PCI (Peripheral Component Interconnect) drug balloon postoperative effect prediction method, device and equipment based on coronary DSA (Differential Scalcification Analysis) and angiography calcification grading and storage medium

Abstract

The invention discloses a coronary DSA and angiography calcification grading-based PCI drug balloon postoperative effect prediction method, device and equipment and a storage medium, and the method comprises the steps: firstly obtaining a PCI drug balloon preoperative DSA image of a to-be-evaluated patient, the DSA image comprising calcified plaque information of a target lesion segment; on the basis of the obtained DSA image, blood vessel morphological parameters are extracted, angiography calcification grading is carried out on calcified plaques of a target lesion segment, and finally the blood vessel morphological parameters of a patient, an angiography calcification grading result, clinical data and instrument data of a to-be-used drug balloon serve as feature data for prediction; and finally, inputting the feature data into a pre-trained machine learning model, and outputting a PCI drug balloon postoperative prediction effect. According to the invention, doctors can be assisted to formulate a more accurate treatment scheme, so that the prognosis of patients is improved.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, and in particular to a method, device, equipment, and storage medium for predicting the postoperative effects of PCI drug balloon based on coronary DSA and angiographic calcification grading. Background Technology

[0002] Coronary atherosclerotic heart disease (CAD) is a group of diseases caused by atherosclerotic plaques and secondary thrombosis in the coronary arteries, leading to myocardial ischemia and even necrosis. It is currently the leading cause of death among urban and rural residents in my country. Percutaneous coronary intervention (PCI) is a key treatment for CAD, aiming to restore myocardial blood supply by reconstructing vascular patency. Among the various PCI techniques, drug-coated balloons (DCBs) have become an important choice for treating specific types of coronary artery lesions (such as small vessel disease and in-stent restenosis) due to their advantages, such as eliminating the need for permanent metal stents, effectively reducing restenosis rates, and decreasing the risk of long-term thrombosis. They are widely used in clinical practice.

[0003] However, accurately predicting the outcome of DCB treatment remains a challenge in clinical decision-making. Currently, physicians primarily rely on preoperative vascular morphological characteristics (such as the degree of stenosis and lesion length) for experience-based judgment. However, these characteristics cannot fully reflect the complexities of the vessel wall, especially the significant impact of calcified lesions on treatment outcomes, which current methods do not adequately consider, resulting in a need to improve the accuracy and reliability of predictions. Summary of the Invention

[0004] Based on this, the present invention addresses the above-mentioned technical problems by providing a method, device, equipment, and storage medium for predicting the postoperative effects of PCI drug balloon based on coronary DSA and angiographic calcification grading.

[0005] On one hand, this invention provides a method for predicting the postoperative efficacy of PCI drug balloon therapy based on coronary DSA and angiographic calcification grading, the method comprising: Obtain preoperative DSA images of patients undergoing PCI drug-eluting balloon surgery, where the DSA images contain information on calcified plaques in the target lesion segment; Based on the acquired DSA images, vascular morphology parameters are extracted, and calcification of calcified plaques in the target lesion segment is graded by angiography. Finally, the patient's vascular morphology parameters, angiography calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction. The feature data is input into a pre-trained machine learning model, which outputs the predicted effect after PCI drug balloon surgery.

[0006] In one embodiment, the contrast-enhanced calcification grading of the calcified plaque in the target lesion segment includes: Each frame in the DSA image sequence is input into a pre-trained deep learning calcification grading model. The deep learning calcification grading model outputs the corresponding calcification degree prediction result based on a single frame of DSA image. The calcification degree includes four levels: no calcification, mild calcification, moderate calcification, and severe calcification. Based on the prediction results of all frames in the DSA image sequence, patient-level calcification grading features of the target blood vessels of the target patient are aggregated and generated.

[0007] In one embodiment, the DSA images are derived from the left anterior descending artery (LAD), the circumflex artery (LCX), and the right coronary artery (RCA), and the number of samples for each vessel type is evenly distributed.

[0008] In one embodiment, the extraction of vascular morphology parameters based on the acquired DSA images includes: Preprocess the DSA images; Spatial scale calibration of DSA images is performed based on reference objects with known physical dimensions to obtain blood vessel images with real physical units; Select the analysis frame from the DSA image where the target lesion segment is clearly visualized and the blood vessel overlap is minimal; Based on the analysis frame, the centerline of the target blood vessel is extracted and the lumen contour is segmented; Based on the lumen profile, calculate the vascular morphology parameters of the target lesion segment.

[0009] In one embodiment, the vascular morphology parameters include at least a reference vessel diameter, a minimum lumen diameter, and the degree of vascular stenosis.

[0010] In one embodiment, the machine learning model is trained in the following manner: The model was trained using a labeled dataset containing historical patients' DSA images, calcification grading results, clinical data, and actual surgical outcome data. The training objective of the machine learning model was to minimize the mean squared error between the predicted and actual results.

[0011] In one embodiment, the predicted effect includes predicted values ​​of vasodilation gain, postoperative minimum lumen diameter, and / or functional indicators.

[0012] On the other hand, the present invention provides a device for predicting the postoperative efficacy of PCI drug-eluting balloon surgery based on coronary DSA and angiographic calcification grading, the device comprising: The image acquisition module is used to acquire DSA images of patients undergoing PCI drug balloon surgery before evaluation. The DSA images contain information on calcified plaques in the target lesion segment. The feature extraction module is used to extract vascular morphology parameters based on the acquired DSA images, and to perform angiographic calcification grading of calcified plaques in the target lesion segment. Finally, the patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction. The prediction module is used to input feature data into a pre-trained machine learning model and output the prediction effect after PCI drug balloon surgery.

[0013] In another aspect, the present invention provides a computer device, including a memory and a processor, wherein the processor performs the following steps when executing the computer program: Obtain preoperative DSA images of patients undergoing PCI drug-eluting balloon surgery, where the DSA images contain information on calcified plaques in the target lesion segment; Based on the acquired DSA images, vascular morphology parameters are extracted, and calcification of calcified plaques in the target lesion segment is graded by angiography. Finally, the patient's vascular morphology parameters, angiography calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction. The feature data is input into a pre-trained machine learning model, which outputs the predicted effect after PCI drug balloon surgery.

[0014] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, performs the following steps: Obtain preoperative DSA images of patients undergoing PCI drug-eluting balloon surgery, where the DSA images contain information on calcified plaques in the target lesion segment; Based on the acquired DSA images, vascular morphology parameters are extracted, and calcification of calcified plaques in the target lesion segment is graded by angiography. Finally, the patient's vascular morphology parameters, angiography calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction. The feature data is input into a pre-trained machine learning model, which outputs the predicted effect after PCI drug balloon surgery.

[0015] Compared with existing technologies, this invention introduces angiographic calcification grading on the basis of preoperative DSA images and integrates patient clinical data and drug-eluting balloon device parameters to construct a multimodal feature-driven machine learning prediction model. This effectively overcomes the limitations of existing technologies that rely solely on vascular morphology features for empirical judgment, and significantly improves the prediction accuracy of immediate and long-term vascular dilation effects after drug-eluting balloon surgery. As a result, it provides reliable data support for the precise formulation of non-implantable interventional strategies, optimization of device selection, and reduction of intraoperative trial and error. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a method for predicting the postoperative outcome of PCI drug balloon surgery based on coronary DSA and angiographic calcification grading in one embodiment.

[0017] Figure 2 Example of coronary DSA image in one embodiment.

[0018] Figure 3 This is a ROC curve of a deep learning calcification grading model in one embodiment.

[0019] Figure 4 This is a partial example of a labeled dataset used for training a machine learning model in one embodiment. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0021] This invention integrates multimodal data from imaging, clinical practice, and medical devices, and utilizes machine learning models to accurately predict postoperative outcomes, thereby assisting physicians in developing better interventional treatment strategies.

[0022] like Figure 1 As shown, the present invention provides a method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading, comprising the following steps: Step S100: Acquire pre-PCI drug-eluting balloon DSA images of the patient to be evaluated, wherein the DSA images contain calcified plaque information of the target lesion segment. At this stage, the primary focus is on acquiring pre-PCI drug-eluting balloon coronary artery DSA image sequences of the patient to be evaluated (e.g., Figure 2 This ensures that the collected data can comprehensively reflect the condition of the lesion site.

[0023] Step S200: Based on the acquired DSA images, extract vascular morphology parameters and perform angiographic calcification grading of calcified plaques in the target lesion segment. Finally, use the patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used as feature data for prediction.

[0024] This step S200 specifically includes the following sub-steps: Step S210: Based on the acquired DSA images, extract vascular morphological parameters. These parameters include at least the reference vessel diameter, minimum lumen diameter, and degree of vascular stenosis. Specifically, step S210 includes: Preprocessing DSA images, such as temporal denoising and artifact suppression of the raw DSA, can improve image quality and reduce subsequent quantization errors.

[0025] Subsequently, the DSA images are spatially calibrated based on reference objects of known physical dimensions to obtain vascular images with real physical units. For example, spatial scale calibration can be performed based on a calibration ruler or guidewire diameter, converting pixel units into real physical units (such as millimeters) to obtain vascular images with true size information.

[0026] Next, the analysis frame with the clearest visualization of the target lesion segment and the least overlap of blood vessels is selected from the DSA image. That is, from the preprocessed DSA image, the key frame with the clearest visualization of the lesion and the least overlap is automatically / semi-automatically selected from multiple projection positions to ensure the consistency and repeatability of quantification to the greatest extent.

[0027] Based on this, the centerline of the target vessel is extracted and the lumen contour is segmented using the selected analysis frames. Specifically, the centerline of the vessel is extracted and the lumen contour is segmented in the selected frames to provide accurate contour and diameter curves for quantitative coronary angiography (QCA) parameter calculation.

[0028] Finally, based on the lumen profile, the vascular morphology parameters of the target lesion segment are calculated. QCA is performed on the target lesion segment, automatically calculating and manually verifying the preoperative reference vessel diameter (RVD), preoperative minimum lumen diameter (MLD), and preoperative stenosis rate, and determining the vessel to which the lesion belongs (e.g., LAD / LCX / RCA / D1). These quantified values ​​and the lesion vessel type are recorded as a subset of image features input to the machine learning model.

[0029] Step S220 involves performing angiographic calcification grading on the calcified plaques in the target lesion segment based on the acquired DSA images. Step S220 inputs each frame of the DSA image sequence into a pre-trained deep learning calcification grading model. After obtaining prediction results based on all frames in the DSA image sequence, the model aggregates these predictions to generate patient-level calcification grading features for the target blood vessel of the target patient.

[0030] Combination Figure 3As shown, the deep learning calcification grading model predicts the degree of calcification based on the output of a single-frame DSA image. This calcification degree includes four levels: no calcification, mild calcification, moderate calcification, and severe calcification. Specifically, 5000-8000 suitable DSA images containing calcification were collected through manual screening (ensuring uniform distribution of LAD / LCX / RCA categories). Cardiology clinicians were asked to classify and label the images according to no calcification, mild calcification, moderate calcification, and severe calcification. Using single-frame DSA images and the four-level calcification labels as deep learning training data, a multi-classification model based on EfficientNet-B4 was constructed. The dataset was stratified by patient into training / validation / test (6 / 2 / 2). The input image size was 512×512, and image data augmentation techniques such as random rotation, slight perspective transformation, brightness / contrast jitter, and random thin-strip occlusion were optional. The model uses weighted cross-entropy as the primary loss (including label smoothing), with AdamW as the optimizer and cosine annealing or OneCycleLR as the learning rate. Early stopping is enabled (based on the validation set macro-average F1 or weighted K-optimal). After model training, key classification performance metrics such as accuracy, sensitivity, specificity, and F1-score are reported. In practical applications, a complete DSA image sequence of the patient is used as input. The calcification grading model predicts the calcification grading features of the target blood vessel frame by frame in the image sequence.

[0031] Step S230: The patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction.

[0032] In step S230, 32-dimensional feature data for prediction are constructed, including: gender, age, systolic blood pressure, diastolic blood pressure, heart rate, height, weight, BMI, history of diabetes, history of hypertension, history of stroke, smoking status, history of myocardial infarction, history of PCI, history of bypass surgery, family history of heart disease, history of chronic coronary syndrome, diseased vessels, calcification grading characteristics, balloon pre-dilation diameter / length / pressure / time, actual balloon dilation pressure / release time, balloon length / diameter, preoperative reference vessel diameter (RVD), preoperative minimum lumen diameter (MLD), preoperative stenosis rate (%), plaque volume, and vessel deformability rate. This forms 32 preoperative and device parameters for a single patient.

[0033] Step S300: Input the feature data into the pre-trained machine learning model and output the predicted effect after PCI drug balloon surgery.

[0034] In step S300, the machine learning model is trained in the following way: Combination Figure 4As shown, the model was trained using a labeled dataset containing historical patients' DSA images, calcification grading results, clinical data, and actual surgical outcome data; the training objective of the machine learning model was to minimize the mean squared error between the predicted and actual results.

[0035] Specifically, a 32-dimensional input feature dictionary was constructed, including: gender, age, systolic blood pressure, diastolic blood pressure, heart rate, height, weight, BMI, history of diabetes, history of hypertension, history of stroke, smoking status, history of myocardial infarction, history of PCI, history of bypass surgery, family history of heart disease, history of chronic coronary syndrome, diseased vessels, calcification grading characteristics, balloon pre-dilation diameter / length / pressure / time, actual balloon dilation pressure / release time, balloon length / diameter, preoperative reference vessel diameter (RVD), preoperative minimum lumen diameter (MLD), preoperative stenosis rate (%), plaque volume, and vessel deformation rate. Units and value ranges were standardized, and missing values ​​were handled using multiple imputation or the model's built-in missing value mechanism. Vascular dilation gain was defined as postoperative MLD minus preoperative MLD (or an equivalent validated standard definition). If necessary, postoperative MLD was remeasured according to the same quantification protocol to ensure label consistency. Outliers exceeding medically reasonable ranges were removed, and the reasons for removal were recorded for future reference.

[0036] Categorical variables (gender, lesion vessels, medical history, and behavioral variables, etc.) are encoded using one-hot encoding; continuous variables retain their original scale to facilitate decision tree model splitting. A 7:1:2 training / validation / independent testing partitioning is performed along the patient dimension to avoid information leakage caused by the same patient crossing sets; simultaneously, a K-fold cross-validation scheme is constructed for robust evaluation.

[0037] Using immediate and long-term vasodilation gains as training objectives, XGBoost / LightGBM / CatBoost models were trained for machine learning regression models. Hyperparameters such as learning rate, maximum depth, subsample rate, column sampling rate, and L1 / L2 regularization were adjusted through Bayesian optimization or grid search. Early stopping was achieved by minimizing the validation set RMSE to prevent overfitting. Evaluation metrics such as MAE, RMSE, R², and 95% confidence intervals were reported on the independent test set. The marginal contribution of each feature to the prediction was quantified based on SHAP values ​​or feature importance ranking, outputting the global importance ranking and individual-level decision paths. Partial dependency plots were generated for key variables such as RVD, MLD, stenosis rate, and balloon pressure / time to verify whether the model conforms to known physiological and drug delivery mechanisms.

[0038] In clinical use, 32 preoperative and device parameters for a single patient are input, and the system automatically performs standardization, coding, and model inference, outputting predicted dilation gain and prediction range. Simultaneously, online quality control is performed (input range verification, image quantification confidence level prompts), providing retest / review suggestions when the quantification confidence level is low or the input exceeds the limits. Furthermore, when used in conjunction with stenosis assessment software (FFR), it can provide real-time postoperative immediate and long-term FFR changes for patients under different balloon models, lengths, and diameters, providing comprehensive planning and real-time guidance for drug-eluting balloon implantation surgery. In other words, the predicted effects include vasodilation gain, postoperative minimum lumen diameter, and / or predicted values ​​of functional indicators.

[0039] Ultimately, the system presents these predictions to doctors in the form of visual reports, helping them to accurately assess the therapeutic effect of drug-eluting balloons before surgery, thereby optimizing device selection and developing more reliable interventional strategies.

[0040] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated in this document, there is no strict order in which these steps are executed; they can be performed in other orders. Furthermore, Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0041] In one embodiment, the present invention provides a device for predicting the postoperative effect of PCI drug-eluting balloon based on coronary DSA and angiographic calcification grading, comprising: an image acquisition module, a feature extraction module, and a prediction module, wherein: The image acquisition module is used to acquire preoperative DSA images of patients undergoing PCI drug-eluting balloon surgery. The DSA images contain information on calcified plaques in the target lesion segment.

[0042] The feature extraction module is used to extract vascular morphology parameters based on the acquired DSA images, and to perform angiographic calcification grading of calcified plaques in the target lesion segment. Finally, the patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction.

[0043] The prediction module is used to input feature data into a pre-trained machine learning model and output the prediction results after PCI drug balloon surgery.

[0044] Specific limitations regarding the PCI drug-eluting balloon post-procedure outcome prediction device based on coronary DSA and angiographic calcification grading can be found in the limitations of the PCI drug-eluting balloon post-procedure outcome prediction method based on coronary DSA and angiographic calcification grading mentioned above, and will not be repeated here. Each module in the aforementioned PCI drug-eluting balloon post-procedure outcome prediction device based on coronary DSA and angiographic calcification grading can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0045] In one embodiment, a computer device is provided, which may be a terminal, including a processor, memory, network interface, display screen, and input device connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for predicting the postoperative effects of PCI drug-eluting balloon surgery based on coronary DSA and angiographic calcification grading. The display screen of the computer device may be a liquid crystal display (LCD) or an e-ink display. The input device of the computer device may be a touch layer covering the display screen, or buttons, a trackball, or a touchpad located on the casing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0046] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps: Step S100: Obtain preoperative DSA images of the patient to be evaluated via PCI drug balloon, wherein the DSA images contain information on calcified plaques in the target lesion segment.

[0047] Step S200: Based on the acquired DSA images, extract vascular morphology parameters and perform angiographic calcification grading of calcified plaques in the target lesion segment. Finally, use the patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used as feature data for prediction.

[0048] Step S300: Input the feature data into the pre-trained machine learning model and output the predicted effect after PCI drug balloon surgery.

[0049] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Step S100: Obtain preoperative DSA images of the patient to be evaluated via PCI drug balloon, wherein the DSA images contain information on calcified plaques in the target lesion segment.

[0050] Step S200: Based on the acquired DSA images, extract vascular morphology parameters and perform angiographic calcification grading of calcified plaques in the target lesion segment. Finally, use the patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used as feature data for prediction.

[0051] Step S300: Input the feature data into the pre-trained machine learning model and output the predicted effect after PCI drug balloon surgery.

[0052] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0053] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0054] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading, characterized in that, The method includes: Obtain preoperative DSA images of patients undergoing PCI drug-eluting balloon surgery, where the DSA images contain information on calcified plaques in the target lesion segment; Based on the acquired DSA images, vascular morphology parameters are extracted, and calcification of calcified plaques in the target lesion segment is graded by angiography. Finally, the patient's vascular morphology parameters, angiography calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction. The feature data is input into a pre-trained machine learning model, which outputs the predicted effect after PCI drug balloon surgery.

2. The method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading as described in claim 1, characterized in that, The imaging calcification grading of calcified plaques in the target lesion segment includes: Each frame in the DSA image sequence is input into a pre-trained deep learning calcification grading model. The deep learning calcification grading model outputs the corresponding calcification degree prediction result based on a single frame of DSA image. The calcification degree includes four levels: no calcification, mild calcification, moderate calcification, and severe calcification. Based on the prediction results of all frames in the DSA image sequence, patient-level calcification grading features of the target blood vessels of the target patient are aggregated and generated.

3. The method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading as described in claim 2, is characterized in that... The DSA images were derived from the left anterior descending artery (LAD), the circumflex artery (LCX), and the right coronary artery (RCA), with a balanced distribution of sample numbers for each vessel type.

4. The method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading as described in claim 1, characterized in that, The extraction of vascular morphology parameters based on the acquired DSA images includes: Preprocess the DSA images; Spatial scale calibration of DSA images is performed based on reference objects with known physical dimensions to obtain blood vessel images with real physical units; Select the analysis frame from the DSA image where the target lesion segment is clearly visualized and the blood vessel overlap is minimal; Based on the analysis frame, the centerline of the target blood vessel is extracted and the lumen contour is segmented; Based on the lumen profile, calculate the vascular morphology parameters of the target lesion segment.

5. The method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading according to claim 4, characterized in that, The vascular morphological parameters include at least the reference vessel diameter, minimum lumen diameter, and degree of vascular stenosis.

6. The method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading as described in claim 1, characterized in that, The machine learning model was trained in the following way: The model was trained using a labeled dataset containing historical patients' DSA images, calcification grading results, clinical data, and actual surgical outcome data. The training objective of the machine learning model was to minimize the mean squared error between the predicted and actual results.

7. The method for predicting the postoperative efficacy of PCI drug-eluting balloon angiography based on coronary DSA and angiographic calcification grading as described in claim 1, characterized in that, The predicted effects include predicted values ​​for vasodilation gain, postoperative minimum lumen diameter, and / or functional parameters.

8. A device for predicting the postoperative effect of PCI drug-eluting balloon surgery based on coronary DSA and angiographic calcification grading, characterized in that, The device includes: The image acquisition module is used to acquire DSA images of patients undergoing PCI drug balloon surgery before evaluation. The DSA images contain information on calcified plaques in the target lesion segment. The feature extraction module is used to extract vascular morphology parameters based on the acquired DSA images, and to perform angiographic calcification grading of calcified plaques in the target lesion segment. Finally, the patient's vascular morphology parameters, angiographic calcification grading results, clinical data, and instrument data of the drug-eluting balloon to be used are used as feature data for prediction. The prediction module is used to input feature data into a pre-trained machine learning model and output the prediction effect after PCI drug balloon surgery.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images