An artificial intelligence identification method for urodynamic examination images

Abstract

The present application relates to the field of image analysis, and particularly relates to an artificial intelligence identification method for urodynamic examination images. The present application obtains urodynamic contrast images and monitoring point displacement information, calculates a velocity time domain curve based on the displacement information, and captures a steady state time period. The present application positions a key contrast frame through time sequence alignment, selects adjacent continuous frames to construct a reference frame group through a sliding window, identifies a reference frame group organization contour to generate a superimposed thermal map, determines a superimposed main contour through frequency threshold screening and contour closure processing, calculates the interval width perpendicular to the contour based on displacement extreme values, constructs a strip-shaped dynamic reference interval along the main contour, and performs local enhancement on the images in the tissue contour reference interval. The present application specifically considers dynamic interference caused by physiological movement, improves the key area development quality, improves the image processing efficiency, and helps doctors to improve the diagnosis speed and efficiency.

Description

Technical Field

[0001] This invention relates to the field of image analysis, and more particularly to an artificial intelligence recognition method for urodynamic examination images. Background Technology

[0002] Urodynamic testing requires a comprehensive assessment of clinical symptoms and lower urinary tract function. Its main goal is to reproduce the patient's symptoms and determine the cause of the symptoms through urodynamic measurements or observations. The interpretation of urodynamic test results should take into account the patient's medical history, including symptoms, comorbidities / conditions, and as much other information as possible, such as residual urine volume, voiding diary, etc.

[0003] For example, Chinese Patent Publication No. CN101933812A discloses a urodynamic testing and analysis method, including the following steps: establishing an elastic element model of the bladder; establishing a urethral model; keeping the anterior urethra perpendicular to the direction of gravity, recording and measuring voiding data, and calculating voiding parameters; calculating the contraction length of the elastic element, and further calculating the contraction velocity, contraction acceleration, and maximum contraction acceleration of the elastic element; and calculating the maximum cross-sectional area of ​​the urethral model. This invention can completely overcome the pain and potential infection caused to patients by traditional invasive urodynamic examination methods. The entire analysis process can be performed automatically with the help of a computer, and the results are clear and easy to understand, facilitating clinical memorization and use. The device made using this method has a simple structure and is easy to maintain, thereby reducing medical costs.

[0004] However, the following problems still exist in the existing technology.

[0005] 1. Current technologies do not consider the dynamic interference caused by physiological displacements such as patient breathing during the examination on urodynamic imaging. This leads to systematic biases in parameter calculations based on static or quasi-static models, making it impossible to accurately capture the true morphological changes of organ boundaries in dynamic scenes. This results in higher difficulty for subsequent physician evaluation and diagnosis, and a greater workload for manual intervention.

[0006] 2. In the existing technology, the image processing method is not adapted, which leads to the degradation of the intra-frame tissue boundary recognition accuracy due to severe motion artifacts in key physiological stages such as bladder filling and voiding, making subsequent doctor evaluation and diagnosis more difficult and requiring a large amount of manual work. Summary of the Invention

[0007] Therefore, this invention provides an artificial intelligence recognition method for urodynamic examination images to overcome the problems of existing technologies that do not consider the dynamic interference caused by the patient's physiological displacement such as breathing during the examination on the urodynamic contrast images, and the image processing methods that are not adaptively adjusted, resulting in high difficulty for doctors to make subsequent evaluation and diagnosis, and a large amount of manual work.

[0008] To achieve the above objectives, the present invention provides an artificial intelligence recognition method for urodynamic examination images, comprising:

[0009] Continuously acquire urodynamic contrast images of the monitored target and simultaneously acquire displacement information of the corresponding monitoring points of the monitored target;

[0010] Based on displacement information, determine the time-domain variation curve of the velocity at the monitoring point in order to capture the steady-state time point of the velocity;

[0011] Align the velocity time-domain change curve with the start time of the urodynamic contrast imaging image, capture key contrast frames in the urodynamic contrast imaging image based on the steady-state time point, and determine several adjacent contrast frames based on the key contrast frames to construct a reference frame group;

[0012] The tissue contours of each contrast frame in the reference frame group are identified, and each contrast frame is superimposed to determine the main superimposed contour. The interval width is determined based on the displacement information, and a confidence tissue contour interval is constructed based on the interval width and the main superimposed contour.

[0013] Based on the confidence tissue contour region, enhancement is performed on the remaining contrast frames of the non-reference frame group in urodynamic contrast imaging, including,

[0014] Determine the confidence tissue contour region in the remaining imaging frames, and perform local enhancement on the image within the confidence tissue contour region;

[0015] The displacement information includes the coordinate data of the monitoring point at each time point.

[0016] Furthermore, the displacement information of the monitoring points corresponding to the monitoring target is obtained, including,

[0017] The coordinate data of several monitoring points of the monitored target in three-dimensional space are collected in real time by a spatial position sensing device.

[0018] The testing points include chest cavity monitoring points and abdominal monitoring points.

[0019] Furthermore, the process of determining the time-domain variation curve of the velocity at the monitoring point based on displacement information to capture the steady-state velocity time point includes:

[0020] The instantaneous velocity of each monitoring point at several moments is calculated based on the displacement information;

[0021] Calculate the instantaneous average velocity at each monitoring point and generate a velocity time-domain variation curve;

[0022] Based on a set speed threshold, a steady-state time period in which the average instantaneous speed is continuously lower than the speed threshold is selected, and the lowest speed point within the time period is determined as the steady-state speed time point.

[0023] Furthermore, the process of aligning the velocity time-domain variation curve with the start time of the urodynamic contrast imaging image, and extracting key contrast frames from the urodynamic contrast imaging image based on the steady-state time point, includes:

[0024] Determine the time points of each frame sequence of urodynamic imaging images from the velocity-time domain variation curve;

[0025] The imaging frame that is closest in time to the steady-state velocity time point is selected as the key imaging frame.

[0026] Furthermore, the process of determining several adjacent contrast frames based on key contrast frames to construct a reference frame group includes,

[0027] Based on the key angiography frame, several consecutive frames forward and backward of the key angiography frame are selected within the steady-state time period to construct a reference frame group.

[0028] Furthermore, the process of identifying the tissue contours of each contrast frame in the reference frame group and overlaying the contrast frames includes,

[0029] Image segmentation processing is performed on each angiography frame in the reference frame group to extract the corresponding contour in each angiography frame;

[0030] Align the images of each contrast frame after placing them in the same spatial coordinate system;

[0031] Based on the aligned images of each angiography frame, the frequency of each spatial location point being covered by the contour is calculated, and a contour overlay heatmap is generated.

[0032] Furthermore, the process of determining the superimposed main contour includes,

[0033] A predetermined frequency threshold is set, and contours with a coverage frequency exceeding the frequency threshold are selected from the contour overlay heatmap and used as the overlay main contours.

[0034] Furthermore, the process of constructing a confidence organization contour interval based on the interval width and the superimposed main contour includes,

[0035] Based on the superimposed main contour, virtual reference main contours are constructed on both sides at a distance from the interval width;

[0036] The band-shaped region formed by the virtual reference master contour is determined as the confidence organization contour interval.

[0037] Furthermore, the interval width is determined based on displacement information, including:

[0038] Determine the average instantaneous velocity of each detection point within the time domain segment corresponding to the reference frame group;

[0039] The width of the interval is positively correlated with the mean instantaneous velocity.

[0040] Furthermore, it also includes marking confidence tissue contour intervals in each of the remaining imaging frames.

[0041] Compared with existing technologies, this invention acquires urodynamic contrast images and monitoring point displacement information, calculates velocity time-domain curves based on displacement information, and captures steady-state time periods. It locates key contrast frames through temporal alignment and constructs reference frame groups by selecting adjacent continuous frames using a sliding window. It identifies tissue contours within the reference frame groups to generate overlay heatmaps, and determines the main overlay contour through frequency thresholding and contour closure processing. Based on displacement extrema, it calculates the width of the interval perpendicular to the contour and constructs a strip-shaped dynamic reference interval along the main contour. Local enhancement is then applied to the image within the tissue contour reference interval. This invention specifically considers dynamic interference caused by physiological movement, improves the imaging quality of key areas, increases image processing efficiency, and helps doctors improve diagnostic speed and efficiency.

[0042] In particular, this invention considers the coordinated processing of physiological displacement and image acquisition. During urodynamic examinations, the displacement of abdominal organs caused by the patient's respiratory movements can lead to distortion in contrast images. Traditional methods lack precise mapping between motion trajectories and image frames. This invention simultaneously acquires displacement information and contrast images, and based on the velocity time-domain change curve, locks in steady-state time periods where the velocity is consistently below a threshold. It accurately establishes the correspondence between physiological motion states and image acquisition, determining steady-state time periods with minimal physiological activity of the monitored target. Within these steady-state time periods, the impact of motion on imaging quality is minimal. Using this as a benchmark, a set of reference frames with better imaging quality is selected as a basis to support the subsequent determination of confident tissue contour intervals. This allows for the use of specific image enhancement methods to focus on key areas, reduce noise introduction, and specifically address dynamic interference caused by physiological motion, thereby improving the imaging quality of key areas.

[0043] In particular, this invention determines the superimposed main contour based on a reference frame group. In practice, physiological activities such as body shaking and coughing of the monitored target can affect the imaging accuracy, produce artifacts, and cause image distortion. In reality, artifacts are mostly derived from the main contour of the main organs and are scattered throughout the main contour. Based on this, a superimposed processing method is adopted to determine the main contour based on the contour superimposed heat map, and then construct a confidence tissue contour interval. The confidence tissue contour interval represents the area where the main contour mainly exists. Since it is determined based on the reference frame group, the confidence level is high, which facilitates subsequent focusing on key areas. It is also convenient to determine the key areas to be focused on for the remaining imaging frames with strong motion effects. Subsequently, a local enhancement method is adopted to save enhancement computing power. Especially when there are many imaging images to be processed in hospitals, it saves processing speed and ensures reliability. Attached Figure Description

[0044] Figure 1 A schematic diagram illustrating the steps of an artificial intelligence recognition method for urodynamic examination images according to an embodiment of the invention;

[0045] Figure 2 A logic block diagram for filtering steady-state time periods according to an embodiment of the invention;

[0046] Figure 3 This is a logic block diagram for filtering contours in a contour overlay heatmap, as an embodiment of the invention. Detailed Implementation

[0047] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0048] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0049] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0050] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0051] Please see Figure 1 The diagram illustrates the steps of an artificial intelligence recognition method for urodynamic examination images according to an embodiment of the present invention. The method includes:

[0052] Step S1: Continuously acquire urodynamic imaging images of the monitored target and simultaneously acquire displacement information of the corresponding monitoring points of the monitored target;

[0053] Step S2: Determine the time-domain change curve of the velocity at the monitoring point based on the displacement information in order to capture the steady-state time point of the velocity;

[0054] Step S3: Align the velocity time-domain change curve with the start time of the urodynamic contrast imaging image, capture key contrast frames in the urodynamic contrast imaging image based on the steady-state time point, and determine several adjacent contrast frames based on the key contrast frames to construct a reference frame group.

[0055] Step S4: Identify the tissue contour of each contrast frame in the reference frame group, perform superposition processing on each contrast frame, determine the superposition main contour, determine the interval width based on displacement information, and construct a confidence tissue contour interval based on the interval width and the superposition main contour.

[0056] Step S5, based on the confidence tissue contour interval, enhance the remaining contrast frames in the non-reference frame group of the urodynamic contrast imaging image, including,

[0057] Determine the confidence tissue contour region in the remaining imaging frames, and perform local enhancement on the image within the confidence tissue contour region;

[0058] The displacement information includes the coordinate data of the monitoring point at each time point.

[0059] Specifically, urodynamic imaging images can be X-ray images or other forms of imaging images, which will not be elaborated further.

[0060] Specifically, there is no limitation on the method of image enhancement. Those skilled in the art can choose any medical imaging image enhancement algorithm in the prior art to enhance local images in order to improve image readability and image display effect, which will not be elaborated here.

[0061] Specifically, this involves acquiring the displacement information of the monitoring points corresponding to the monitoring target, including...

[0062] The coordinate data of several monitoring points of the monitored target in three-dimensional space are collected in real time by a spatial position sensing device.

[0063] The testing points include chest cavity monitoring points and abdominal monitoring points.

[0064] There are no restrictions on the form of the spatial position sensing device. For example, a binocular camera device can be used, with pre-determined chest and abdominal monitoring points as detection points. The binocular camera device can then acquire the coordinate data of the detection points in three-dimensional space. Of course, other forms are also possible, which will not be elaborated here.

[0065] Specifically, obtaining the patient's authorization is required to collect displacement information and other information of the monitored target, which will not be elaborated further here.

[0066] Specifically, please refer to Figure 2 As shown, Figure 2This is a logic block diagram for filtering steady-state time periods according to an embodiment of the present invention. The process of determining the time-domain change curve of the monitoring point velocity based on displacement information to capture the steady-state velocity time point includes:

[0067] The instantaneous velocity of each monitoring point at several moments is calculated based on the displacement information;

[0068] Calculate the instantaneous average velocity at each monitoring point and generate a velocity time-domain variation curve;

[0069] Based on a set speed threshold, a steady-state time period in which the average instantaneous speed is continuously lower than the speed threshold is selected, and the lowest speed point within the time period is determined as the steady-state speed time point.

[0070] In practice, the instantaneous velocity is calculated based on the distance between the coordinate data of the current moment and the coordinate data of the adjacent previous moment, and the velocity calculated based on the time length between the moments is used as the instantaneous velocity of the current moment.

[0071] In practice, the velocity threshold is predetermined. The instantaneous velocity average value of corresponding monitoring points during the urodynamic contrast imaging process of several patients is recorded in advance. The mean value of each instantaneous velocity average value is calculated to characterize the subtle movement of the patient under normal conditions. The velocity threshold is set as the product of the mean value and the error coefficient, and the error coefficient is selected in the interval [0.9, 0.95].

[0072] Specifically, the process of aligning the velocity time-domain variation curve with the start time of the urodynamic contrast imaging image, and extracting key contrast frames from the urodynamic contrast imaging image based on the steady-state time point, includes:

[0073] Determine the time points of each frame sequence of urodynamic imaging images from the velocity-time domain variation curve;

[0074] The imaging frame closest in time to the steady-state velocity time point is selected as the key imaging frame. In reality, the instantaneous velocity at the steady-state velocity time point is relatively small, and the imaging quality is less affected by motion, so it is selected as the key imaging frame.

[0075] Understandably, the velocity time-domain change curve needs to be aligned with the time domain of the urodynamic imaging image for subsequent analysis.

[0076] Specifically, the process of determining several adjacent contrast-enhanced frames based on key contrast-enhanced frames to construct a reference frame group includes,

[0077] Based on the key angiography frame, several consecutive frames forward and backward of the key angiography frame are selected within the steady-state time period to construct a reference frame group.

[0078] In practice, other angiography frames close to the key angiography frame are less affected by motion. In practice, consecutive angiography frames within 0.3 to 0.5 seconds in the forward and backward directions can be selected, with 0.3 seconds being preferred.

[0079] This invention considers the coordinated processing of physiological displacement and image acquisition. During urodynamic examinations, the displacement of abdominal organs caused by the patient's respiratory movements can lead to distortion in contrast images. Traditional methods lack precise mapping between motion trajectories and image frames. This invention simultaneously acquires displacement information and contrast images, and based on the velocity time-domain change curve, locks in steady-state time periods where the velocity is consistently below a threshold. It accurately establishes the correspondence between physiological motion states and image acquisition, determining steady-state time periods with minimal physiological activity of the monitored target. Within these steady-state time periods, the impact of motion on imaging quality is minimal. Using this as a benchmark, a set of reference frames with better imaging quality is selected as a basis to support the subsequent determination of confident tissue contour regions. This allows for the use of specific image enhancement methods, thereby specifically considering the dynamic interference caused by physiological motion and improving the imaging quality of key areas.

[0080] Specifically, the process of identifying the tissue contours of each contrast frame in the reference frame group and overlaying the contrast frames includes,

[0081] Image segmentation processing is performed on each angiography frame in the reference frame group to extract the corresponding contour in each angiography frame;

[0082] Align the images of each contrast frame after placing them in the same spatial coordinate system;

[0083] Based on the aligned images of each angiography frame, the frequency of each spatial location point being covered by the contour is calculated, and a contour overlay heatmap is generated.

[0084] In practice, the contour overlay heatmap reflects the frequency of each spatial point in the imaging frame being covered by a contour. Even if there are artifacts in the contour of the main tissue, their existence is normal. Therefore, if there are some spatial points covered by contours in the contour overlay heatmap with a high frequency, they can be identified as the main contour.

[0085] Specifically, the process of determining the overlay master contour includes,

[0086] A predetermined frequency threshold is set, and contours with a coverage frequency exceeding the frequency threshold are selected from the contour overlay heatmap and used as the overlay main contours.

[0087] In practice, the frequency threshold is predetermined. Under authorized conditions, urodynamic imaging images of several patients are statistically analyzed in advance. The contours of the main organs in the urodynamic imaging images are pre-annotated by a person skilled in the art. The average coverage frequency of the marked contours in the contour overlay heat map is determined. For each patient, there is an average coverage frequency. Finally, the mean of the average coverage frequency is calculated and set as the frequency threshold.

[0088] Specifically, the process of constructing a confidence organization contour interval based on the interval width and the superimposed main contour includes,

[0089] Based on the superimposed main contour, virtual reference main contours are constructed on both sides at a distance from the interval width;

[0090] The band-shaped region formed by the virtual reference master contour is determined as the confidence organization contour interval.

[0091] It is understandable that the proposed reference master contour can be obtained by copying the superimposed master contour. After copying and superimposing the master contour, it is moved to form a virtual reference master contour. After the two virtual reference master contours are sealed at both ends, a strip-shaped region can be formed, forming a confidence tissue contour interval, which will not be elaborated further.

[0092] In practice, the superimposed main contour can be divided into several contour segments before constructing a virtual reference main contour, and then a strip region can be constructed. The strip regions can be spliced ​​together to form a confidence organization contour interval.

[0093] Specifically, the interval width is determined based on displacement information, including:

[0094] Determine the average instantaneous velocity of each detection point within the time domain segment corresponding to the reference frame group;

[0095] The width of the interval is positively correlated with the mean instantaneous velocity.

[0096] In practice, optional

[0097] The interval width is set as the product of the width reference value and the scaling factor. The scaling factor is the ratio of the instantaneous average speed to the speed threshold. The width reference value is determined based on the maximum width of the reference main contour and is set between 3 and 5 times the reference main contour, with 3 times being the preferred setting.

[0098] Specifically, this also includes marking confidence tissue contour intervals in each of the remaining contrast frames.

[0099] This invention determines the superimposed main contour based on a reference frame group. In practice, physiological activities such as body shaking and coughing can affect the imaging accuracy, producing artifacts and causing image distortion. In reality, artifacts are mostly derived from the main contour of major organs and are scattered throughout the main contour. Based on this, a superimposed processing method is adopted to determine the main contour based on the contour superimposed heat map, and then construct a confidence tissue contour interval. The confidence tissue contour interval represents the main area where the main contour exists. Since it is determined based on the reference frame group, the confidence level is high, which facilitates subsequent focusing on key areas. It is also convenient to determine the key areas to be focused on for the remaining imaging frames with strong motion effects. Subsequently, a local enhancement method is adopted to save enhancement computing power, especially when there are many imaging images to be processed in hospitals, saving processing speed and ensuring reliability.

[0100] If the artificial intelligence recognition method for urodynamic examination images of the present invention is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0101] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An artificial intelligence recognition method for urodynamic examination images, characterized in that, include: Continuously acquire urodynamic contrast images of the monitored target and simultaneously acquire displacement information of the corresponding monitoring points of the monitored target; Based on displacement information, determine the time-domain variation curve of the velocity at the monitoring point in order to capture the steady-state time point of the velocity; Align the velocity time-domain change curve with the start time of the urodynamic contrast imaging image, capture key contrast frames in the urodynamic contrast imaging image based on the steady-state time point, and determine several adjacent contrast frames based on the key contrast frames to construct a reference frame group; The tissue contours of each contrast frame in the reference frame group are identified, and each contrast frame is superimposed to determine the main superimposed contour. The interval width is determined based on the displacement information, and a confidence tissue contour interval is constructed based on the interval width and the main superimposed contour. Based on the confidence tissue contour region, enhancement is performed on the remaining contrast frames of the non-reference frame group in urodynamic contrast imaging, including, Determine the confidence tissue contour region in the remaining imaging frames, and perform local enhancement on the image within the confidence tissue contour region; The displacement information includes the coordinate data of the monitoring point at each time point.

2. The artificial intelligence recognition method for urodynamic examination images according to claim 1, characterized in that, Obtain displacement information of the monitoring points corresponding to the monitoring target, including: The coordinate data of several monitoring points of the monitored target in three-dimensional space are collected in real time by a spatial position sensing device. The testing points include chest cavity monitoring points and abdominal monitoring points.

3. The artificial intelligence recognition method for urodynamic examination images according to claim 1, characterized in that, The process of determining the time-domain variation curve of the velocity at the monitoring point based on displacement information, in order to capture the steady-state velocity time point, includes the following: The instantaneous velocity of each monitoring point at several moments is calculated based on the displacement information; Calculate the instantaneous average velocity at each monitoring point and generate a velocity time-domain variation curve; Based on a set speed threshold, a steady-state time period in which the average instantaneous speed is continuously lower than the speed threshold is selected, and the lowest speed point within the time period is determined as the steady-state speed time point.

4. The artificial intelligence recognition method for urodynamic examination images according to claim 1, characterized in that, Aligning the velocity time-domain variation curve with the start time of the urodynamic contrast imaging image, the process of extracting key contrast frames from the urodynamic contrast imaging image based on the steady-state time point includes: Determine the time points of each frame sequence of urodynamic imaging images from the velocity-time domain variation curve; The imaging frame that is closest in time to the steady-state velocity time point is selected as the key imaging frame.

5. The artificial intelligence recognition method for urodynamic examination images according to claim 3, characterized in that, The process of determining several adjacent contrast frames based on key contrast frames to construct a reference frame group includes, Based on the key angiography frame, several consecutive frames forward and backward of the key angiography frame are selected within the steady-state time period to construct a reference frame group.

6. The artificial intelligence recognition method for urodynamic examination images according to claim 1, characterized in that, The process of identifying the tissue contours of each contrast frame in the reference frame group and overlaying the contrast frames includes: Image segmentation processing is performed on each angiography frame in the reference frame group to extract the corresponding contour in each angiography frame; Align the images of each contrast frame after placing them in the same spatial coordinate system; Based on the aligned images of each angiography frame, the frequency of each spatial location point being covered by the contour is calculated, and a contour overlay heatmap is generated.

7. The artificial intelligence recognition method for urodynamic examination images according to claim 6, characterized in that, The process of determining the superimposed main contour includes, A predetermined frequency threshold is set, and contours with a coverage frequency exceeding the frequency threshold are selected from the contour overlay heatmap and used as the overlay main contours.

8. The artificial intelligence recognition method for urodynamic examination images according to claim 1, characterized in that, The process of constructing a confidence organization contour interval based on the interval width and the superimposed main contour includes: Based on the superimposed main contour, virtual reference main contours are constructed on both sides at a distance from the interval width; The band-shaped region formed by the virtual reference master contour is determined as the confidence organization contour interval.

9. The artificial intelligence recognition method for urodynamic examination images according to claim 8, characterized in that, The interval width is determined based on displacement information, including: Determine the average instantaneous velocity of each detection point within the time domain segment corresponding to the reference frame group; The width of the interval is positively correlated with the mean instantaneous velocity.

10. The artificial intelligence recognition method for urodynamic examination images according to claim 1, characterized in that, It also includes marking the confidence tissue contour interval in each of the remaining imaging frames.

Images