motion), and v is the displacement velocity in the y-direction (vertical motion). , It is the spatial gradient. It is the time gradient.

[0043] If we assume that the velocity of a local pixel block remains constant within the window size W, then we have: ; Solve for u and v to obtain the displacement vector of each pixel.

[0044] Calculate the average motion velocity of all pixels in the pelvic floor muscle region to assess the overall dynamic deformation of the muscles. The calculation expression is as follows: ;in: The dynamic deformation velocity of the pelvic floor muscles is represented by u and v, which are the horizontal and vertical movement velocities, respectively. The dynamic deformation index (PFDC) of the pelvic floor muscles is calculated using the following expression: Where M is the total number of pixels in the pelvic floor muscle region. It is the deformation speed of each pixel.

[0045] The tissue boundary stability index and pelvic floor muscle dynamic deformation index are converted into a comprehensive feature vector. The comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses the prediction of the quality analysis value label of the real-time ultrasound image for each set of comprehensive feature vectors as the prediction objective. The training objective is to minimize the sum of prediction errors of the quality analysis value labels of all real-time ultrasound images. The machine learning model is trained until the sum of prediction errors converges and the model training stops. The quality analysis value of the real-time ultrasound image is determined based on the model output. The machine learning model is a multinomial regression model.

[0046] The method for obtaining the quality analysis value of real-time ultrasound images is as follows: The corresponding function expression is obtained from the training data of the comprehensive feature vector of the trained machine learning model. In the formula, This is the model's output function, where TBSI is the tissue boundary stability index and PFDC is the pelvic floor muscle dynamic deformation index. This represents the quality analysis value of real-time ultrasound images.

[0047] The quality analysis value of the acquired real-time ultrasound image is compared with a pre-set reference threshold based on historical data. If the quality analysis value of the real-time ultrasound image is greater than or equal to the pre-set reference threshold, it indicates that the quality of the acquired real-time ultrasound image is high, and it is classified as a high-quality real-time ultrasound image. For high-quality real-time ultrasound images, key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, are automatically calculated, and abnormality annotations are provided. If the quality analysis value of the real-time ultrasound image is less than the pre-set reference threshold, it indicates that the quality of the acquired real-time ultrasound image is low, and it is classified as a low-quality real-time ultrasound image, requiring re-acquisition of the image or adjustment of ultrasound parameters.

[0048] For high-quality real-time ultrasound images, the system automatically calculates key anatomical parameters such as bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness to accurately assess pelvic floor function. Bladder neck mobility refers to the displacement of the bladder neck between resting and Valsalva maneuvers (such as coughing or straining), and can be used to assess the risk of stress urinary incontinence; urethral rotation angle represents the change in urethral tilt under different physiological states and is a key indicator for assessing urethral overactivity; pelvic floor muscle thickness reflects the muscle condition of the pelvic floor muscles, and thinness or uneven thickening may indicate muscle atrophy, dysfunction, or the effects of postoperative scar tissue.

[0049] After acquiring these anatomical parameters, the system will highlight any abnormalities within the normal reference range. Excessive bladder neck mobility may indicate a risk of urinary incontinence; urethral rotation angles exceeding the normal range may be related to pelvic organ prolapse; and significantly reduced or uneven pelvic floor muscle thickness may indicate decreased muscle strength or injury. The system will provide doctors with intuitive feedback through color coding, warning signs, or risk scores to assist clinical decision-making and ensure the accuracy of diagnosis and the appropriateness of treatment plans.

[0050] The user interaction module is a crucial component of the transperineal pelvic floor ultrasound examination system. It primarily displays real-time ultrasound images, presents the analysis results from the data analysis module, and supports doctors in inputting patient information. Through its intuitive interface and interactive design, this module enhances the convenience of ultrasound examinations, enabling doctors to efficiently complete image evaluation, data input, and diagnostic decisions.

[0051] Real-time ultrasound imaging provides clear and smooth ultrasound images, allowing doctors to observe changes in pelvic floor anatomy and function in real time. High-definition real-time ultrasound display: Supports 2D grayscale ultrasound, color Doppler (CFD), tissue harmonic imaging (THI), and other modes, with dynamic adjustment of parameters such as brightness, contrast, and gain. 4D ultrasound playback: Allows for pausing and replaying dynamic images, enabling doctors to analyze changes in pelvic floor tissues under different physiological states, such as the Valsalva maneuver and cough test. Multi-view switching: Supports 3D / 4D image reconstruction from different angles to obtain complete structural information of the bladder neck, urethra, and pelvic floor muscles. AI-assisted boundary annotation: Automatically outlines the bladder neck, urethra, and pelvic floor muscles, with manual adjustment to improve annotation accuracy.

[0052] The data analysis results are presented intuitively, showcasing key anatomical parameters and quality assessment results calculated by the data analysis module to assist physicians in decision-making. Automatic calculation and visualization of key pelvic floor anatomical parameters include: Bladder Neck Mobility: Displaying bladder neck displacement using numerical values ​​and dynamic graphs, and providing normal reference ranges. Urethral Rotation Angle: Presenting changes in urethral tilt using angle values ​​and dynamic graphs, and marking abnormalities. Pelvic Floor Muscle Thickness: Displaying muscle thickness measurement results and comparing them with historical data.

[0053] Image quality score (1-10 points), calculated by the data analysis module. High-quality images are highlighted (green) and low-quality images are indicated (red), allowing doctors to quickly determine whether re-acquiring images is necessary. Abnormal parameters are highlighted (e.g., bladder neck mobility exceeding the normal range). Color coding (green / yellow / red) visually indicates the ultrasound assessment results. Intelligent risk scoring (AI-based) predicts whether the patient has pelvic floor dysfunction and suggests further examinations or treatment options.

[0054] The doctor input and patient information management system facilitates the entry of patient information, supports data storage and historical comparison, and improves follow-up management capabilities. Basic patient information entry includes: name, age, medical history, symptom description, etc. Electronic medical record (EMR) integration allows direct access to hospital database information.

[0055] Image data is automatically matched with patient information for easy retrieval later. It supports the DICOM standard and can interface with PACS image storage systems, facilitating remote consultations or long-term follow-up.

[0056] Doctors can adjust the AI-calculated markers for the bladder neck, urethra, and pelvic floor muscles to ensure accuracy. Customizable diagnostic assessments: Doctors can manually enter diagnostic conclusions and store them in the patient's file. Voice input support: Reduces the burden of manual input for doctors and improves work efficiency.

[0057] The user interface is optimized for use with ultrasound equipment touchscreens, allowing for quick selection of menu items and adjustment of parameters. Users can switch between ultrasound images, data analysis, and patient information with a single click, improving operational fluency. Doctors can remotely view images and data via tablets or mobile devices, enhancing follow-up convenience.

[0058] In this embodiment, high-precision ultrasound imaging and intelligent analysis of pelvic floor tissues are achieved by integrating an ultrasound probe module, an imaging module, an image optimization module, a 3D imaging module, a data analysis module, and a user interaction module. The ultrasound probe module transmits and receives ultrasound signals; the imaging module processes the echo signals and generates real-time images; and the image optimization module reduces artifact interference and improves image quality by automatically adjusting gain, contrast, filtering, and tissue harmonic imaging modes. The 3D imaging module supports 3D reconstruction and 4D imaging, capturing changes in pelvic floor muscles in both static and dynamic states. The data analysis module assesses the quality of the optimized ultrasound images and automatically calculates bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness in high-quality images, providing abnormality annotations. The user interaction module displays real-time images and analysis results and supports doctors inputting patient information, enabling precise and intelligent pelvic floor function assessment and clinical decision support.

[0059] Example 2, the method of using transperineal pelvic floor ultrasound examination described in this example includes: S1: The patient lies supine or on their side, relaxes their pelvic floor muscles, and prepares for the examination. S2: The operator applies coupling gel to the ultrasound probe and places it on the patient's perineum, then adjusts the probe angle to obtain the best imaging view. S3: The ultrasound images of the pelvic floor tissues are displayed in real time by the imaging device and enhanced by the image optimization module to reduce artifact interference. S4: Enable the 3D imaging module to acquire 3D reconstructed images of the pelvic floor muscles and analyze key anatomical structures; S5: By combining ultrasound image data with reduced artifacts and pelvic floor muscle change data, the quality of real-time ultrasound images is evaluated, pelvic floor function parameters are calculated, and abnormal indications are provided in conjunction with patient symptoms. S6: Based on ultrasound images and analysis results, doctors determine whether the patient has pelvic floor dysfunction and develop corresponding treatment or rehabilitation plans.

[0060] In step S2, the operator applies a coupling gel to the ultrasound probe, places it on the patient's perineum, and adjusts the probe angle to obtain the optimal imaging view. This includes: To optimize the ultrasound probe angle and ensure high-quality real-time ultrasound imaging, the probe angle was adjusted based on the ultrasound image quality assessment results (quality analysis value Q). Dynamic adjustments are made. By constructing a feedback control system, the probe angle is adaptively adjusted in the direction of optimal image quality, reducing artifacts and enhancing the clarity of tissue boundaries.

[0061] Let the current angle of the ultrasound probe be... (The angle after the nth adjustment), then the dynamic adjustment of the probe angle is expressed as: ;in: The current angle of the ultrasound probe (unit: degrees). The adjusted ultrasound probe angle (unit: degrees). Q is the quality analysis value of the real-time ultrasound image, calculated by the data analysis module (range: 0≤Q≤1). This represents the gradient of the image quality analysis value with respect to the probe angle, i.e., the rate of quality change at the current angle. This is the learning rate (step size factor), used to control the adjustment range. A typical value is 0.1-0.5 to prevent oscillations or slow convergence.

[0062] Taking the partial derivative with respect to the probe angle, we get: ;in: , The weighting coefficients determine the contributions of TBSI and PFDC to image quality calculations, and are typically set to [value missing]. .

[0063] This refers to the effect of probe angle on tissue boundary stability. If the boundary is blurred, the partial derivative of this term is small, and the probe angle needs to be adjusted. To assess the impact of probe angle on the dynamic deformation of the pelvic floor muscles, if the pelvic floor muscle contour is abnormally deformed, this partial derivative is small, and the probe angle needs to be adjusted.

[0064] To achieve dynamic adjustment, an iterative optimization strategy is adopted: Initial angle setting: Set the initial probe angle. It is usually between 30° and 60°.

[0065] Calculate image quality analysis value : Calculate using current imagery .

[0066] Calculate gradient : Find the rate of change of mass with respect to angle.

[0067] Update probe angle: .

[0068] Convergence criteria: If the change in image quality is less than the set threshold ϵ (e.g., 0.01), stop adjusting; otherwise, continue iterating.

[0069] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A transperineal pelvic floor color Doppler ultrasound examination system, characterized in that: It includes an ultrasound probe module, an imaging module, an image optimization module, a 3D imaging module, a data analysis module, and a user interaction module; The ultrasound probe module is used to transmit and receive ultrasound signals and acquire echo signals from the pelvic floor tissues. The imaging module, connected to the ultrasound probe module, is used to receive echo signals and generate real-time ultrasound images; The image optimization module is used to automatically adjust ultrasound imaging parameters, including gain, contrast, filtering, and tissue harmonic imaging mode, to reduce artifacts caused by the complex structure of the pelvic floor tissue. The 3D imaging module is used for 3D reconstruction of pelvic floor tissues, capturing data on changes in pelvic floor muscles in static and dynamic states through 4D imaging technology. The data analysis module is used to combine ultrasound image data with artifact reduction and pelvic floor muscle change data to evaluate the quality of real-time ultrasound images. For high-quality real-time ultrasound images, it automatically calculates key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle, and pelvic floor muscle thickness, and provides abnormal annotation prompts. Among them, a tissue boundary stability index is generated by measuring whether the boundaries of pelvic floor tissues are clear and stable in ultrasound images. The method for obtaining the tissue boundary stability index is as follows: Preprocess the ultrasound image. If the input image is color, convert it to a grayscale image I(x,y). Smooth the image using a Gaussian kernel. Calculate the expression: Where: I(x,y) is the grayscale value of the original ultrasound image. It is a two-dimensional Gaussian filter kernel. This represents the convolution operation, where σ is the standard deviation of the Gaussian filter. This is a smoothed grayscale image; Calculate the horizontal direction and vertical direction The gradient is obtained using a Sobel filter: , ; Calculate the gradient magnitude: ; Calculate the gradient direction: ; Along the gradient direction Refine along the gradient direction, only keep the local maxima, set a high threshold Th and a low threshold Tl, and screen the edges: if G(x,y) > Th, mark it as a strong edge point; if Tl < G(x,y) < Th, mark it as a weak edge point, if it is connected to a strong edge point, keep it, otherwise discard it; if G(x,y) < Tl, discard it; Calculate the local contrast of pixels within the boundary region The expression is: Where: N is the number of boundary points, It is the average gray value of all boundary points, where E represents the set of edge pixels in the image. The gradient direction consistency value of the boundary points is calculated. The expression is: ;in: It is the average gradient direction at all boundary points; The tissue boundary stability index is obtained by weighted average summation of the boundary contrast change rate and gradient consistency value. Among them, the dynamic deformation index of the pelvic floor muscles is generated by measuring the consistency and repeatability of the deformation of the pelvic floor muscles under different physiological states. The method for obtaining the dynamic deformation index of the pelvic floor muscles is as follows: Preprocess ultrasound image sequences; if the ultrasound images are in RGB format, convert them to grayscale images. Calculate the horizontal direction and vertical direction The gradient is obtained using a Sobel filter: , ; Calculate the time gradient: ; The fundamental equation for optical flow is as follows: (This is used to calculate the velocity of each pixel.) Where: u is the displacement velocity in the x-direction, and v is the displacement velocity in the y-direction. , It is the spatial gradient. It is the time gradient; If we assume that a local pixel block maintains a constant velocity within a window of size W, then: ; Solve for u and v to obtain the displacement vector of each pixel; Calculate the average motion velocity of all pixels in the pelvic floor muscle region using the following expression: ;in: The dynamic deformation index (PFDC) of the pelvic floor muscles represents the local deformation velocity, where u and v are the horizontal and vertical movement velocities, respectively. The calculation expression is as follows: Where M is the total number of pixels in the pelvic floor muscle region. It is the deformation speed of each pixel; The tissue boundary stability index and pelvic floor muscle dynamic deformation index are converted into a comprehensive feature vector. The comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses the prediction of the quality analysis value label of the real-time ultrasound image for each set of comprehensive feature vectors as the prediction objective. The training objective is to minimize the sum of prediction errors of the quality analysis value labels of all real-time ultrasound images. The machine learning model is trained until the sum of prediction errors converges and the model training stops. The quality analysis value of the real-time ultrasound image is determined based on the model output. The machine learning model is a multinomial regression model. The user interaction module displays real-time ultrasound images and the analysis results from the data analysis module, and supports doctors in inputting patient information.

2. The transperineal pelvic floor color Doppler ultrasound examination system according to claim 1, characterized in that: In the data analysis module, the quality analysis value of the acquired real-time ultrasound image is compared with the preset quality analysis value reference threshold based on historical data. If the quality analysis value of the real-time ultrasound image is greater than or equal to the preset quality analysis value reference threshold, it indicates that the quality of the acquired real-time ultrasound image is high and it is classified as a high-quality real-time ultrasound image. For high-quality real-time ultrasound images, key pelvic floor anatomical parameters, including bladder neck mobility, urethral rotation angle and pelvic floor muscle thickness, are automatically calculated and abnormal annotation prompts are provided. If the quality analysis value of the real-time ultrasound image is less than the preset quality analysis value reference threshold, it indicates that the quality of the acquired real-time ultrasound image is low. It should be classified as a low-quality real-time ultrasound image, and the image needs to be acquired again or the ultrasound parameters need to be adjusted.

3. A method for using transperineal pelvic floor color Doppler ultrasound, characterized in that: The method of use employs a transperineal pelvic floor color Doppler ultrasound examination system as described in any one of claims 1-2, comprising: S1: The patient lies supine or on their side, relaxes their pelvic floor muscles, and prepares for the examination. S2: The operator applies coupling gel to the ultrasound probe and places it on the patient's perineum, then adjusts the probe angle to obtain the best imaging view. S3: The ultrasound images of the pelvic floor tissues are displayed in real time by the imaging device and enhanced by the image optimization module to reduce artifact interference. S4: Enable the 3D imaging module to acquire 3D reconstructed images of the pelvic floor muscles and analyze key anatomical structures; S5: By combining ultrasound image data with reduced artifacts and pelvic floor muscle change data, the quality of real-time ultrasound images is evaluated, pelvic floor function parameters are calculated, and abnormal indications are provided in conjunction with patient symptoms. S6: Based on ultrasound images and analysis results, doctors determine whether the patient has pelvic floor dysfunction and develop corresponding treatment or rehabilitation plans.

4. The method of using transperineal pelvic floor color Doppler ultrasound examination according to claim 3, characterized in that: In S2, the operator applies coupling gel to the ultrasound probe, places it on the patient's perineum, and adjusts the probe angle to obtain the optimal imaging view. This includes: Let the current angle of the ultrasound probe be... The dynamic adjustment of the probe angle is then expressed as: ;in: The current angle of the ultrasound probe. The adjusted ultrasound probe angle is represented by Q, which is the real-time ultrasound image quality analysis value calculated by the data analysis module. The gradient of the image quality analysis value with respect to the probe angle is the rate of quality change at the current angle, and λ is the learning rate, which is used to control the adjustment range. Taking the partial derivative with respect to the probe angle, we get: Where: α and β are weighting coefficients, set as α + β = 1; The influence of probe angle on tissue boundary stability is considered. If the boundary is blurred, the partial derivative of this term is small, and the probe angle needs to be adjusted. To assess the impact of probe angle on the dynamic deformation of the pelvic floor muscles, if the pelvic floor muscle contour is abnormally deformed, this partial derivative is small, and the probe angle needs to be adjusted.

5. The method of using a transperineal pelvic floor color Doppler ultrasound examination according to claim 4, characterized in that: In S2, an iterative optimization strategy is adopted to achieve dynamic adjustment. The iterative optimization strategy includes: Set the initial probe angle Between 30° and 60°; calculated using current imagery. ; Calculate the rate of change of mass with respect to angle, and update the probe angle: If the change in image quality is less than the set threshold ϵ, stop adjusting; otherwise, continue iterating.

Images