A system and method for evaluating the transmembrane state of the bladder three-dimensional morphology

Abstract

This invention provides a transmembrane assessment system and method for three-dimensional bladder morphology, applied in the field of data processing technology. This application preprocesses bladder ultrasound image information of a target patient to generate target ultrasound image sequence information; processes the target ultrasound image sequence information to generate target enhanced ultrasound images; processes the patient's urinary system physiological parameters based on a target bladder morphology assessment model to generate target physiological state characteristics; processes the target physiological state characteristics, target enhanced ultrasound images, and functional imaging data based on the target bladder morphology assessment model to generate a predicted CT image of the target patient; and processes the predicted CT image and simulated localization CT image information of the target patient based on the target bladder morphology assessment model combined with bladder dynamic image information to generate a three-dimensional bladder morphology assessment result.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a transmembrane assessment system and method for three-dimensional bladder morphology. Background Technology

[0002] In external beam radiotherapy for pelvic tumors, variations in bladder fullness can significantly impact the radiotherapy outcome. Inconsistent bladder fullness may result in insufficient coverage of the target area, failing to effectively kill tumor cells and leading to poor treatment results; or the irradiation area may be too large, causing unnecessary damage to surrounding normal tissues and increasing radiotherapy side effects.

[0003] Existing technologies for bladder capacity measurement mainly include two-dimensional and three-dimensional ultrasound imaging, CT imaging, and methods based on bladder bioelectrical impedance analysis. However, these methods are limited to bladder capacity measurement and lack in-depth application in bladder three-dimensional morphological assessment, failing to meet the clinical need for precise bladder morphological evaluation. Furthermore, while some patents address the anterior-posterior registration of ultrasound in the bladder, they only focus on registration between ultrasound images, lacking cross-modal assessment between different modalities of imaging such as ultrasound and CT. In actual radiotherapy, localization CT is the gold standard for treatment planning, and subsequent registration with ultrasound images is necessary to monitor bladder three-dimensional morphology in real time to determine when treatment should begin. However, existing technologies struggle to achieve accurate cross-modal assessment, affecting the accuracy and effectiveness of radiotherapy planning.

[0004] Finally, traditional cone-beam computed tomography (CBCT) assessment methods, due to limitations in their radiation characteristics and imaging modes, cannot achieve real-time assessment of the bladder's three-dimensional morphology and capacity. This is extremely inconvenient in scenarios requiring real-time monitoring of the bladder to adjust radiotherapy plans, and cannot provide timely, accurate, and dynamic bladder information for clinical treatment.

[0005] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore includes information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide a cross-membranous assessment system and method for three-dimensional bladder morphology, which at least to some extent overcomes the problems existing in the prior art. By sequentially preprocessing and enhancing bladder ultrasound images, image quality is improved, highlighting bladder structural information. Multi-source data is processed using a model. This includes classifying, extracting, normalizing, analyzing, and fusing physiological parameters of the urinary system to generate target physiological state features; and mining, fusing, and processing features from ultrasound images and functional imaging data in multiple steps to generate predicted CT images. Finally, by combining dynamic bladder images, multi-stage processing is performed on the predicted CT images and simulated localization CT images to complete the three-dimensional bladder morphology assessment. The entire process utilizes deep learning and multimodal data fusion technology to convert cross-modal assessment into same-modal assessment, significantly improving the intuitiveness and accuracy of the assessment. This effectively assists physicians in developing precise radiotherapy plans, reducing radiotherapy side effects caused by changes in bladder filling, and providing strong support for pelvic tumor radiotherapy.

[0007] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0008] According to one aspect of this application, a transmembrane assessment method for three-dimensional bladder morphology is provided, comprising: acquiring bladder ultrasound image information of a target patient, simulated positioning CT image information of the target patient, urinary system physiological parameter information of the target patient, functional imaging data, and bladder dynamic image information, wherein the bladder dynamic image information is used to characterize the morphological changes of the bladder under different physiological states; preprocessing the bladder ultrasound image information of the target patient to generate target ultrasound image sequence information; processing the target ultrasound image sequence information to generate target enhanced ultrasound image; processing the urinary system physiological parameter information of the patient based on a target bladder morphology assessment model to generate target physiological state characteristics; processing the target physiological state characteristics, target enhanced ultrasound image, and functional imaging data based on the target bladder morphology assessment model to generate a predicted CT image of the target patient; and processing the predicted CT image and simulated positioning CT image information of the target patient based on the target bladder morphology assessment model combined with the bladder dynamic image information to generate a three-dimensional bladder morphology assessment result.

[0009] Another aspect of this application discloses a transmembrane assessment device for three-dimensional bladder morphology, characterized by comprising: an acquisition module for acquiring bladder ultrasound image information of a target patient, simulated positioning CT image information of the target patient, urinary system physiological parameter information of the target patient, functional imaging data, and bladder dynamic image information, wherein the bladder dynamic image information is used to characterize the morphological changes of the bladder under different physiological states; a processing module for preprocessing the bladder ultrasound image information of the target patient to generate target ultrasound image sequence information; processing the target ultrasound image sequence information to generate target enhanced ultrasound images; processing the urinary system physiological parameter information of the patient based on a target bladder morphology assessment model to generate target physiological state characteristics; processing the target physiological state characteristics, target enhanced ultrasound images, and functional imaging data based on the target bladder morphology assessment model to generate a predicted CT image of the target patient; and processing the predicted CT image and simulated positioning CT image information of the target patient based on the target bladder morphology assessment model combined with the bladder dynamic image information to generate a three-dimensional bladder morphology assessment result.

[0010] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described transmembrane assessment method for the three-dimensional morphology of the bladder.

[0011] This application provides a cross-membranous assessment system and method for three-dimensional bladder morphology. The server acquires multi-source data, including bladder ultrasound images, simulated localization CT images, urinary system physiological parameters, functional imaging data, and bladder dynamic image information, providing a comprehensive basis for assessment. The bladder ultrasound images undergo preprocessing and enhancement to improve image quality and highlight bladder structural information. A model is used to process the multi-source data. For example, urinary system physiological parameters are classified, feature-extracted, normalized, correlated, and fused to generate target physiological state features. Ultrasound images and functional imaging data undergo feature mining, fusion, and multi-step processing to generate predicted CT images. Finally, combined with bladder dynamic images, the predicted CT images and simulated localization CT images undergo multi-stage processing to complete the three-dimensional bladder morphology assessment. The entire process utilizes deep learning and multi-modal data fusion technology to transform cross-modal assessment into same-modal assessment, significantly improving the intuitiveness and accuracy of the assessment. This effectively assists physicians in developing precise radiotherapy plans, reducing radiotherapy side effects caused by changes in bladder filling, and providing strong support for pelvic tumor radiotherapy.

[0012] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0013] Figure 1A flowchart illustrating a transmembrane assessment method for three-dimensional morphology of the bladder provided in an embodiment of this application is shown.

[0014] Figure 2 A schematic diagram of the structure of a transmembrane assessment device for three-dimensional morphology of the bladder provided in an embodiment of this application is shown. Detailed Implementation

[0015] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0016] The following is combined with Figure 1 This application describes a transmembrane assessment method for the three-dimensional morphology of the bladder according to exemplary embodiments thereof. In one embodiment, this application also proposes a transmembrane assessment system and method for the three-dimensional morphology of the bladder. Figure 1 A schematic flowchart illustrating a transmembrane assessment method for three-dimensional bladder morphology according to an embodiment of this application is shown. Figure 1 As shown, this method is applied to a server and includes:

[0017] S101, acquires bladder ultrasound image information, simulated positioning CT image information, urinary system physiological parameter information, functional imaging data, and bladder dynamic image information of the target patient.

[0018] In one implementation, ultrasound diagnostic equipment is used to generate images by emitting ultrasound waves towards the patient's bladder and receiving the reflected echoes. For example, the patient lies on an examination table, and the doctor applies coupling gel to the ultrasound probe and places it on the patient's lower abdomen, adjusting the probe's angle and position to acquire ultrasound images of different sections of the bladder. These images provide a real-time view of the bladder's approximate shape, but may suffer from poor clarity or noise interference; some minute lesions may be difficult to visualize clearly. During the radiotherapy simulation and localization phase, the patient lies on a CT scan table, positioned according to radiotherapy requirements, and then undergoes a CT scan. During the scan, X-rays penetrate the patient's body from different angles, and a detector receives the attenuated radiation signals after passing through the body. These signals are then processed by a computer to reconstruct tomographic images of the bladder. These CT images accurately show the bladder's location, shape, size, and relationship with surrounding tissues, serving as crucial information for radiotherapy planning. For instance, doctors can use these images to determine the target area for radiotherapy.

[0019] Physiological parameters of the urinary system can be obtained through various medical testing methods, including but not limited to laboratory tests and physiological function tests. For example, blood samples can be collected to test kidney function indicators such as creatinine and blood urea nitrogen to assess whether kidney excretion function is normal; urine samples can be collected for routine urinalysis to check for abnormal components such as red blood cells, white blood cells, and proteins, helping to determine whether there is inflammation or infection in the urinary system; bladder pressure measurement can also be used to measure pressure changes in the bladder during filling and emptying to understand the functional status of the bladder. Functional imaging data can be obtained using functional imaging techniques, such as positron emission tomography (PET)-CT or magnetic resonance spectroscopy (MRS). Taking PET-CT as an example, the patient is first injected with a drug labeled with a radionuclide. These drugs participate in the metabolic process in the body and accumulate in metabolically active sites. Then, a PET-CT scan is performed, which detects the gamma rays emitted by the radionuclide to obtain metabolic information of the bladder tissue and determine the functional status of the bladder tissue. If there is a tumor in the bladder, the metabolism of the tumor site is usually more active than that of normal tissue, which will show different signals on the image.

[0020] Dynamic imaging information of the bladder can be obtained using real-time ultrasound monitoring or cystography. Real-time ultrasound monitoring continuously observes the morphological changes of the bladder during different physiological processes such as filling and emptying. The ultrasound equipment can continuously acquire images, forming a dynamic image sequence. Cystography, on the other hand, involves injecting a contrast agent into the bladder and then using X-ray fluoroscopy to dynamically observe changes in the bladder's morphology over time, such as observing bladder wall peristalsis and the smoothness of bladder emptying, thereby comprehensively understanding the morphological changes of the bladder under different physiological states.

[0021] In another implementation, acquiring a target bladder morphology assessment model includes obtaining a training sample set and a pre-defined bladder morphology assessment model. The training sample set includes multimodal data and corresponding three-dimensional bladder morphology assessment results. The hospital collected relevant data from 500 pelvic tumor patients over the past three years from its case database as the training sample set. This data contained multimodal information, specifically: bladder ultrasound images of each patient during radiotherapy simulation positioning, recording the real-time bladder morphology; high-resolution simulation positioning CT images, accurately showing the bladder's location, shape, size, and relationship with surrounding tissues; urinary system physiological parameters obtained through laboratory tests and physiological function tests, such as renal function indicators (creatinine, blood urea nitrogen, etc.), urinalysis results (red blood cell, white blood cell, protein, etc. content), and bladder pressure measurement data; functional imaging data obtained using PET-CT, reflecting the metabolic status of bladder tissue; and dynamic bladder images acquired through real-time ultrasound dynamic monitoring or cystography dynamic imaging technology, presenting morphological changes of the bladder under different physiological states. Simultaneously, the hospital's medical imaging team, based on clinical experience and professional knowledge, assessed the three-dimensional bladder morphology of each patient and incorporated these assessment results into the training sample set. In addition, a neural network based on the Cyclegan architecture is used as the preset bladder morphology evaluation model. This model has powerful feature extraction and processing capabilities and is suitable for processing multimodal data.

[0022] The number of features for each modality in the training sample set was statistically analyzed, and a sampling ratio was generated based on the feature distribution balance. Data analysis tools were used to extract and statistically analyze features from the multimodal data in the training sample set. For example, features of ultrasound images include bladder wall thickness and the uniformity of echoes within the bladder; features of CT images include bladder volume, shape index, and distance from surrounding organs; features of urinary system physiological parameters cover the numerical range and trends of various indicators; features of functional imaging data involve the distribution and intensity of metabolically active areas; and features of bladder dynamic images include bladder wall peristalsis frequency and emptying velocity. Statistical analysis revealed that CT image data has a larger number of features, while functional imaging data has a relatively smaller number. To ensure that data from each modality is fully utilized during training and to avoid any one modality dominating the training results, a sampling ratio was calculated based on the feature distribution balance. Assuming that the number of CT image features is three times the number of functional imaging data features, the sampling ratio for functional imaging data is set to three times that of CT images to balance their influence, and this pattern is used to determine the sampling ratios for other modalities.

[0023] Multimodal stratified sampling is performed on the training sample set based on the sampling ratio to generate a preset number of sampling feature combinations. Taking ultrasound and CT images as examples, a certain number of samples are extracted from the ultrasound image data layer according to its sampling ratio, and samples are simultaneously extracted from the CT image data layer according to the corresponding sampling ratio. The extracted ultrasound and CT image samples are then combined. Similarly, a similar operation is performed on urinary system physiological parameters, functional imaging data, and bladder dynamic image information. Finally, samples extracted from different modalities are combined to form individual sampling feature combinations. Assuming a preset of 1000 sampling feature combinations are generated, this stratified sampling method ensures that each combination contains multimodal data with a relatively balanced proportion of each modality, thus more comprehensively reflecting the characteristics of the bladder.

[0024] Statistical tests are performed on any key feature and each sampling feature combination to divide the data into a response group and a poor response group. Each group contains a predetermined number of data samples, and at least one sample carries lesion identification information. "Whether there is a tumor lesion in the bladder wall" is selected as the key feature for analysis. For each sampling feature combination, statistical tests are used to determine the relationship between the combination and bladder wall tumor lesions. For example, the chi-square test is used to analyze the correlation between features such as abnormal bladder wall echoes in ultrasound images, bladder wall morphological changes in CT images, and metabolically abnormal areas in functional imaging data with tumor lesions. Based on the test results, the data is divided into a response group and a poor response group. Assuming that each group contains 200 data samples, during the division process, it is ensured that each group has at least one sample with clear lesion identification information, such as a sample diagnosed as bladder cancer by pathological biopsy. The purpose of this division is to allow the model to learn the differences between normal bladder morphology and diseased bladder morphology during training.

[0025] A pre-defined bladder morphology assessment model was iteratively trained using a response group and a poor response group. Model parameters were optimized through cross-validation, generating the trained model and performance metrics. Data from the response group and the poor response group were input into the pre-defined Cyclegan-based bladder morphology assessment model for training. During training, the model continuously adjusted its parameters to learn features and patterns in the data. For example, the model learned the relationship between different echogenic features in ultrasound images and bladder lesions, and the connection between bladder morphology parameters in CT images and bladder function. To avoid overfitting, cross-validation was used, dividing the training data into five parts. Four parts were used for training and one part for validation in each iteration, repeated five times. The average of the five validation results was used as the model's performance metrics, such as accuracy, recall, and F1 score. Through multiple rounds of iterative training, the model's parameters were continuously optimized, and its performance gradually improved.

[0026] If the lesion marker information in the training results is identified by the model as a key feature for morphological assessment, then the trained model will be used as the target bladder morphological assessment model. After multiple rounds of training and validation, the model's learning results will be analyzed. If the model can accurately identify the importance of lesion marker information for bladder morphological assessment when predicting bladder 3D morphology, that is, if the lesion marker information plays a key role in the model's decision-making process—for example, if the model can accurately determine the trend of bladder morphological changes based on the characteristics of bladder wall tumors—then the model is considered to have learned effective features, and the trained model will be determined as the target bladder morphological assessment model. This model will be used for subsequent assessment of the bladder 3D morphology of new patients, helping doctors to more accurately assess the patient's bladder status and develop more precise radiotherapy plans.

[0027] S102, preprocess the bladder ultrasound image information of the target patient to generate target ultrasound image sequence information.

[0028] In one implementation, denoising and contrast enhancement are performed on the bladder ultrasound images of the target patient to generate an initial bladder image. The patient lies on an examination bed, and the doctor uses ultrasound diagnostic equipment to acquire bladder ultrasound images. Due to the principles of ultrasound imaging and factors such as scattering and reflection of ultrasound by human tissue, the acquired raw ultrasound images contain noise, and the contrast between the bladder and surrounding tissues is low, affecting the observation of the bladder's morphology. Technicians import these raw ultrasound images into image processing software and use a Gaussian filtering algorithm to denoise the images. Gaussian filtering effectively removes random noise from the image by weighted averaging of each pixel and its neighboring pixels, making the image smoother. Next, a histogram equalization algorithm is used to enhance image contrast. This algorithm expands the dynamic range of image grayscale by redistributing the grayscale values ​​of image pixels, making the boundary between the bladder and surrounding tissues clearer. After these two steps, an initial bladder image is generated. Compared to the original image, noise is significantly reduced, and the bladder's outline and internal structure are easier to observe.

[0029] The initial bladder image is segmented into bladder regions to generate a target bladder image, which contains only the bladder region. To more accurately analyze bladder morphology, the bladder region needs to be segmented separately from the initial bladder image. A deep learning-based U-Net network model is used for bladder region segmentation. The U-Net network has an encoder and decoder structure; the encoder extracts image features, and the decoder uses these features for image reconstruction and segmentation. The initial bladder image is input into the U-Net network. After training, the network learns the morphological features of the bladder, accurately identifying the bladder region and segmenting it from the entire image to generate a target bladder image containing only the bladder region. This eliminates interference from surrounding tissues, providing a foundation for more accurate subsequent assessment of bladder morphology.

[0030] Image quality assessment processing is performed on the target bladder image to generate quality defect information. This quality defect information characterizes quality problems such as image blurring and artifacts, as well as their location and type. Although the image quality is improved after the previous processing, some quality problems may still exist. A specialized image quality assessment algorithm is used to evaluate the target bladder image. This algorithm analyzes each pixel and its neighborhood to detect the presence of quality problems such as blurring and artifacts. For example, it determines whether the image is blurry by calculating the gradient information of the image. If the gradient value of a certain area is small, it indicates that the edge of that area is not clear, and there may be a blurring problem. Artifacts are detected by analyzing abnormal changes in pixel values ​​in the image. The algorithm records the location and type of these quality problems and generates quality defect information. Suppose that blurring is found in the upper left corner of the image, which may be caused by uneven scanning angle or pressure of the ultrasound probe in that area. At the same time, there are some artifacts caused by ultrasound reflection in the image. The algorithm accurately records detailed information such as the location, degree of blurring, and type of artifacts in these blurred areas and artifacts.

[0031] The system processes quality defect information to generate repair strategy parameters, including blurring parameters, denoising parameters, artifact repair parameters, and image enhancement parameters. Based on the generated quality defect information, the system automatically calculates the repair strategy parameters. For blurring issues, the system determines blurring parameters based on factors such as the size and severity of the blurred region. For example, if the blurred region is small and the blurriness is mild, a smaller Gaussian kernel parameter might be set for further filtering to improve the blur without losing too much image detail. For artifact issues, artifact repair parameters are determined based on the type and characteristics of the artifacts, such as using specific artifact removal algorithms and setting appropriate thresholds. For the overall image quality, denoising and image enhancement parameters are also determined to further optimize the image quality. Finally, a repair strategy parameter set containing blurring, denoising, artifact repair, and image enhancement parameters is generated.

[0032] Based on the repair strategy parameters, the target bladder image is processed to generate a target ultrasound image sequence. According to the blurring parameters, an appropriate filtering algorithm is used to process the blurred areas, making them clearer. Based on the denoising parameters, the image is denoised again to further reduce the impact of noise. According to the artifact repair parameters, corresponding algorithms are used to remove artifacts from the image. According to the image enhancement parameters, image enhancement algorithms are used to optimize the image and improve its overall quality. After this series of processing steps, the processed images are arranged in chronological order or a specific logical order to generate the target ultrasound image sequence. This image sequence more accurately reflects the true morphology of the bladder, providing high-quality data support for subsequent operations such as bladder 3D morphological assessment and the generation of predictive CT images.

[0033] S103, Process the target ultrasound image sequence information to generate a target enhanced ultrasound image.

[0034] In one implementation, the target ultrasound image sequence information is normalized to generate a normalized image sequence. An ultrasound image at a specific moment is selected as an example from the target ultrasound image sequence information obtained in the previous steps. The pixel values ​​of this frame may range from 0 to 255, but differences in brightness and contrast exist between different images, which is detrimental to subsequent unified processing. Through normalization, the image pixel values ​​are mapped to the standard range of 0-1. The normalization formula used is:

[0035] (I(x,y) is the pixel value of the original image at coordinates (x,y), I min and I maxThese are the minimum and maximum pixel values ​​in the original image, respectively. This process is performed on each image in the target ultrasound image sequence to obtain a normalized image sequence. After normalization, the pixel value range of all images is uniform, providing a consistent basis for subsequent processing.

[0036] Contrast enhancement is performed on a normalized image sequence based on a preset enhancement algorithm to generate a contrast-enhanced image sequence. The preset contrast enhancement algorithm formula is used, such as the improved histogram equalization algorithm CLAHE (Constrained Contrast Adaptive Histogram Equalization). The CLAHE algorithm divides the image into multiple small blocks, performs histogram equalization on each block separately, and simultaneously limits the degree of contrast enhancement to avoid excessive noise amplification. The CLAHE algorithm was applied to the patient's normalized image sequence, with a contrast constraint parameter set to 40 and a block size of 8x8 pixels. After processing, the contrast between the bladder tissue and surrounding tissues in the image was significantly enhanced, and structures that were previously difficult to distinguish under low contrast became clearer, generating a contrast-enhanced image sequence.

[0037] Noise filtering is applied to the contrast-enhanced image sequence to generate a low-noise image sequence. Edge sharpening is then applied to this low-noise image sequence to generate a sharpened edge image sequence. Since noise is inevitably introduced during ultrasound imaging, it may become more pronounced after contrast enhancement. Therefore, noise filtering is performed on the contrast-enhanced image sequence. A Gaussian filtering algorithm is used, with the standard deviation of the Gaussian kernel set to 1.5. Gaussian filtering effectively smooths the image and removes most of the noise by weighted averaging of each pixel and its neighboring pixels. After processing, noise in the image is suppressed, generating a low-noise image sequence. At this point, the outline and internal structure of the bladder remain clear, and no important information is lost due to filtering. To highlight the edge information of the bladder, edge sharpening is applied to the low-noise image sequence. The Laplacian operator is used for edge sharpening. The Laplacian operator can detect the second derivative in the image, highlighting edges and details. The Laplacian operator is applied to each image in the low-noise image sequence, and then the result is added to the original image to achieve edge sharpening. After processing, the edges of the bladder become sharper, and the boundary with surrounding tissues becomes clearer, generating a sequence of sharpened edge images. This allows for more accurate identification of the bladder's boundaries during subsequent assessments of bladder morphology.

[0038] A detail-enhanced image sequence is generated by convolving a Gaussian high-pass filter with a sharpened edge image sequence. To further enhance the detail in bladder images, a Gaussian high-pass filter is used to convolve with the sharpened edge image sequence, highlighting high-frequency details. The parameters of the Gaussian high-pass filter are set to effectively enhance the minute structures and textures in the bladder images. Convolution is performed on each image in the sharpened edge image sequence. After processing, the minute structures inside the bladder, such as mucosal folds, become more clearly visible, generating a detail-enhanced image sequence.

[0039] A target-enhanced ultrasound image is generated by fusing a sequence of detail-enhanced images according to a preset image fusion rule. The preset fusion rule can be based on a weighted average method, where multiple images in the detail-enhanced image sequence are weighted according to certain weights. For example, three adjacent detail-enhanced images can be assigned weights of 0.3, 0.4, and 0.3 respectively, and then a weighted average is performed. This fusion process integrates the advantageous information from different images to generate the target-enhanced ultrasound image. This image not only has clear edges and rich details but also low noise levels, more accurately reflecting the true morphology of the bladder and providing high-quality image data for subsequent analysis based on a target bladder morphology assessment model.

[0040] S104, based on the target bladder morphology assessment model, processes the patient's urinary system physiological parameter information to generate target physiological state characteristics.

[0041] In one implementation, a target bladder morphology assessment model is used to classify and process the patient's urinary system physiological parameters, generating abnormal and normal parameter information. The hospital obtains the patient's urinary system physiological parameters, including renal function indicators such as creatinine and blood urea nitrogen from blood tests, routine urinalysis indicators such as red blood cell, white blood cell, and protein content from urine tests, and bladder pressure data during bladder filling and emptying from bladder pressure measurements. These parameters are then input into a pre-trained target bladder morphology assessment model. The model classifies the parameters based on built-in classification rules and learned normal range knowledge. For example, normal creatinine levels are 53-106 μmol / L for men and 44-97 μmol / L for women. If a patient's creatinine value exceeds this range, the model classifies it as an abnormal parameter. Similarly, normal urine should contain very few red blood cells, and white blood cell counts also have a normal range; values ​​exceeding these ranges are considered abnormal. Thus, the model divides the patient's urinary system physiological parameters into two groups: abnormal and normal parameters.

[0042] Based on the target bladder morphology assessment model, feature extraction is performed on abnormal and normal parameter information to generate initial physiological state features. For the abnormal and normal parameter information obtained in the previous step, the target bladder morphology assessment model further extracts features. Taking creatinine as an example, the model extracts features such as the specific value of creatinine, the degree of deviation from the normal range, and the recent trend of changes in the measured value; for white blood cells in urine, features such as the number of white blood cells and the fluctuation of the white blood cell count are extracted. By performing similar processing on all parameters, the key features of each parameter are extracted and combined to form the initial physiological state features. These features reflect the physiological condition of the patient's urinary system from different perspectives. For example, a large deviation of creatinine from the normal range may indicate a significant problem with kidney function, affecting the normal metabolic and excretory functions of the bladder.

[0043] The initial physiological state characteristics are normalized to generate standardized feature vectors. Because the dimensions and value ranges of different parameters vary significantly, the initial physiological state characteristics are normalized to facilitate subsequent unified analysis. For example, creatinine values ​​may range from tens to hundreds, while the number of white blood cells in urine is usually in the single digits to tens; directly comparing the two values ​​is meaningless. A normalization formula is used, such as... (x is the original eigenvalue, x min and x max The minimum and maximum values ​​in the training sample set are used to map the values ​​of all initial physiological state features to the standard range of (0-1). After normalization, all features are on the same scale, forming a standardized feature vector, making different features comparable and providing a unified data foundation for subsequent analysis.

[0044] The initial physiological state characteristics are analyzed using correlation analysis to generate information on the relationships between parameters. The target bladder morphology assessment model performs correlation analysis on standardized feature vectors to explore the intrinsic connections between different parameters. For example, it was found that an increase in white blood cell count in the patient's urine was accompanied by abnormal fluctuations in bladder pressure during urination. The model, through data analysis, concluded that there is a positive correlation between the two, meaning that an increase in white blood cells may affect bladder emptying function, thereby causing abnormal bladder pressure. The model may also find correlations between renal function indicators and routine urinalysis indicators; for example, when creatinine is elevated, the protein content in urine also tends to increase. By comprehensively analyzing the correlations between various parameters, information on the relationships between parameters is generated. This information helps to gain a deeper understanding of the overall condition of the patient's urinary system and the interaction mechanisms between various parameters.

[0045] The model fuses the correlation information between parameters and the standardized feature vectors to generate target physiological state features. It comprehensively considers the standardized values ​​of each parameter and their correlations. For example, if a strong correlation is found between several parameters and their standardized values ​​deviate from the normal range, the model will assign greater weight to these parameters. Through a specific fusion algorithm, all information is integrated to generate the target physiological state features. These features comprehensively reflect the overall physiological state of the patient's urinary system, including both abnormalities in individual parameters and the interactions between them. This provides crucial information for subsequent 3D bladder morphology assessment and CT image prediction based on these features, helping doctors more accurately determine the patient's current bladder physiological state and providing strong support for radiotherapy planning.

[0046] S105 processes the target physiological characteristics, enhanced ultrasound images, and functional imaging data based on the target bladder morphology assessment model to generate a predicted CT image of the target patient.

[0047] In one implementation, a target physiological state feature encoding process is performed on the target bladder morphology assessment model to generate physiological state feature codes. Assuming a patient with a pelvic tumor, after processing their urinary system physiological parameters, target physiological state features are obtained. These features include information such as abnormal renal function indicators (e.g., abnormal creatinine and blood urea nitrogen), abnormal urinalysis indicators (abnormal red blood cell and white blood cell counts), and abnormal bladder pressure. The target bladder morphology assessment model performs feature encoding on these target physiological state features. The model converts each feature into a specific encoding form. For example, for the feature of creatinine values ​​exceeding the normal range, the model encodes it into a set of numbers based on the degree of deviation and trend. This set of numbers not only represents the abnormality of creatinine but also contains potential correlation information with other physiological state features. By encoding all target physiological state features, physiological state feature codes are generated. These codes store the patient's urinary system physiological information in a compact and easily processed manner by the model.

[0048] Feature extraction processing was performed on the target enhanced ultrasound image to generate ultrasound image features. The target enhanced ultrasound image was obtained after a series of processing steps on the patient's bladder ultrasound images. The target bladder morphology assessment model extracted features from this image. The model identifies features such as bladder wall thickness, homogeneity of internal bladder echoes, and bladder shape and contour. For example, if the model detects localized bladder wall thickening, it extracts the location and extent of the thickening as features; for areas with heterogeneous internal bladder echoes, it extracts features such as their extent and echo intensity differences. Through these operations, key information about bladder morphology and structure in the target enhanced ultrasound image is transformed into ultrasound image features that can be understood and processed by the model. Feature mining processing was performed on functional imaging data to generate functional imaging features. The patient also underwent a PET-CT scan, obtaining functional imaging data. The target bladder morphology assessment model performed feature mining on this functional imaging data. PET-CT images reflect the metabolic status of bladder tissue, and the model analyzes features such as the distribution and intensity of metabolically active areas. If a region within the bladder is found to have significantly higher metabolism than surrounding tissues, the model extracts information such as the location of that region and the ratio of its metabolic activity intensity to that of normal tissue as functional imaging features. These characteristics provide functional information about the bladder at the metabolic level, helping to provide a more comprehensive understanding of the bladder's condition.

[0049] Physiological state feature encoding, ultrasound image features, and functional imaging features are fused to generate a fused feature vector. During the fusion process, the model assigns different weights to different types of features based on their importance and relevance. For example, considering the significant potential impact of the patient's current renal function indicators on bladder morphology, physiological state feature encoding may be given a higher weight; similarly, the key morphological feature of bladder wall thickening in ultrasound images is also assigned a relatively high weight. Then, a specific fusion algorithm combines these features to generate the fused feature vector. This vector integrates information from physiological, morphological, and functional aspects, comprehensively describing the state of the patient's bladder.

[0050] The fused feature vector undergoes dimensionality reduction and reconstruction to generate initial predicted features. These initial predicted features are then subjected to convolution and pooling to generate the target predicted features. While the fused feature vector contains a wealth of information, it may contain redundant and complex high-dimensional data, which is detrimental to subsequent processing. Therefore, the model performs dimensionality reduction and reconstruction on the fused feature vector. The model employs dimensionality reduction algorithms such as Principal Component Analysis (PCA) to remove redundant information from the data, extract principal components, and transform the high-dimensional fused feature vector into low-dimensional initial predicted features. During dimensionality reduction, the model uses reconstruction algorithms to retain as much key information as possible, ensuring that the initial predicted features not only reflect the main content of the original features but also have a more concise form, facilitating subsequent convolution and pooling operations.

[0051] The model performs convolution and pooling operations on the initial predicted features. Convolution operations extract local features by sliding different convolution kernels across the initial predicted features. For example, convolution kernels of specific sizes and weights are used to detect local information related to bladder morphology, such as edges and textures, highlighting key morphological features. Pooling operations further compress the convolutional features, reducing data dimensionality while retaining important features. For instance, max pooling is used, selecting the maximum value within a certain region as the pooling result, reducing the amount of data while enhancing the representativeness of the features. After multiple rounds of convolution and pooling, target predicted features are generated. These features are more focused on the key morphological and functional features of the bladder, providing more accurate information for generating predicted CT images.

[0052] The model performs upsampling and deconvolution on the target predicted features to generate a predicted CT image of the target patient. Upsampling increases the data resolution through interpolation and other methods, gradually bringing the feature map size closer to the size of the actual CT image. Deconvolution, the inverse of convolution, reconstructs the target predicted features into an image with richer details. In this process, the model progressively converts the target predicted features into a predicted CT image based on learned CT image features and patterns. For example, the model supplements missing details based on learning from a large number of CT images during training, making the predicted CT image closer to the actual CT image in terms of morphology, structure, and function, thus generating a predicted CT image of the target patient. This predicted CT image provides crucial data support for subsequent comparison with simulated localization CT images and evaluation of the bladder's three-dimensional morphology, helping doctors more accurately determine the actual state of the patient's bladder and develop more precise radiotherapy plans.

[0053] S106, Based on the target bladder morphology assessment model and combined with bladder dynamic image information, the predicted CT image and simulated positioning CT image information of the target patient are processed to generate bladder three-dimensional morphology assessment results.

[0054] In one implementation, a target bladder morphology assessment model is used to perform feature alignment processing on the predicted CT image and simulated localization CT image information of a target patient, generating aligned image features. Assume a patient with pelvic tumors has a simulated localization CT image acquired during the radiotherapy localization phase, and a predicted CT image generated based on the target bladder morphology assessment model is obtained through previous steps. The target bladder morphology assessment model performs feature alignment processing on these two images. The model first identifies key feature points of the bladder in both images, such as key outline points and landmark features of internal structures. Then, an image registration algorithm is used to match and align the features of the predicted CT image with those of the simulated localization CT image. For example, the edges, corners, and other key parts of the bladder in the two images are made to overlap as much as possible in spatial location. In this way, aligned image features are generated, laying the foundation for more accurate comparison and analysis of the differences between the two images, ensuring that bladder morphology is evaluated under the same standard.

[0055] The bladder dynamic images are preprocessed using a target bladder morphology assessment model to generate bladder wall elastic features. Real-time ultrasound dynamic monitoring was used to acquire dynamic bladder images of the patient. The target bladder morphology assessment model preprocesses the dynamic bladder images to generate bladder wall elastic features. The model analyzes the deformation of the bladder wall during filling and emptying. By tracking the positional changes of specific points on the bladder wall at different times, the strain and stress of the bladder wall are calculated. For example, during bladder filling, the degree of stretching in a certain region of the bladder wall is observed, and the elastic modulus of that region is calculated based on mechanical principles, thus serving as part of the bladder wall elastic features. These elastic features reflect the elastic condition of the bladder wall and are of great significance for assessing bladder function and morphological changes.

[0056] The target bladder morphology assessment model processes the feature vectors of dynamic bladder images with features from aligned predicted CT images and simulated localization CT images to generate fused features. The bladder wall elasticity feature vectors are then fused with the aligned image features. The model weights these features based on their importance and relevance. For example, areas of abnormal elasticity in the bladder wall elasticity features are given higher weights, as this may indicate bladder lesions or functional abnormalities; morphological features such as bladder shape and size in the predicted and simulated localization CT images are also weighted according to their importance in bladder morphology assessment.

[0057] Different types of features have varying importance in reflecting bladder status; therefore, the first step in a specific fusion algorithm is to determine the weight of each feature. For example, in a patient with pelvic tumors, abnormal renal function indicators (such as creatinine and blood urea nitrogen) can significantly impact bladder metabolism and excretion. During fusion, the weight of physiological state features related to renal function will be increased accordingly. Similarly, for key morphological features in ultrasound images, such as bladder wall thickening or abnormal echoes, the weight of ultrasound image features will also be increased because they may be important indicators for diagnosing bladder lesions. In dynamic bladder images, dynamic features such as bladder wall peristalsis frequency and emptying velocity will also be assigned higher weights if abnormal changes occur. The method for determining weights can be based on statistical analysis of large amounts of clinical data, or optimized using expert experience and machine learning algorithms to ensure that the impact of important features on the fusion results is highlighted.

[0058] After determining the weights, the feature fusion stage begins. Assuming the patient's physiological state feature codes, ultrasound image features, functional imaging features, and bladder dynamic image feature vectors have been extracted, a weighted summation fusion method is used to linearly combine these features according to the determined weights. For example, if the physiological state feature codes are represented by vector A, ultrasound image features by vector B, functional imaging features by vector C, and bladder dynamic image feature vector by vector D, with corresponding weights ω1, ω2, ω3, and ω4 respectively, then the fused feature vector F can be expressed as F = ω1A + ω2B + ω3C + ω4D. In this way, features from different modalities are organically combined, enabling the fused features to comprehensively reflect multiple aspects of bladder information.

[0059] This fusion process integrates and complements information on bladder morphology, function, and dynamic changes. Physiological state features reflect the overall physiological condition of the patient's urinary system; for example, abnormal kidney function may indicate impaired bladder excretion. Ultrasound image features display the immediate morphological structure of the bladder, such as bladder wall thickness and internal echogenicity, which can be used to determine the presence of morphological abnormalities. Functional imaging features provide functional information about the bladder from a metabolic perspective, enabling the identification of potential lesion areas. Bladder dynamic imaging features demonstrate changes in the bladder under different physiological states, such as bladder wall elasticity and peristaltic function. Through fusion, these information complement each other, forming a comprehensive description of the bladder's condition. For example, when ultrasound images show localized thickening of the bladder wall, functional imaging features detect metabolic abnormalities in this thickened area, and bladder dynamic imaging shows abnormal peristalsis in this area during urination, combining this information allows the fused features to more accurately reflect potential bladder lesions and their impact on bladder function and dynamic morphological changes.

[0060] Following the steps outlined above, a comprehensive fusion feature is ultimately generated. This fusion feature is no longer information from a single modality, but rather a complex feature vector encompassing multiple aspects of the bladder's state. It deeply fuses information from different sources, providing a rich and comprehensive data foundation for subsequent bladder 3D morphological assessment. In subsequent processing, based on this fusion feature, the model can more accurately perform feature enhancement, mapping, and classification, thereby obtaining more precise bladder 3D morphological assessment results. This assists physicians in gaining a more comprehensive understanding of the patient's bladder status and developing a more appropriate radiotherapy plan.

[0061] Feature enhancement processing is performed on the fused features to generate a fused enhanced feature sequence. This enhancement process highlights key information. The model employs techniques from Convolutional Neural Networks (CNNs), such as using specific convolutional kernels. These kernels are designed to extract features related to bladder morphology and function, such as edge features and texture features. Through multiple convolutions and nonlinear activation functions, key information in the fused features is enhanced while noise and irrelevant information are suppressed. For example, the contrast and clarity of detailed features of the bladder wall edges and internal structures are improved after convolution. After this series of processing steps, a fused enhanced feature sequence is generated, making the bladder features more prominent and facilitating subsequent analysis and classification.

[0062] The fused and enhanced feature sequences are processed through feature mapping and classification to generate bladder morphology classification information. The model maps the fused and enhanced feature sequences to a specific feature space, where different feature combinations correspond to different bladder morphology categories. Through training, the model learns feature patterns for each category, including normal bladder morphology, bladder morphology with varying degrees of abnormal filling, and bladder morphology with lesions. Then, based on the input fused and enhanced feature sequences, the model uses classification algorithms, such as Support Vector Machines (SVM) or classification layers of neural networks, to determine the current bladder morphology category, generating bladder morphology classification information. For example, it determines whether the bladder is in a normal filling state, an overfilled state, or has morphological abnormalities caused by a tumor.

[0063] The bladder morphology classification information is comprehensively evaluated and quantified to generate a three-dimensional bladder morphology assessment result. The model considers information from each category within the bladder morphology classification information, as well as other relevant information obtained during previous processing, such as bladder wall elasticity characteristics and the degree of difference between predicted CT images and simulated localization CT images. The three-dimensional morphology of the bladder is comprehensively evaluated using preset evaluation rules and quantitative indicators. For example, an evaluation score is set, calculated based on factors such as the similarity of the bladder morphology to a normal morphology and the degree of abnormality in bladder wall elasticity. A higher score indicates a bladder morphology closer to normal, with a lower risk of adverse effects during radiotherapy; a lower score indicates a more severe bladder morphological abnormality, with a greater likelihood of needing to adjust the radiotherapy plan. The final three-dimensional bladder morphology assessment result provides doctors with accurate and quantitative evidence for developing radiotherapy plans, helping them determine whether adjustments to parameters such as the target area and dose are necessary to reduce radiotherapy side effects caused by changes in bladder filling.

[0064] The server acquires multi-source data, including bladder ultrasound images, simulated localization CT images, urinary system physiological parameters, functional imaging data, and bladder dynamic imaging information, providing a comprehensive basis for assessment. Bladder ultrasound images undergo preprocessing and enhancement to improve image quality and highlight bladder structural information. A model is used to process the multi-source data. This includes classifying, extracting, normalizing, performing correlation analysis and fusion of urinary system physiological parameters to generate target physiological state features; and mining, fusing, and processing features from ultrasound images and functional imaging data in multiple steps to generate predicted CT images. Finally, combined with bladder dynamic images, the predicted CT images and simulated localization CT images undergo multi-stage processing to complete the three-dimensional morphological assessment of the bladder. The entire process utilizes deep learning and multimodal data fusion technology to transform cross-modal assessment into same-modal assessment, significantly improving the intuitiveness and accuracy of the assessment. This effectively assists physicians in developing precise radiotherapy plans, reducing radiotherapy side effects caused by changes in bladder filling, and providing strong support for pelvic tumor radiotherapy.

[0065] In one implementation, such as Figure 2 As shown, this application also provides a transmembrane assessment device for the three-dimensional morphology of the bladder, comprising:

[0066] The acquisition module 201 is used to acquire the target patient's bladder ultrasound image information, the target patient's simulated positioning CT image information, the target patient's urinary system physiological parameter information, functional imaging data, and bladder dynamic image information. Among them, the bladder dynamic image information is used to characterize the morphological changes of the bladder under different physiological states.

[0067] The processing module 202 is used to preprocess the bladder ultrasound image information of the target patient to generate target ultrasound image sequence information; process the target ultrasound image sequence information to generate target enhanced ultrasound images; process the patient's urinary system physiological parameter information based on the target bladder morphology assessment model to generate target physiological state characteristics; process the target physiological state characteristics, target enhanced ultrasound images, and functional imaging data based on the target bladder morphology assessment model to generate predicted CT images of the target patient; and process the predicted CT images and simulated positioning CT images of the target patient based on the target bladder morphology assessment model combined with bladder dynamic image information to generate bladder three-dimensional morphology assessment results.

[0068] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of the transmembrane assessment method, electronic device, electronic device, and readable storage medium for evaluating the three-dimensional morphology of the bladder are basically similar to the above-described embodiments of the transmembrane assessment method for evaluating the three-dimensional morphology of the bladder, and therefore are described relatively simply. Relevant parts can be referred to in the descriptions of the above-described embodiments of the transmembrane assessment method for evaluating the three-dimensional morphology of the bladder.

Claims

1. A transmembrane assessment method for the three-dimensional morphology of the bladder, characterized in that, include: The study aims to acquire bladder ultrasound images, simulated localization CT images, urinary system physiological parameters, functional imaging data, and bladder dynamic images of the target patient. The bladder dynamic images are used to characterize the morphological changes of the bladder under different physiological states. Preprocess the bladder ultrasound image information of the target patient to generate the target ultrasound image sequence information; The target ultrasound image sequence information is processed to generate an enhanced ultrasound image of the target; Based on the target bladder morphology assessment model, the patient's urinary system physiological parameter information is processed to generate target physiological state characteristics; Based on the target bladder morphology assessment model, the target physiological state characteristics, target enhanced ultrasound images, and functional imaging data are processed to generate predictive CT images of the target patient; Based on the target bladder morphology assessment model, combined with bladder dynamic image information, the predicted CT images and simulated localization CT images of the target patient are processed to generate bladder three-dimensional morphology assessment results.

2. The method as described in claim 1, characterized in that, Preprocessing of bladder ultrasound images of the target patient generates target ultrasound image sequence information, including: The bladder ultrasound images of the target patient are denoised and contrast-enhanced to generate initial bladder images; The initial bladder image is segmented into bladder regions to generate a target bladder image, wherein the target bladder image is an image containing only the bladder region; Image quality assessment processing is performed on the target bladder image to generate quality defect information, wherein the quality defect information is used to characterize quality problems including image blurring and artifacts, as well as their location and type; The quality defect information is processed to generate repair strategy parameters, which include blur processing parameters, denoising parameters, artifact repair parameters, and image enhancement parameters. The target bladder image is processed based on the repair strategy parameters to generate the target ultrasound image sequence information.

3. The method as described in claim 2, characterized in that, The target ultrasound image sequence information is processed to generate an enhanced ultrasound image of the target, including: The target ultrasound image sequence information is normalized to generate a normalized image sequence; The normalized image sequence is subjected to contrast enhancement processing based on a preset enhancement algorithm to generate an enhanced contrast image sequence. Noise filtering is applied to the contrast-enhanced image sequence to generate a low-noise image sequence. Edge sharpening is performed on a low-noise image sequence to generate a sharpened edge image sequence. A detail-enhanced image sequence is generated by convolution processing with a Gaussian high-pass filter and a sharpened edge image sequence. Based on preset image fusion rules, the image sequence for detail enhancement is fused to generate the target enhanced ultrasound image.

4. The method as described in claim 1, characterized in that, Obtain a target bladder morphology assessment model, including: Obtain a training sample set and a preset bladder morphology assessment model. The training sample set includes multimodal data and corresponding bladder three-dimensional morphology assessment results. The number of features in each modality of the training sample set is counted, and a sampling ratio is generated based on the feature distribution balance. Based on the sampling ratio, the training sample set is subjected to multimodal hierarchical sampling processing to generate a preset number of sampling feature combinations; Statistical tests are performed based on any key feature combined with each sampling feature to divide the data into a response group and a poor response group. Each group contains a preset number of data samples, and at least one sample carries lesion identification information. The pre-defined bladder morphology assessment model is iteratively trained based on the response group and the poor response group. The model parameters are optimized through cross-validation to generate the trained model and performance indicators. If the lesion identification information in the training results is recognized by the model as a key feature for morphological assessment, then the trained model will be used as the target bladder morphological assessment model.

5. The method as described in claim 1, characterized in that, Based on the target bladder morphology assessment model, the patient's urinary system physiological parameters are processed to generate target physiological state characteristics, including: Based on the target bladder morphology assessment model, the patient's urinary system physiological parameters are classified and processed to generate abnormal and normal parameter information. Based on the target bladder morphology assessment model, feature extraction processing is performed on abnormal and normal parameter information to generate initial physiological state features; The initial physiological state characteristics are normalized to generate standardized feature vectors; The initial physiological state characteristics are analyzed to generate correlation information between parameters. The correlation information between parameters and the standardized feature vector are fused to generate the target physiological state features.

6. The method as described in claim 1, characterized in that, Based on the target bladder morphology assessment model, the target physiological state characteristics, target enhanced ultrasound images, and functional imaging data are processed to generate predictive CT images of the target patient, including: Based on the target bladder morphology assessment model, feature encoding processing is performed on the target physiological state characteristics to generate physiological state feature codes; Feature extraction processing is performed on the enhanced ultrasound image of the target to generate ultrasound image features; Feature mining is performed on functional image data to generate functional image features; Physiological state feature encoding, ultrasound image features, and functional imaging features are fused to generate a fused feature vector. The fused feature vector is subjected to dimensionality reduction and reconstruction to generate initial prediction features; The initial predicted features are processed by convolution and pooling to generate the target predicted features; The predicted features of the target are upsampled and deconvolutioned to generate the predicted CT image of the target patient.

7. The method as described in claim 6, characterized in that, Based on the target bladder morphology assessment model and combined with dynamic bladder image information, the predicted CT images and simulated localization CT images of the target patient are processed to generate a three-dimensional bladder morphology assessment result, including: Based on the target bladder morphology assessment model, feature alignment processing is performed on the predicted CT images and simulated localization CT images of the target patient to generate aligned image features; Based on the target bladder morphology assessment model, the dynamic images of the bladder are preprocessed to generate bladder wall elasticity features; Based on the target bladder morphology assessment model, the feature vectors of the bladder dynamic image are processed with the features of the aligned predicted CT image and the simulated localization CT image to generate fused features; The fused features are enhanced to generate a fused enhanced feature sequence; The fused and enhanced feature sequences are subjected to feature mapping and classification processing to generate bladder morphology classification information; The bladder morphology classification information is comprehensively evaluated and quantified to generate a three-dimensional bladder morphology evaluation result.

8. A transmembrane assessment device for three-dimensional bladder morphology, characterized in that, The device includes: The acquisition module is used to acquire bladder ultrasound image information, simulated positioning CT image information, urinary system physiological parameter information, functional imaging data, and bladder dynamic image information of the target patient. Among them, the bladder dynamic image information is used to characterize the morphological changes of the bladder under different physiological states. The processing module is used to preprocess the bladder ultrasound image information of the target patient to generate target ultrasound image sequence information; process the target ultrasound image sequence information to generate target enhanced ultrasound images; process the patient's urinary system physiological parameters based on the target bladder morphology assessment model to generate target physiological state characteristics; process the target physiological state characteristics, target enhanced ultrasound images, and functional imaging data based on the target bladder morphology assessment model to generate predicted CT images of the target patient; and process the predicted CT images and simulated positioning CT images of the target patient based on the target bladder morphology assessment model combined with bladder dynamic image information to generate bladder three-dimensional morphology assessment results.

9. An electronic device, characterized in that, include: First processor; and memory for storing executable instructions of the first processor; The first processor is configured to execute the transmembrane assessment method for the three-dimensional morphology of the bladder according to any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the second processor, it implements the transmembrane assessment method for the three-dimensional morphology of the bladder as described in any one of claims 1 to 7.

Images