Thrombus property detection method and device, medical equipment and storage medium

Abstract

The embodiment of the invention is suitable for the technical field of medical treatment, and provides a thrombus property detection method and device, medical equipment and a storage medium, and the method comprises the steps: obtaining a first blood vessel image of a patient; the first blood vessel image is a preoperative image or an intraoperative image of a patient, and the first blood vessel image comprises a thrombus area; extracting features of thrombus from the first blood vessel image; and inputting the thrombus characteristics into a trained analysis model to obtain a thrombus property detection result of the patient. By adopting the method, the thrombus property can be analyzed before the operation so as to provide help for the clinical vascular recanalization operation treatment.

Description

Technical Field

[0001] The present application belongs to the field of medical technology, and in particular relates to a method, apparatus, medical equipment and storage medium for detecting properties of thrombus. Background Art

[0002] Thrombi are small masses of blood formed on the surface of a broken or repaired cardiovascular vessel. They are composed of insoluble fibrin, deposited platelets, accumulated white blood cells, and trapped red blood cells. Analyzing the properties of thrombi can provide important guidance for patient prognosis.

[0003] Currently, the analysis of the nature of thrombus mainly relies on histopathological analysis of thrombus removed from patients after surgery to determine the assessment results of the patient's prognosis.

[0004] However, this method requires the patient to undergo surgery before analyzing the nature of the thrombus, and cannot provide assistance for clinical vascular recanalization surgery. Summary of the Invention

[0005] The embodiments of the present application provide a method, apparatus, medical device and storage medium for detecting thrombus properties, which can solve the problem of being unable to perform thrombus property analysis before surgery to provide assistance for clinical vascular recanalization surgery.

[0006] In a first aspect, embodiments of the present application provide a method for detecting properties of a thrombus, the method comprising:

[0007] Acquiring a first blood vessel image of the patient; the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area;

[0008] extracting features of the thrombus from the first blood vessel image;

[0009] The characteristics of the thrombus are input into the trained analysis model to obtain the patient's thrombus property detection results.

[0010] In a second aspect, an embodiment of the present application provides a device for detecting properties of a thrombus, the device comprising:

[0011] A first acquisition module is configured to acquire a first blood vessel image of the patient; the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area;

[0012] an extraction module, configured to extract features of the thrombus from the first blood vessel image;

[0013] The first input module is used to input the characteristics of the thrombus into the trained analysis model to obtain the patient's thrombus property detection results.

[0014] In a third aspect, an embodiment of the present application provides a medical device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method of the first aspect described above when executing the computer program.

[0015] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method of the first aspect described above.

[0016] In a fifth aspect, an embodiment of the present application provides a computer program product, which, when executed on a medical device, enables the medical device to execute the method of the first aspect described above.

[0017] Compared to the prior art, the present embodiment of the present invention offers the following advantages: a medical device can acquire a first vascular image of a patient, then extract thrombus features from the first vascular image to determine the relevant properties of the thrombus. Finally, the thrombus features are analyzed using a pre-trained analysis model to obtain a thrombus property detection result. The first vascular image can be a preoperative or intraoperative image of the patient, and the first vascular image contains a thrombus region. In this embodiment, this method eliminates the need to perform surgery and remove the thrombus before analyzing the patient's thrombus properties. Thus, through this method, the medical device can obtain thrombus property detection results before surgery, thereby assisting in clinical vascular recanalization surgery. Furthermore, compared to relying on the doctor's personal experience to analyze the first vascular image by reading the image, using the above analysis model to analyze the thrombus features and obtain thrombus property detection results is simpler and more operational, and the thrombus property detection results derived based on the thrombus features are also relatively accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0019] Figure 1 This is a flowchart of a method for detecting thrombus properties provided in one embodiment of the present application;

[0020] Figure 2 This is a schematic diagram of an implementation method for determining anatomical structural features of a thrombus in a thrombus property detection method provided in one embodiment of the present application;

[0021] Figure 3This is a schematic diagram of an application scenario of a thrombus location in a thrombus property detection method provided in an embodiment of the present application;

[0022] Figure 4 This is a schematic diagram of an implementation method for determining thrombus perfusion characteristics in a thrombus property detection method provided in one embodiment of the present application;

[0023] Figure 5 This is a flowchart of a method for detecting thrombus properties provided in another embodiment of the present application;

[0024] Figure 6 This is a schematic structural diagram of a thrombus property detection device provided in one embodiment of the present application;

[0025] Figure 7 This is a structural diagram of a medical device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0026] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0027] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0028] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0029] Thrombi are small masses of blood formed on the surface of a broken or repaired cardiovascular vessel. They are composed of insoluble fibrin, deposited platelets, accumulated white blood cells, and trapped red blood cells. Analyzing the properties of thrombi can provide important guidance for patient prognosis.

[0030] Thrombi can be divided into two types based on the relative content of tissue components and permeability. Specifically, the first type of thrombus is a low-permeability thrombus rich in fibrin or platelets. Patients with this type of thrombus have a poor prognosis. In other words, the patient's treatment effect is unsatisfactory and may have sequelae. The second type is a high-permeability thrombus rich in red blood cells. Patients with this type of thrombus have a better prognosis. In other words, the patient's treatment effect is excellent.

[0031] Currently, the analysis of the nature of thrombus mainly relies on histopathological analysis of thrombus removed from patients after surgery to determine the assessment results of the patient's prognosis.

[0032] However, this method requires the patient to undergo surgery before analyzing the nature of the thrombus, and cannot provide assistance for clinical revascularization surgery. For example, different surgical strategies can be used for different types of thrombus properties.

[0033] Based on this, in order to determine the nature of the patient's thrombus before surgery so as to carry out targeted clinical vascular recanalization surgery, the present application provides a thrombus property detection method, which can be applied to mobile phones, tablet computers, laptops, ultra-mobile personal computers (Ultra-Mobile Personal Computer, UMPC), netbooks, thrombus detection intelligent analyzers and other devices. The embodiments of the present application do not impose any restrictions on the specific type of medical equipment.

[0034] See also Figure 1 , Figure 1 The flowchart of a method for detecting thrombus properties provided in an embodiment of the present application is shown. The method includes the following steps:

[0035] S101 . Acquire a first blood vessel image of a patient; the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area.

[0036] In one embodiment, the patient may have a thrombotic disorder, in which case the medical device may perform the above method to determine the nature of the thrombus in the patient. Alternatively, the patient may not have a thrombotic disorder, in which case the medical device may perform the above method to detect whether the patient has a thrombus. This embodiment uses the example of a patient with a thrombotic disorder. Specifically, the first vascular image is a preoperative or intraoperative image of the patient, and the first vascular image contains a thrombus region.

[0037] The first blood vessel image may be one or more of a digital subtraction angiography (DSA) image and a cone beam CT (CBCT) image, and is not limited thereto.

[0038] It is understandable that the DSA and CBCT images can be acquired by angiography equipment. In this embodiment, the method for acquiring the first blood vessel image is not limited.

[0039] The first vascular image can be one or more of a two-dimensional DSA image, a three-dimensional DSA image, and a CBCT image. That is, the number of first vascular images can be one or more. Furthermore, when there are multiple first vascular images, the types of the multiple first vascular images can be the same or different, without limitation.

[0040] S102: Extracting features of the thrombus from the first blood vessel image.

[0041] In one embodiment, the above-mentioned thrombus characteristics can be used to describe the characteristics of the patient's thrombus, wherein the thrombus characteristics can include at least one of thrombus anatomical structure characteristics and thrombus perfusion characteristics.

[0042] It should be noted that the aforementioned thrombus anatomical structural characteristics may describe characteristics such as thrombus size, length, shape, and connection with the vessel wall (whether it is adhered to the vessel wall). However, thrombus anatomical structural characteristics may vary depending on the nature of the thrombus or the location of the thrombus.

[0043] The aforementioned thrombus perfusion characteristics describe the permeability of blood through the thrombus at the thrombus location. The medical device can determine the thrombus perfusion characteristics based on the signal intensity of the contrast agent flowing through the thrombus over a preset duration. Alternatively, the medical device can measure signal intensity using color Doppler ultrasound equipment or angiographic CT angiography equipment.

[0044] For example, the medical device may use the average value of the signal intensity within a preset time period when the contrast agent flows through the thrombus, or the maximum value of the signal intensity within the preset time period as the thrombus perfusion feature, which is not limited to this.

[0045] It is understandable that determining the thrombus perfusion characteristics based on the signal intensity within a preset time period can avoid errors caused by occasional excessive or insufficient signal intensity, thereby improving the accuracy of the thrombus perfusion characteristics.

[0046] As an example, in order to accurately describe the anatomical structure characteristics of the thrombus, in this embodiment, the medical device can use the following Figure 2 The steps S201-S203 shown generate the anatomical structure features of the thrombus. The details are as follows:

[0047] S201 : Determine a target region containing a thrombus from a first blood vessel image, extract the morphology of the thrombus from the target region, and determine the image intensity in the target region.

[0048] In one embodiment, the target region may be a ROI (Region of Interest) in the first vascular image. The target region may be pre-annotated by medical personnel, or a thrombus recognition model in a medical device may be used to identify the thrombus location in the first vascular image. A region of a predetermined size, centered on the identified thrombus location, is then defined as the target region. The predetermined size may be a circular region with a predetermined radius R, or a rectangular region with a length a and a width b, without limitation.

[0049] Specifically, refer to Figure 3 , Figure 3 This is a schematic diagram of an application scenario of a thrombus location in a thrombus property detection method provided in an embodiment of the present application. Figure 3 When the first blood vessel image is a DSA image, the thrombus position in the patient's brain in the first blood vessel image.

[0050] In one embodiment, the thrombus morphology may include posture information such as the shape, size, and position of the thrombus. The medical device may describe the thrombus morphology using boundary features of the region where the thrombus is located, or may use the ratio of the area occupied by the thrombus to the area of the target region to represent the thrombus morphology, without limitation.

[0051] It should be noted that using the boundary features of the area where the thrombus is located to describe the morphology of the thrombus can intuitively represent the thrombus morphology information, which helps the subsequent analysis model understand the thrombus morphology. In addition, using the area ratio to represent the morphology of the thrombus can reflect the coverage or severity of the thrombus. It can be understood that when the area ratio is larger, it can be considered that the patient's thrombus is larger. In other words, the thrombus is more serious. Conversely, when the area ratio is smaller, it can be considered that the patient's thrombus is smaller. In other words, the thrombus is not serious. Based on this, it can be considered that using the area ratio to represent the morphology of the thrombus can also help the analysis model understand the nature of the thrombus.

[0052] The image intensity can be considered as the intensity of the image pixels. In this embodiment, the image intensity can be represented by the pixel value or grayscale value of the image. For example, the average grayscale value of all pixels in the target area can be used as the image intensity.

[0053] It should be noted that the aforementioned thrombi are classified into low-permeability thrombi rich in fibrin or platelets and high-permeability thrombi rich in red blood cells. Furthermore, because different types of thrombi contain different components (one type is rich in red blood cells, the other is rich in fibrin or platelets), their corresponding image intensities typically differ, which will not be explained in detail.

[0054] S202. Extract high-order image features of the thrombus from the first blood vessel image based on radiomics.

[0055] In one embodiment, radiomics is the process of extracting large amounts of image information from images (such as CT, MRI, and PET) at high throughput, enabling image segmentation, feature extraction, and model building. High-order image features refer to image feature variables extracted by applying filters. Extraction methods include, but are not limited to, high-order image features obtained by applying wavelet transform, fractal dimension, and metric function processing to the first vascular image.

[0056] S203. Determine the morphology, image intensity, and high-order image features as thrombus anatomical structure features.

[0057] It should be noted that the aforementioned thrombus morphology and image intensity and other anatomical structural features of the thrombus can be considered as low-order image features of the thrombus. It is understood that low-order image features are features such as the contour, edge, color, texture, and shape of the target (thrombus) in the image.

[0058] The thrombus morphology can be considered as the semantic features of the thrombus in the first vessel image. That is, it uses machine language to represent what the human eye can see (the shape and edges of the thrombus). Furthermore, the image intensity features can be considered as the texture features of the thrombus in the first vessel image. Texture features (image intensity features) can reflect the heterogeneity of signal intensity within the thrombus.

[0059] It should be noted that low-level image features usually contain less semantic information but accurately locate the target, while high-level image features usually contain richer semantic information but roughly locate the target.

[0060] Based on this, using low-order image features and high-order image features as thrombus anatomical structure features can fully reflect the essential information of thrombus properties, thereby improving the accuracy of subsequent thrombus property analysis based on thrombus characteristics.

[0061] As another example, in order to accurately describe the thrombus perfusion characteristics, in this embodiment, the medical device can use the following method: Figure 4 The steps S401-S403 shown generate the anatomical structure features of the thrombus. The details are as follows:

[0062] S401: Determine a thrombus position from a first blood vessel image.

[0063] S402 : Generate a signal intensity variation curve showing the signal intensity varying with time according to the signal intensity at the thrombus position.

[0064] In one embodiment, the above S201 has described how to determine the location of the thrombus, and the above S102 has described how to measure the signal strength, which will not be described again.

[0065] In this embodiment, to accurately characterize thrombus perfusion characteristics, the medical device may collect signal intensity at the thrombus location at predetermined collection intervals or in real time to generate a signal intensity variation curve of signal intensity over time. The predetermined collection interval may be set based on actual conditions and is not limited thereto.

[0066] S403: Determine thrombus perfusion characteristics based on the signal intensity change curve.

[0067] It is understood that the signal intensity curve can reflect the permeability of the thrombus, which is related to its cellular composition, including tissue components such as fibrin, platelets, and red blood cells. Therefore, the thrombus perfusion characteristics obtained based on the signal intensity curve can accurately characterize the permeability of the thrombus to blood and its tissue composition.

[0068] The medical device may use the curve characteristics of the signal intensity change curve as the thrombus perfusion characteristic. For example, the medical device may determine the curve characteristics from the signal intensity change curve; further, the medical device may fit the signal intensity change curve to determine a fitting form of the signal intensity change curve. Finally, both the curve characteristics and the fitting form are determined as the thrombus perfusion characteristic.

[0069] Exemplarily, the curve features include but are not limited to the maximum peak value in the signal strength change curve, the peak time corresponding to the maximum peak value, the rising slope and the falling slope in the signal strength change curve, which are not limited.

[0070] In one embodiment, the maximum peak value can reflect the maximum blood volume within the lesion corresponding to the thrombus. Marking the peak time corresponding to each peak value can more intuitively display the trend of signal intensity changes, thereby enabling the analysis model to better analyze and predict thrombi and improve the accuracy of thrombus property detection results.

[0071] The above-mentioned fitting of the signal intensity variation curve includes, but is not limited to, a linear fitting form and a Gaussian fitting form, which are not limited thereto.

[0072] It should be noted that different thrombus components and properties typically generate signal intensity curves with different fitting forms. In other words, it can be assumed that the fitting form of the signal intensity curve can characterize the properties and components of the thrombus, thereby assisting the analysis model in accurately generating thrombus property detection results.

[0073] For example, a linear fit usually indicates a high-permeability thrombus rich in red blood cells, which is believed to indicate a better prognosis for the patient, while a Gaussian fit usually indicates a low-permeability thrombus rich in fibrin or platelets, which is believed to indicate a worse prognosis for the patient.

[0074] It should be noted that by analyzing the time curve of the signal intensity at the thrombus location in the first vascular image, quantitative information about blood perfusion can be obtained. Based on this, the medical device can discern the details of the thrombus perfusion process to obtain analysis parameters during the perfusion process (thrombus perfusion characteristics). Furthermore, in subsequent analysis, the analysis parameters that can quantify the blood perfusion process are input into the analysis model as thrombus perfusion characteristics, further improving the accuracy of thrombus property detection results.

[0075] S103. Input the characteristics of the thrombus into the trained analysis model to obtain the patient's thrombus property detection results.

[0076] In one embodiment, the thrombus properties have been described above and will not be further explained. It should be noted that the above example is only an illustration of processing a first blood vessel image to obtain a thrombus property detection result for the first blood vessel image.

[0077] However, if there are multiple first blood vessel images, the medical device can process each first blood vessel image separately to obtain a thrombus property detection result corresponding to each blood vessel image. Then, based on the multiple thrombus property detection results, the number of thrombus property detection results corresponding to each thrombus property type is counted. Finally, the thrombus property type corresponding to the largest number is determined as the thrombus property detection result for the patient.

[0078] The analysis model may be obtained by pre-training the model based on the thrombus property training labels corresponding to the postoperative thrombus and the training thrombus characteristics of the postoperative thrombus.

[0079] Specifically, medical equipment can be pre-installed based on Figure 5 The training and analysis model shown in S501-S506 is described in detail as follows:

[0080] S501 : Acquire training data, where the training data includes training thrombus features determined from a second blood vessel image in a post-operative thrombus, and thrombus property training labels corresponding to the post-operative thrombus.

[0081] In one embodiment, the training data is raw medical data from thrombosis treatments at major hospitals. For example, after treating thrombosis in other patients, doctors can perform histopathological analysis on the thrombi removed post-operatively to obtain training labels for thrombus properties. Furthermore, a second vessel image of the thrombus is acquired and, using the method described in S102 above, training thrombus features are extracted. This will not be described in detail.

[0082] S502: Input the training thrombus features into the initial model for training to obtain a prediction label corresponding to the postoperative thrombus.

[0083] In one embodiment, the initial model can be a pre-set neural network model, which is not limited to this. Inputting the training thrombus features into the initial model for training is a forward propagation process. This process typically requires processing through the input layer, hidden layer, and output layer of the initial model to obtain a predicted label.

[0084] S503: Determine the training loss corresponding to the postoperative thrombus based on the thrombus property training label and the prediction label.

[0085] In one embodiment, the medical device may pre-characterize the thrombus property training label with a corresponding numerical value, for example, 0 or 1. The training loss is then calculated based on the numerical value corresponding to the predicted label and the numerical value corresponding to the thrombus property training label.

[0086] The numerical value corresponding to the predicted label can be the probability value that the initial model predicts the thrombus property to be the predicted label; alternatively, when the predicted label and the thrombus property training label corresponding to the thrombus are the same, the numerical value corresponding to the thrombus property training label can be used as the numerical value corresponding to the predicted label; and when the predicted label and the thrombus property training label corresponding to the thrombus are different, the numerical value corresponding to another thrombus property training label can be used as the numerical value corresponding to the predicted label. In this embodiment, the method for determining the numerical value corresponding to the predicted label is not limited.

[0087] In one embodiment, the training loss may be calculated using a preset loss function, for example, a squared loss function or an absolute loss function, which is not limited thereto.

[0088] S504: Iteratively update model parameters in the initial model based on the training loss.

[0089] In one embodiment, the medical device may update the model parameters in the initial model based on the training loss through back propagation. The updating method of the model parameters may be a gradient descent method, an Adam algorithm, or the like, which is not limited thereto.

[0090] S505: If the training loss converges during the iterative update process, the training of the initial model is terminated, and the current initial model is determined as the analysis model.

[0091] S506. If the training loss does not converge during the iterative update process, repeat the target step and the steps after the target step until the training loss converges; the target step includes inputting the training thrombus features into the initial model for training to obtain the prediction label corresponding to the postoperative thrombus.

[0092] In one embodiment, the training loss convergence includes but is not limited to the training loss being less than a preset training loss, or the number of iterative updates reaching a preset number, which are not limited thereto. The preset training loss and the preset number of updates can both be set by medical personnel.

[0093] It should be noted that when the training loss has not converged, the medical device may repeatedly execute the above steps S502-S506 until an analysis model is obtained.

[0094] In this embodiment, a medical device can acquire a first blood vessel image of a patient and then extract thrombus features from the first blood vessel image to determine the relevant properties of the thrombus. Finally, the thrombus features are analyzed using a pre-trained analysis model to obtain a thrombus property detection result. The first blood vessel image can be a preoperative or intraoperative image of the patient, and the first blood vessel image contains a thrombus area. In this embodiment, this method eliminates the need to perform surgery and remove the thrombus before analyzing the patient's thrombus properties. Thus, through this method, the medical device can obtain thrombus property detection results before surgery, providing assistance for clinical vascular recanalization surgery. Furthermore, compared to relying on the doctor's personal experience to analyze the first blood vessel image by reading the image, using the above analysis model to analyze the thrombus features and obtain thrombus property detection results is simpler and more operational, and the thrombus property detection results obtained based on the thrombus features are relatively accurate.

[0095] In another embodiment, after obtaining the thrombus property detection results, the medical device may also generate a surgical strategy for the thrombus and / or an assessment result of the patient's prognosis based on the thrombus property detection results.

[0096] For example, it has been described above that patients with low-permeability thrombi rich in fibrin or platelets have a poor prognosis. Furthermore, patients with high-permeability thrombi rich in red blood cells have a better prognosis. Therefore, when determining the thrombus nature test results, the medical device can determine the above prognostic assessment results.

[0097] In addition, it has been explained above that the thrombus property detection method in this embodiment is a method for detecting a patient before surgery. Therefore, the medical device can also provide guidance for doctors' clinical surgical treatment based on the thrombus property detection results.

[0098] For example, when the thrombus property test result indicates a low-permeability thrombus rich in fibrin or platelets, the medical device may generate a surgical strategy using stent thrombectomy. Furthermore, when the thrombus property test result indicates a high-permeability thrombus containing red blood cells, the medical device may generate a surgical strategy using aspiration thrombectomy or intravenous thrombolysis.

[0099] The surgical strategy corresponding to each thrombus type can be pre-set by medical staff in the medical device. Furthermore, the medical device can quickly and accurately generate the surgical strategy and prognostic assessment results based on the thrombus property detection results to assist in clinical revascularization surgery.

[0100] See also Figure 6 , Figure 6 This is a schematic diagram of the structure of a thrombus property detection device provided in an embodiment of the present application. The thrombus property detection device in this embodiment includes modules for performing Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 Each step in the corresponding embodiment. Please refer to Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 as well as Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 For the convenience of explanation, only the parts related to this embodiment are shown. Figure 6 The thrombus property detection device 600 may include: a first acquisition module 610, an extraction module 620, and a first input module 630, wherein:

[0101] The first acquisition module 610 is configured to acquire a first blood vessel image of the patient; the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area.

[0102] The extraction module 620 is configured to extract the characteristics of the thrombus from the first vascular image; the first vascular image is a preoperative image or an intraoperative image of the patient, and the first vascular image contains a thrombus area.

[0103] The first input module 630 is used to input the characteristics of the thrombus into the trained analysis model to obtain the thrombus property detection result of the patient.

[0104] In one embodiment, the characteristic of the thrombus includes at least one of a thrombus anatomical structure characteristic and a thrombus perfusion characteristic.

[0105] In one embodiment, the extraction module 620 is further configured to:

[0106] A target area containing a thrombus is determined from the first vascular image, and the morphology of the thrombus is extracted from the target area, and the image intensity in the target area is determined; high-order image features of the thrombus are extracted from the first vascular image based on imaging omics; and the morphology, image intensity, and high-order image features are determined as anatomical structural features of the thrombus.

[0107] In one embodiment, the extraction module 620 is further configured to:

[0108] The thrombus position is determined from the first blood vessel image; a signal intensity variation curve showing the signal intensity varying with time is generated according to the signal intensity at the thrombus position; and a thrombus perfusion characteristic is determined based on the signal intensity variation curve.

[0109] In one embodiment, the extraction module 620 is further configured to:

[0110] Determine thrombus characteristics from the signal intensity change curve; fit the signal intensity change curve to determine the fitting form of the signal intensity change curve; determine the curve characteristics and the fitting form as thrombus perfusion characteristics.

[0111] In one embodiment, the thrombus property detection device 600:

[0112] The generation module is used to generate a surgical strategy for thrombosis and / or an evaluation result of the patient's prognosis based on the thrombus property detection result.

[0113] In one embodiment, the thrombus property detection device 600 further includes:

[0114] The second acquisition module is used to acquire training data, where the training data includes training thrombus features determined from the second blood vessel image in the postoperative thrombus and thrombus property training labels corresponding to the postoperative thrombus.

[0115] The second input module is used to input the training thrombus features into the initial model for training to obtain the prediction label corresponding to the postoperative thrombus.

[0116] The first determination module is used to determine the training loss corresponding to the postoperative thrombus based on the training label and the prediction label of the thrombus property.

[0117] An update module is used to iteratively update the model parameters in the initial model based on the training loss.

[0118] A second determination module is configured to terminate the training of the initial model and determine the current initial model as the analysis model if the training loss converges during the iterative update process;

[0119] The iterative module is used to repeatedly execute the target step and the steps after the target step until the training loss converges if the training loss does not converge during the iterative update process; the target step includes inputting the training thrombus features into the initial model for training to obtain the predicted label corresponding to the postoperative thrombus.

[0120] When it is understood that Figure 6 In the schematic diagram of the structure of the thrombus property detection device shown, each module is used to perform Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 The steps in the corresponding embodiments, and Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 Each step in the corresponding embodiment has been explained in detail in the above embodiment. Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 as well as Figure 1 、 Figure 2 、 Figure 4 as well as Figure 5 The relevant descriptions in the corresponding embodiments will not be repeated here.

[0121] Figure 7 This is a schematic diagram of the structure of a medical device provided in one embodiment of the present application. Figure 7 As shown, the medical device 700 of this embodiment includes: a processor 710, a memory 720, and a computer program 730 stored in the memory 720 and executable on the processor 710, such as a program for a method for detecting thrombus properties. When the processor 710 executes the computer program 730, the steps in each embodiment of the above-mentioned method for detecting thrombus properties are implemented, such as Figure 1 Alternatively, the processor 710 executes the computer program 730 to implement the above Figure 6 The functions of each module in the corresponding embodiment are, for example, Figure 6 For details on the functions of each module shown, please refer to Figure 6 Related description in the corresponding embodiment.

[0122] Exemplarily, the computer program 730 can be divided into one or more modules, one or more of which are stored in the memory 720 and executed by the processor 710 to implement the thrombus property detection method provided in the embodiments of the present application. One or more modules can be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program 730 in the medical device 700. For example, the computer program 730 can implement the thrombus property detection method provided in the embodiments of the present application.

[0123] The medical device 700 may include, but is not limited to, a processor 710 and a memory 720. Those skilled in the art will appreciate that Figure 7 This is merely an example of the medical device 700 and does not constitute a limitation of the medical device 700 . The medical device 700 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the medical device may also include input and output devices, network access devices, buses, etc.

[0124] The processor 710 may be a central processing unit, or other general-purpose processor, a digital signal processor, an application-specific integrated circuit, an off-the-shelf programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0125] The memory 720 may be an internal storage unit of the medical device 700, such as a hard disk or memory of the medical device 700. The memory 720 may also be an external storage device of the medical device 700, such as a plug-in hard disk, smart memory card, flash memory card, etc. equipped on the medical device 700. Furthermore, the memory 720 may include both an internal storage unit of the medical device 700 and an external storage device.

[0126] An embodiment of the present application provides a computer-readable storage medium, which stores a computer program. The computer program is executed by a processor to execute the thrombus property detection method in each of the above embodiments.

[0127] An embodiment of the present application provides a computer program product. When the computer program product is run on a medical device, the medical device executes the thrombus property detection method in each of the above embodiments.

[0128] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for detecting thrombus properties, characterized in that: The method comprises: Acquiring a first blood vessel image of a patient; the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area; extracting features of the thrombus from the first blood vessel image; The characteristics of the thrombus are input into the trained analysis model to obtain the thrombus property detection results of the patient.

2. The method according to claim 1, characterized in that The thrombus characteristics include at least one of thrombus anatomical structure characteristics and thrombus perfusion characteristics.

3. The method according to claim 2, characterized in that The extracting the feature of the thrombus from the first blood vessel image includes: determining a target region containing the thrombus from the first blood vessel image, extracting the morphology of the thrombus from the target region, and determining an image intensity in the target region; extracting high-order image features of the thrombus from the first blood vessel image based on radiomics; The morphology, the image intensity, and the high-order image features are determined as the thrombus anatomical structure features.

4. The method according to claim 2, characterized in that The extracting the feature of the thrombus from the first blood vessel image includes: determining a thrombus location from the first blood vessel image; generating a signal intensity variation curve of the signal intensity varying with time according to the signal intensity at the thrombus position; The thrombus perfusion characteristic is determined based on the signal intensity change curve.

5. The method according to claim 4, characterized in that Determining the thrombus perfusion characteristic based on the signal intensity change curve includes: determining a curve characteristic from the signal intensity variation curve; Fitting the signal intensity variation curve to determine a fitting form of the signal intensity variation curve; The curve characteristic and the fitting form are determined as the thrombus perfusion characteristic.

6. The method according to claim 1, characterized in that After inputting the characteristics of the thrombus into the trained analysis model to obtain the thrombus property detection result of the patient, the method further includes: Based on the thrombus property detection result, a surgical strategy for the thrombus and / or an evaluation result of the patient's prognosis is generated.

7. The method according to any one of claims 1 to 6, characterized in that The method further comprises: Acquiring training data, the training data including training thrombus features determined from a second blood vessel image in a post-operative thrombus and a thrombus property training label corresponding to the post-operative thrombus; Inputting the training thrombus features into an initial model for training to obtain a prediction label corresponding to the postoperative thrombus; determining a training loss corresponding to the postoperative thrombus according to the thrombus property training label and the prediction label; Iteratively updating model parameters in the initial model based on the training loss; If the training loss converges during the iterative update process, the training of the initial model is terminated, and the current initial model is determined as the analysis model; If the training loss does not converge during the iterative update process, the target step and the steps after the target step are repeatedly executed until the training loss converges; the target step includes inputting the training thrombus features into the initial model for training to obtain a prediction label corresponding to the postoperative thrombus.

8. A device for detecting properties of thrombus, characterized in that: The device comprises: A first acquisition module is configured to acquire a first blood vessel image of a patient; the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area; an extraction module, configured to extract features of the thrombus from the first blood vessel image; The first input module is used to input the characteristics of the thrombus into the trained analysis model to obtain the thrombus property detection result of the patient.

9. A medical device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the following is achieved: Acquiring a first blood vessel image of a patient, where the first blood vessel image is a preoperative image or an intraoperative image of the patient, and the first blood vessel image contains a thrombus area; extracting features of the thrombus from the first blood vessel image; The characteristics of the thrombus are input into the trained analysis model to obtain the thrombus property detection results of the patient.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images