Abdominal aortic aneurysm rupture detection method and device and electronic equipment

Abstract

The invention relates to the technical field of medical images.The invention discloses an abdominal aortic aneurysm rupture detection method and device and electronic equipment, and the abdominal aortic aneurysm rupture detection method comprises the steps that effective data containing the retroperitoneal space is obtained from original scanning data of abdominal aortic aneurysm, and the effective data comprises a plurality of data units; identifying whether a data unit in the valid data has a specific target, and counting the number of valid units of the data unit with the specific target; and determining whether abdominal aortic aneurysm rupture occurs or not according to the number of the effective units. According to the method, the detection result of abdominal aortic aneurysm rupture can be accurately and quickly obtained.

Description

Technical Field

[0001] The present disclosure relates to the technical field of medical imaging, for example, to a method, device, and electronic device for detecting abdominal aortic aneurysm rupture. Background Art

[0002] Abdominal aortic aneurysm (AAA) is defined as a local dilation of the abdominal aorta that exceeds 1.5 times the normal diameter. If an AAA ruptures, the situation becomes very dangerous and can cause massive internal bleeding. For example, if an AAA ruptures directly into the abdominal cavity, hemorrhagic shock will occur rapidly; if an AAA ruptures into the retroperitoneal space, a localized hematoma may form.

[0003] Ruptured Abdominal Aortic Aneurysm (RAAA) is a surgical emergency and if not effectively treated, it will lead to a high mortality rate. Open surgery is a quick and effective way to treat ruptured abdominal aortic aneurysm, so quickly detecting ruptured abdominal aortic aneurysm is crucial to saving the patient's life. In related technologies, the focus is on the early prediction before the formation of abdominal aortic aneurysm and the risk prediction of abdominal aortic aneurysm rupture. There is a lack of related research on the detection of ruptured abdominal aortic aneurysm, which makes it impossible for related technologies to detect ruptured abdominal aortic aneurysm in a timely, accurate and rapid manner.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] In order to provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. The summary is not an extensive review, nor is it intended to identify key / critical components or delineate the scope of protection of these embodiments, but rather serves as a prelude to the detailed description that follows.

[0006] The embodiments of the present disclosure provide a method, device, and electronic device for detecting abdominal aortic aneurysm rupture, which can accurately and quickly obtain the detection results of abdominal aortic aneurysm rupture.

[0007] According to a first aspect of the present disclosure, a method for detecting abdominal aortic aneurysm rupture is provided, comprising:

[0008] Acquire valid data including the retroperitoneal space from the original scan data of the abdominal aortic aneurysm, wherein the valid data includes a plurality of data units;

[0009] Identify whether a data unit in valid data has a specific target, and count the number of valid units of data units that have the specific target;

[0010] The presence of abdominal aortic aneurysm rupture was determined based on the number of effective units.

[0011] In some embodiments, the number of valid units is the number of data units that are continuously identified to have a specific target.

[0012] In some embodiments, identifying whether a data unit in valid data has a specific target and counting the number of valid units of the data unit having the specific target includes:

[0013] Identify whether the data unit in the valid data has a specific target, and count the number of valid units in real time;

[0014] When a data unit without a specific target is identified, the number of valid units is cleared, and the remaining data units are continuously identified to see whether they have a specific target, and the number of valid units is recounted until the number of valid units reaches a threshold or all data units have been identified.

[0015] In some embodiments, determining whether abdominal aortic aneurysm rupture occurs according to the effective number of units comprises:

[0016] When the number of valid units does not reach the quantity threshold, it is determined that no abdominal aortic aneurysm rupture has occurred;

[0017] When the number of valid units reaches a number threshold, it is determined that abdominal aortic aneurysm rupture occurs.

[0018] In some embodiments, the data unit is a single layer image or a single cuboid unit in the original scan data.

[0019] In some embodiments, the specific target is a site of retroperitoneal bleeding and / or a site of contrast agent extravasation.

[0020] In some embodiments, valid data including the retroperitoneal space is obtained from the original scan data of the abdominal aortic aneurysm, including: inputting the original scan data of the abdominal aortic aneurysm into a retroperitoneal space recognition model, and using the retroperitoneal space recognition model to obtain valid data including the retroperitoneal space.

[0021] In some embodiments, the retroperitoneal space recognition model is trained by:

[0022] The retroperitoneal space area is marked in each sample original scan data, and then all the sample original scan data are divided into a training set and a test set;

[0023] Using the training set to train the initial retroperitoneal space recognition model;

[0024] After the retroperitoneal space recognition model is trained using the training set, the retroperitoneal space recognition model is tested using the test set.

[0025] In some embodiments, identifying whether a data unit in valid data has a specific target includes: inputting the data unit in the valid data into a specific target detection model, and using the target detection model to determine whether the data unit has a specific target.

[0026] In some embodiments, the specific target detection model is trained by:

[0027] Mark a specific target in each sample valid data containing the retroperitoneal space, and then divide all the sample valid data into a training set and a test set;

[0028] Using the training set to train an initial specific target detection model, where the target detection model includes a convolution module, an RPN module, and a Roi Head module;

[0029] After the specific target detection model is trained using the training set, the specific target detection model is tested using the test set.

[0030] According to a second aspect of the present disclosure, there is provided an abdominal aortic aneurysm rupture detection device, comprising:

[0031] An effective data acquisition module is configured to acquire effective data including the retroperitoneal space in the original scan data of the abdominal aortic aneurysm, wherein the effective data includes a plurality of data units;

[0032] A specific target identification module is configured to identify whether a data unit in the valid data has a specific target, and count the number of valid units of the data unit with the specific target;

[0033] The rupture result determination module is configured to determine whether the abdominal aortic aneurysm rupture occurs according to the number of valid units.

[0034] According to a third aspect of the present disclosure, an electronic device is provided, comprising a processor and a memory storing program instructions, wherein the processor is configured to execute the abdominal aortic aneurysm rupture detection method provided by the first aspect of the present disclosure when running the program instructions.

[0035] The abdominal aortic aneurysm rupture detection method, device, and electronic device provided by the embodiments of the present disclosure can achieve the following technical effects:

[0036] The abdominal aortic aneurysm rupture detection method provided by the embodiment of the present disclosure takes into account that after the rupture of the abdominal aortic aneurysm, a specific target (bleeding or extravasation of contrast agent) will appear in the retroperitoneal space. Whether the scanning data has a specific target can be used as an important basis for whether the abdominal aortic aneurysm rupture occurs. Therefore, firstly, valid data containing the retroperitoneal space is obtained in the original scanning data of the abdominal aortic aneurysm, so that the data processing range can be narrowed to improve the subsequent detection efficiency. On this basis, taking the data unit in the valid data as a unit, it is identified one by one whether the data unit has a specific target, and the number of valid units of the data unit with the specific target is counted. Finally, it is determined whether the abdominal aortic aneurysm rupture occurs based on the number of valid units, so as to accurately and quickly obtain the detection result of the abdominal aortic aneurysm rupture.

[0037] The foregoing general description and the following description are exemplary and explanatory only and are not intended to limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] One or more embodiments are exemplarily described by corresponding drawings, which do not limit the embodiments. Elements with the same reference numerals in the drawings are shown as similar elements, and the drawings do not constitute a scale limitation, and wherein:

[0039] Figure 1 is a schematic diagram of a method for detecting abdominal aortic aneurysm rupture provided by an embodiment of the present disclosure;

[0040] Figure 2 is a schematic diagram of another abdominal aortic aneurysm rupture detection method provided by an embodiment of the present disclosure;

[0041] Figure 3 is a schematic diagram of another abdominal aortic aneurysm rupture detection method provided by an embodiment of the present disclosure;

[0042] Figure 4 is a schematic diagram of another abdominal aortic aneurysm rupture detection method provided by an embodiment of the present disclosure;

[0043] Figure 5 is a schematic diagram of a specific target detection model provided by an embodiment of the present disclosure;

[0044] Figure 6 is a schematic diagram of an abdominal aortic aneurysm rupture detection device provided by an embodiment of the present disclosure;

[0045] Figure 7 It is a schematic diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0046] In order to be able to understand the features and technical contents of the embodiments of the present disclosure in more detail, the implementation of the embodiments of the present disclosure is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present disclosure. In the following technical description, for the convenience of explanation, a full understanding of the disclosed embodiments is provided through multiple details. However, one or more embodiments can still be implemented without these details. In other cases, to simplify the drawings, well-known structures and devices can be simplified for display.

[0047] The terms "first", "second", etc. in the specification and claims of the embodiments of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchanged where appropriate, so that the embodiments of the embodiments of the present disclosure described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions.

[0048] Unless otherwise stated, the term "plurality" means two or more.

[0049] In the embodiment of the present disclosure, the character " / " indicates that the preceding and following objects are in an "or" relationship. For example, A / B indicates: A or B.

[0050] The term "and / or" is a description of the association relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or, A and B.

[0051] The term "correspondence" may refer to an association relationship or a binding relationship. The correspondence between A and B means that there is an association relationship or a binding relationship between A and B.

[0052] Abdominal aortic aneurysm is defined as a local dilation of the abdominal aorta that exceeds 1.5 times the normal diameter. If an abdominal aortic aneurysm ruptures, the situation can become very dangerous and can cause massive internal bleeding. For example, if an abdominal aortic aneurysm ruptures directly into the abdominal cavity, hemorrhagic shock can occur rapidly; if an abdominal aortic aneurysm ruptures into the retroperitoneal space, a localized hematoma can form. Ruptured abdominal aortic aneurysm is a critical surgical emergency and can result in a high mortality rate if not effectively treated. Open surgery is a quick and effective way to treat ruptured abdominal aortic aneurysm, so quickly detecting ruptured abdominal aortic aneurysm is crucial to saving the patient's life.

[0053] In the related technologies, the focus is on the early prediction of abdominal aortic aneurysm before it forms and the prediction of the risk of abdominal aortic aneurysm rupture. The early prediction of abdominal aortic aneurysm before it forms is based on the patient's age, gender, smoking, obesity, hypertension, and family history of abdominal aortic aneurysm as a data set. The data set is trained based on the random forest algorithm to obtain a prediction model based on the random forest algorithm. The random forest consists of multiple decision trees, and all trees are independent of each other. When predicting a new sample set, which category the sample set belongs to depends on the judgment of each decision tree, and the prediction model itself is composed of multiple classifiers, and then the final prediction result is obtained by voting and averaging. For the prediction of the risk of abdominal aortic aneurysm rupture, a prediction system for the risk of abdominal aortic aneurysm rupture based on machine learning is usually used for prediction, in which the prediction system includes a three-dimensional reconstruction model module of the lesion area, a morphological parameter module, a finite element analysis module, a hemodynamic parameter module, and a clinical data module. The three-dimensional reconstruction model module is used to establish a three-dimensional model of the abdominal aortic aneurysm after image segmentation of the medical imaging data; the morphological parameter module is used to calculate the shape of the abdominal aorta, the size of the abdominal aorta, etc.; the finite element analysis module is used to determine the wall stress of the abdominal aortic aneurysm and the wall thickness of the abdominal aortic aneurysm strain result. The prediction system constructed based on the above modules can accurately predict the risk of abdominal aortic aneurysm rupture. However, the above solutions of the related technology are all applied to the stage before the rupture of the abdominal aortic aneurysm occurs, and there is a lack of relevant research on the detection of abdominal aortic aneurysm rupture, which results in the inability of the related technology to detect the rupture of the abdominal aortic aneurysm in a timely, accurate and rapid manner.

[0054] The embodiment of the present disclosure provides an electronic device, which may be a device with computing capability such as a computer, a terminal, or a server. The electronic device may determine whether an abdominal aortic aneurysm ruptures according to the presence of a specific target in the scan data of the abdominal aortic aneurysm.

[0055] In conjunction with the electronic device provided by the embodiment of the present disclosure (hereinafter referred to as the device), the embodiment of the present disclosure provides a method for detecting abdominal aortic aneurysm rupture, such as Figure 1 As shown, the method for detecting abdominal aortic aneurysm rupture includes:

[0056] S101, the device obtains valid data including the retroperitoneal space from the original scanning data of the abdominal aortic aneurysm.

[0057] In the disclosed embodiment, the original scan data is obtained by scanning a patient using a scanning device. For example, the scanning device may be a computed tomography (CT) device and a magnetic resonance imaging (MRI) device.

[0058] Taking a computed tomography device as an example, during the scanning process, the X-ray source and detector rotate around the patient while the X-rays pass through different parts of the body. The detector measures the changes in X-ray intensity at each angle and records a series of projection data. The collected projection data usually needs to go through a preprocessing step to remove noise and other undesirable factors. After the projection data is preprocessed, a filter is first applied to the projection data at each angle, and then the filtered projection data are back-projected along their original path onto a virtual two-dimensional plane. This process is repeated until all the projection data are back-projected back to the same plane to obtain a series of two-dimensional cross-sectional images. Here, this series of two-dimensional cross-sectional images can be used as the original scan data. Furthermore, three-dimensional reconstruction can be performed based on a series of two-dimensional cross-sectional images to obtain a three-dimensional image, and here, the three-dimensional image can also be used as the original scan data.

[0059] Take the MRI device as an example. After the patient is in position, the MRI device applies a strong uniform magnetic field to align the hydrogen nuclei (protons) in the body. Next, the coils in the MRI device emit radio waves of a specific frequency, which interact with the protons and cause them to deviate from their alignment. When the radio waves are turned off, the protons release energy and return to their original alignment. As the protons return to their original state, the energy they release is captured by the receiving coil in the form of electromagnetic signals. Different tissue types have different relaxation characteristics, so they return to their original state at different rates, which produces different signal strengths. The MRI device processes the received signals and converts them into corresponding three-dimensional images. Here, the three-dimensional image can be used as the original scan data.

[0060] The retroperitoneal space refers to an anatomical area located behind the peritoneal cavity, between the peritoneum and the posterior abdominal wall and the anterior spine. This area contains a variety of important organs and structures, such as the kidneys, adrenal glands, part of the pancreas, the abdominal aorta, the inferior vena cava, and other lymph nodes and blood vessels.

[0061] When an abdominal aortic aneurysm ruptures, a specific target will appear in the retroperitoneal space area of ​​the original scan data, wherein the specific target is the retroperitoneal bleeding location and / or the contrast agent extravasation location. Optionally, the computed tomography device can be a non-contrast computed tomography (NCCT) device or a computed tomography angiography (CTA) device. In the case where the original scan data is obtained after scanning a patient using an NCCT device, the specific target is the retroperitoneal bleeding location; in the case where the original scan data is obtained after scanning a patient using a CTA device, the specific target is the retroperitoneal bleeding location and / or the contrast agent extravasation location.

[0062] S102, the device identifies whether a data unit in the valid data has a specific target, and counts the number of valid units of the data unit that has the specific target.

[0063] In the embodiment of the present disclosure, the valid data includes multiple data units, and the data unit is a single layer image or a single rectangular unit in the original scan data.

[0064] When the data unit is a single-layer image in the original scan data, the device can obtain the single-layer image layer by layer in the valid data and identify whether there is a specific target in the single-layer image. Here, the original scan number is a series of two-dimensional cross-sectional images, and the single-layer image is a two-dimensional cross-sectional image. It can be understood that if there is a specific target in the single-layer image, the value of the number of valid units can be increased by 1.

[0065] When the data unit is a single cuboid unit in the original scan data, the device can intercept the single cuboid unit in the valid data based on the 3D sliding window to identify whether there is a specific target in the single cuboid unit. Here, the original scan number is a three-dimensional image. It can be understood that if there is a specific target in the single cuboid unit, the value of the number of valid units can be increased by 1.

[0066] S103, the device determines whether abdominal aortic aneurysm rupture occurs based on the number of effective units.

[0067] In the disclosed embodiment, the number of effective units can reflect the credibility of abdominal aortic aneurysm rupture. Generally speaking, the number of effective units is positively correlated with the credibility of abdominal aortic aneurysm rupture, that is, the larger the number of effective units, the higher the credibility of abdominal aortic aneurysm rupture. Based on the relationship between the number of effective units and the credibility of abdominal aortic aneurysm rupture, the device can accurately determine whether abdominal aortic aneurysm rupture occurs according to the number of effective units.

[0068] The abdominal aortic aneurysm rupture detection method provided by the embodiment of the present disclosure takes into account that after the rupture of the abdominal aortic aneurysm, a specific target (bleeding or extravasation of contrast agent) will appear in the retroperitoneal space. Whether the scanning data has a specific target can be used as an important basis for whether the abdominal aortic aneurysm rupture occurs. Therefore, firstly, valid data containing the retroperitoneal space is obtained in the original scanning data of the aortic aneurysm. This can narrow the data processing range to improve the subsequent detection efficiency, and can avoid false positives caused by the calculation of useless areas. On this basis, taking the data unit in the valid data as a unit, it is identified one by one whether the data unit has a specific target, and the number of valid units of the data unit with the specific target is counted. Finally, it is determined whether the abdominal aortic aneurysm rupture occurs based on the number of valid units, so as to accurately and quickly obtain the detection result of the abdominal aortic aneurysm rupture.

[0069] In some embodiments, the valid unit number is the number of data units with a specific target that are continuously identified, or the valid unit number is the total number of data units with a specific target.

[0070] In some embodiments, when the number of valid units is the number of data units with specific targets that are continuously identified, it is identified whether the data units in the valid data have specific targets, and the number of valid units of the data units with specific targets is counted, including: identifying whether the data units in the valid data have specific targets, and counting the number of valid units in real time; when data units without specific targets are identified, the number of valid units is cleared to zero, and the remaining data units are continuously identified whether they have specific targets, and the number of valid units is re-counted until the number of valid units reaches a threshold or all data units have been identified.

[0071] Combination Figure 2 As shown, the embodiment of the present disclosure provides another abdominal aortic aneurysm rupture detection method, the abdominal aortic aneurysm rupture detection method comprising:

[0072] S201, the device obtains valid data including the retroperitoneal space from the original scanning data of the abdominal aortic aneurysm.

[0073] S202, the device identifies whether the data unit in the valid data has a specific target, and counts the number of valid units in real time.

[0074] S203, when the device identifies a data unit without a specific target, it clears the number of valid units to zero, continues to identify whether the remaining data units have a specific target, and recounts the number of valid units until the number of valid units reaches a threshold or all data units have been identified.

[0075] In the embodiment of the present disclosure, the initial value of the number of valid units is zero. When the device recognizes a data unit with a specific target for the first time, the value of the number of valid units becomes 1. If the device recognizes a specific target in the next data unit, the value of the number of valid units is increased by 1, otherwise the number of valid units is cleared to zero. That is to say, in the process of identifying a specific target for a data unit, the value of the number of valid units is accumulated only when the device continuously identifies a data unit with a specific target; once a data unit with a specific target is not continuously identified, the number of valid units is cleared to zero.

[0076] In the embodiment of the present disclosure, a quantity threshold may be pre-set, and when the number of valid units reaches the quantity threshold, the identification process for a specific target of the data unit is terminated. If the number of valid units has not reached the quantity threshold, the identification process for a specific target of the data unit is terminated when all data units in the valid data have been identified. Since the quantity threshold is set, the identification process for a specific target of the data unit can be terminated when the number of valid units reaches the quantity threshold, so in some cases it is not necessary to identify all data units, which helps to save process time and improve determination efficiency.

[0077] In the embodiment of the present disclosure, when the data unit is a single-layer image in the original scan data, the device can obtain the single-layer image layer by layer in the valid data to identify whether there is a specific target in the single-layer image. The initial value of the number of valid units is zero. When the device recognizes the single-layer image with the specific target for the first time, the value of the number of valid units becomes 1. If the device recognizes the specific target in the next single-layer image, the value of the number of valid units is increased by 1, otherwise the number of valid units is cleared. That is, in the process of identifying a specific target for a single-layer image, the value of the number of valid units is accumulated only when the device continuously identifies the single-layer image with the specific target; once the single-layer image with the specific target is not continuously identified, the number of valid units is cleared. When the number of valid units reaches the number threshold, the process of identifying the specific target for the single-layer image is terminated. If the number of valid units has not reached the number threshold, the process of identifying the specific target for the single-layer image is terminated when all the single-layer images in the valid data have been identified.

[0078] In an embodiment of the present disclosure, when the data unit is a single cuboid unit in the original scan data, the device can intercept a single cuboid unit in the valid data based on the 3D sliding window to identify whether there is a specific target in the single cuboid unit. The initial value of the number of valid units is zero. When the device recognizes a single cuboid unit with a specific target for the first time, the value of the number of valid units becomes 1. If the device recognizes a specific target in the next single cuboid unit, the value of the number of valid units is increased by 1, otherwise the number of valid units is cleared. That is, in the recognition process of a specific target for a single cuboid unit, the value of the number of valid units is accumulated only when the device continuously recognizes a single cuboid unit with a specific target; once a single cuboid unit with a specific target is not continuously recognized, the number of valid units is cleared. When the number of valid units reaches the number threshold, the recognition process of the specific target for the single cuboid unit is terminated. If the number of valid units has not reached the number threshold, the recognition process of the specific target for the single cuboid unit is terminated when all the single cuboid units in the valid data have been recognized.

[0079] S204, the device determines whether abdominal aortic aneurysm rupture occurs based on the comparison result of the effective unit quantity and the quantity threshold.

[0080] In some embodiments, determining whether an abdominal aortic aneurysm rupture occurs based on the number of effective units includes: when the number of effective units does not reach a number threshold, determining that an abdominal aortic aneurysm rupture does not occur; when the number of effective units reaches a number threshold, determining that an abdominal aortic aneurysm rupture occurs.

[0081] Combination Figure 3 As shown, the embodiment of the present disclosure provides another abdominal aortic aneurysm rupture detection method, the abdominal aortic aneurysm rupture detection method comprising:

[0082] S301, the device obtains valid data including the retroperitoneal space from the original scanning data of the abdominal aortic aneurysm.

[0083] S302, the device identifies whether a data unit in the valid data has a specific target, and counts the number of valid units of the data unit that has the specific target.

[0084] S303: When the number of valid units does not reach the number threshold, the device determines that the abdominal aortic aneurysm has not ruptured.

[0085] S304: When the number of valid units reaches a threshold, the device determines that an abdominal aortic aneurysm rupture occurs.

[0086] In some embodiments, valid data including the retroperitoneal space is obtained from the original scan data of the abdominal aortic aneurysm, including: inputting the original scan data of the abdominal aortic aneurysm into a retroperitoneal space recognition model, and using the retroperitoneal space recognition model to obtain valid data including the retroperitoneal space.

[0087] In some embodiments, identifying whether a data unit in valid data has a specific target includes: inputting the data unit in the valid data into a specific target detection model, and using the target detection model to determine whether the data unit has a specific target.

[0088] Combination Figure 4 As shown, the embodiment of the present disclosure provides another abdominal aortic aneurysm rupture detection method, the abdominal aortic aneurysm rupture detection method comprising:

[0089] S401, the device inputs the original scanning data of the abdominal aortic aneurysm into the retroperitoneal space recognition model, and uses the retroperitoneal space recognition model to obtain valid data including the retroperitoneal space.

[0090] In the disclosed embodiment, the retroperitoneal space recognition model can be an artificial intelligence model with image recognition and segmentation functions, and the retroperitoneal space recognition model can identify the area of ​​the retroperitoneal space in the original scan data and obtain a valid number containing the retroperitoneal space. For example, the retroperitoneal space recognition model is a U-Net (U-shaped Network) model, a nnU-Net (Neural Network U-Net) model, or a U-Net 3D (3D U-shaped Network) model. Of course, the type of the retroperitoneal space recognition model is not limited to this.

[0091] S402, the device inputs the data unit in the valid data into the specific target detection model, uses the target detection model to determine whether the data unit has a specific target, and counts the number of valid units of the data unit that has the specific target.

[0092] In the disclosed embodiment, the specific target detection model may be an artificial intelligence model with image recognition function, and the specific target detection model may identify whether a specific target exists in the data unit of valid data. For example, the specific target detection model is a FasterRCNN (Faster Region-based Convolutional Neural Network) model, a Yolo (You Only Look Once) model, a nnDetection (Neural Network Detection, Neural Network Detection) model, a Unet model or a nnUnet model. Of course, the type of the specific target detection model is not limited thereto. When the specific target detection model is a two-dimensional (Two-Dimensional, 2D) detection model, the data unit is a single-layer image in the original scan data; when the specific target detection model is a three-dimensional (Three-Dimensional, 3D) detection model, the data unit is a single rectangular unit in the original scan data.

[0093] S403, the device determines whether abdominal aortic aneurysm rupture occurs based on the number of effective units.

[0094] In some embodiments, the retroperitoneal space recognition model is trained by marking the retroperitoneal space region in each sample original scan data, and then dividing all sample original scan data into a training set and a test set; using the training set to train the initial retroperitoneal space recognition model; after the retroperitoneal space recognition model is trained using the training set, the retroperitoneal space recognition model is tested using the test set. The retroperitoneal space recognition model is a U-Net model, a nnU-Net model, or a U-Net 3D model.

[0095] In the disclosed embodiment, the corresponding software can be used to mark the retroperitoneal space region layer by layer in each raw scan data, and save the retroperitoneal space region label. Then, all sample raw scan data are divided into a training set and a test set according to a preset ratio (such as 8:2 or 9:1).

[0096] After obtaining the training set and the test set, the original scan data in the training set and the test set are preprocessed. Specifically, the bed board and air related data in the original scan data are removed by image processing methods such as connected domain extraction and threshold segmentation, and then the values ​​of the data points in the original scan data are normalized to [0, 1] according to the appropriate window width and window position.

[0097] After preprocessing the raw scan data in the training set and the test set, the raw scan data in the training set is input into the initial retroperitoneal space recognition model after hot encoding. In the retroperitoneal space recognition model, the 2-channel data is first output. After the 2-channel data passes through softmax (normalized exponential function), the value at the input position corresponding to each channel represents the probability that the data point belongs to the channel category (the 0th channel corresponds to the background, and the 1st channel corresponds to the retroperitoneal space). Then, the output obtained after softmax is directly used to calculate the function loss, and the parameters of the retroperitoneal space recognition model are adjusted based on the channel function loss. Among them, the function loss can be cross entropy, etc.

[0098] After the training of the retroperitoneal space recognition model is completed using the training set, the original scan data in the test set is input into the retroperitoneal space recognition model to obtain the retroperitoneal space region label. The retroperitoneal space region label is multiplied by the preprocessed original scan data to obtain valid data containing the retroperitoneal space.

[0099] In the disclosed embodiment, when the data unit is a single layer image in the original scan data, the single layer image is the circumscribed rectangle of the retroperitoneal space region, and each single layer image needs to be resampled to the same size. When the data unit is a single cuboid unit in the original scan data, the single cuboid unit is the circumscribed cuboid of the retroperitoneal space region, and each single cuboid unit is isotropically processed in the x, y, and z directions.

[0100] In some embodiments, the specific target detection model is trained in the following manner: a specific target is marked in each sample valid data containing the retroperitoneal space, and then all sample valid data are divided into a training set and a test set; the training set is used to train the initial specific target detection model; after the specific target detection model is trained using the training set, the specific target detection model is tested using the test set. The specific target detection model can be a FasterRCNN model, a Yolo model, a nnDetection model, a Unet model, or a nnUnet model. Figure 5 As shown, taking the FasterRCNN model as an example, the specific target detection model includes a convolution module, an RPN (Region Proposal Network) module and a Roi Head (Region of Interest Head) module.

[0101] In the disclosed embodiment, a specific target is marked in each sample valid data containing the retroperitoneal space. Specifically, when the original scan data is obtained by scanning a patient using an NCCT device, the specific target is the location of retroperitoneal bleeding, and the range of the location of retroperitoneal bleeding can be marked layer by layer in the original scan data; when the original scan data is obtained by scanning a patient using a CTA device, the specific target is the location of retroperitoneal bleeding and / or the location of contrast agent extravasation, and the range of the location of retroperitoneal bleeding and the range of the location of contrast agent extravasation can be marked layer by layer in the original scan data. After marking the specific target in each sample valid data containing the retroperitoneal space, all sample valid data are divided into a training set and a test set according to a preset ratio (such as 8:2 or 9:1).

[0102] After preprocessing the raw scan data in the training set and the test set, the valid sample data in the training set is input into the convolution module to generate the feature map. Here, the convolution module that can be selected can be ResNet (Residual Network), Inception Network or VGG16 (Visual Geometry Group Network 16-layer version), etc.

[0103] First, nine anchors are generated based on the upper left corner of the feature map. Each anchor has three scales and three different length-to-width ratios. Then, it is mapped to the 2D image of the retroperitoneal cross section and anchors are generated. The anchors that exceed its boundaries are removed, leaving R anchors. All B target boxes with real annotations are recorded as bbox. When the intersection over union (IOU) between an anchor and the corresponding target box is greater than 0.7, the anchor is recorded as a positive sample anchor; when the intersection over union (IOU) between an anchor and the corresponding target box is less than 0.3, the anchor is recorded as a negative sample anchor. At the same time, the anchor with the maximum intersection over union ratio corresponding to each target box is also recorded as a positive sample anchor. Finally, 128 anchors are randomly selected from the positive sample anchors and the negative sample anchors, and the labels of the 128 positive sample anchors are set to 1, the labels of the 128 negative sample anchors are set to 0, and the labels of the remaining R-256 anchors are all set to 0.

[0104] After the anchor points are marked on the feature map, the feature map is input into the RPN module. Here, the RPN module has a classification branch and a regression branch.

[0105] When the original scan data is obtained by scanning the patient using an NCCT device, the output of the classification branch has a channel number of 9×2, 9 represents the 9 anchor points of each feature map, and 2 represents the probability that each anchor point is a normal area or a probability of retroperitoneal hemorrhage (p0, p1), p0 is the probability of a normal area, and p1 is the probability of retroperitoneal hemorrhage.

[0106] When the original scan data is obtained by scanning the patient using a CTA device, the output of the classification branch has a channel number of 9×3, where 9 represents the 9 anchor points of each feature map, and 3 represents the probability that each anchor point is a normal area, retroperitoneal hemorrhage, or contrast agent extravasation (p0, p1, p2), where p0 is the probability of a normal area, p1 is the probability of retroperitoneal hemorrhage, and p2 is the probability of contrast agent extravasation.

[0107] When the original scan data is obtained by scanning the patient using an NCCT device or a CTA device, the number of output channels of the regression branch is 9×4, where 9 represents the 9 anchor points of each feature map, and 4 represents the 4 position parameter prediction values ​​(x, y, w, h) of each anchor point. x and y represent the center point coordinates of the predicted anchor point, and w and h represent the width and height of the anchor point, respectively.

[0108] The first 0.6R anchor points with higher probability p1 calculated by the RPN module are selected and corrected using the position regression parameters. 0.1R regions of interest (ROI) and corresponding position parameters are selected using non-maximum suppression.

[0109] Concatenate 0.1R regions of interest and B target frames to get RoisRB, calculate the intersection and union ratio of each RoisRB and B target frames, and find which target frame has the largest intersection and union ratio between each RoisRB. The label value of each RoisRB is the label value of the target frame with the largest intersection and union ratio plus 1, that is, labelRBi=labelBi+1. Among them, labelRBi is the label value of RoisRB, and labelBi is the label value of RoisRB, which is the label value of the target frame with the largest intersection and union ratio.

[0110] Randomly select 64 RoisRBs with an IoU greater than 0.5 as positive samples, and randomly select 64 RoisRBs with an IoU less than 0.5 as negative samples, where the negative sample label labelRBi = 0. The prepared 128 RoisRBs are projected from the valid data to the corresponding area on the feature map and then input, with two branches: classification score and regression parameter.

[0111] When the original scan data is obtained by scanning the patient using an NCCT device, the classification branch can obtain prediction scores (score0, score1) for the two categories of background and retroperitoneal bleeding, where score0 is the prediction score for the background category and score1 is the prediction score for the retroperitoneal bleeding category. When the original scan data is obtained by scanning the patient using a CTA device, the classification branch can obtain prediction scores (score0, score1, score2) for the three categories of background, retroperitoneal bleeding and contrast agent extravasation, where score0 is the prediction score for the background category, score1 is the prediction score for the retroperitoneal bleeding category, and score2 is the prediction score for the contrast agent extravasation category. When the original scan data is obtained by scanning the patient using an NCCT device or a CTA device, the regression branch can obtain parameter prediction values ​​(xo, yo, wo, ho) for each position.

[0112] After the specific target detection model is trained using the training set, the valid sample data in the test set is input into the convolution module of the specific target detection model to obtain the feature map. The feature map is input into the RPN module to obtain 0.1N regions of interest, and only 0.015N regions of interest are selected after non-maximum suppression. Then, the 0.015N regions of interest are input into the Roi Head module to obtain the category scores and position parameters of the 0.015N regions of interest.

[0113] When the original scan data is obtained by scanning the patient using NCCT equipment, the score threshold of retroperitoneal bleeding is set to P, and the regions of interest with score1 less than P in the (score0, score1) obtained by the classification branch are discarded. When the original scan data is obtained by scanning the patient using CTA equipment, the score thresholds of retroperitoneal bleeding and contrast agent extravasation are set to P1 and P2 respectively, and the regions of interest with score1 less than P1 and score2 less than P2 in the (score0, score1, score2) obtained by the classification branch are discarded. After discarding the corresponding regions of interest, all remaining regions of interest after screening are recorded as RoisR.

[0114] RoisR is fine-tuned in combination with the position parameters, and the detection box Roio is obtained after maximum suppression. When the original scan data is obtained by scanning the patient using an NCCT device, the detection box Roio indicates that there may be retroperitoneal bleeding at this location; when the original scan data is obtained by scanning the patient using a CTA device, the detection box Roio indicates that there may be retroperitoneal bleeding or extravasation of contrast agent at this location.

[0115] In some embodiments, the data units in the original scan data of the abdominal aortic aneurysm are sequentially input into the retroperitoneal space recognition model, the retroperitoneal space recognition model is used to obtain the retroperitoneal space segmentation label, the retroperitoneal space segmentation label is multiplied by the data unit to obtain a new data unit, the new data unit is input into the specific target detection model, the target detection model is used to determine whether the data unit has a specific target, and the number of valid units of the data unit with the specific target is counted. The above process can be repeated. Among them, the number of valid units is the number of data units with specific targets that are continuously identified. When a data unit without a specific target is identified, the number of valid units is cleared, and the remaining data units are continuously identified to see whether they have a specific target, and the number of valid units is re-counted until the number of valid units reaches the number threshold or the data units have been identified.

[0116] In some embodiments, when the number of valid units is the number of data units with specific targets that are continuously identified, the device determines that abdominal aortic aneurysm rupture has not occurred when the number of valid units does not reach the number threshold, including: when the number of valid units does not reach the number threshold, determining a first number of valid units and a second number of valid units from the recorded number of valid units, wherein the first number of valid units and the second number of valid units are the two largest numbers of valid units; determining a first data unit corresponding to the first number of valid units and a second data unit corresponding to the second number of valid units in the valid data, taking all data units between the first data unit and the second data unit as third data units, and counting the number of third data units; calculating the ratio of data units with specific targets to data units without specific targets in the third data units, and taking the ratio as the valid unit ratio; determining whether abdominal aortic aneurysm rupture has occurred based on the number of third data units and the valid unit ratio.

[0117] In some embodiments, based on the number of third data units and the ratio of valid units, determining whether an abdominal aortic aneurysm rupture has occurred includes: when the number of third data units is less than a preset upper limit and the ratio of valid units is greater than a preset lower limit, determining that an abdominal aortic aneurysm rupture has occurred; when the number of third data units is not less than a preset upper limit or the ratio of valid units is not greater than a preset lower limit, determining that an abdominal aortic aneurysm rupture has occurred, and determining that an abdominal aortic aneurysm rupture has not occurred.

[0118] Combination Figure 6 As shown, an embodiment of the present disclosure provides an abdominal aortic aneurysm rupture detection device 600 , which includes a valid data acquisition module 601 , a specific target recognition module 602 , and a rupture result determination module 603 .

[0119] The valid data acquisition module 601 is configured to acquire valid data containing the retroperitoneal space in the original scan data of the abdominal aortic aneurysm, wherein the valid data includes a plurality of data units.

[0120] The specific target identification module 602 is configured to identify whether a data unit in valid data has a specific target, and count the number of valid units of the data unit having the specific target.

[0121] The rupture result determination module 603 is configured to determine whether the abdominal aortic aneurysm rupture occurs according to the number of valid units.

[0122] The abdominal aortic aneurysm rupture detection device provided by the embodiment of the present disclosure takes into account that after the rupture of the abdominal aortic aneurysm, a specific target (bleeding or extravasation of contrast agent) will appear in the retroperitoneal space. Whether the scanning data has a specific target can be used as an important basis for whether the abdominal aortic aneurysm rupture occurs. Therefore, firstly, valid data containing the retroperitoneal space is obtained in the original scanning data of the aortic aneurysm, so that the data processing range can be narrowed to improve the subsequent detection efficiency. On this basis, taking the data unit in the valid data as a unit, it is identified one by one whether the data unit has a specific target, and the number of valid units of the data unit with the specific target is counted. Finally, it is determined whether the abdominal aortic aneurysm rupture occurs based on the number of valid units, so as to accurately and quickly obtain the detection result of the abdominal aortic aneurysm rupture.

[0123] In some embodiments, the number of valid units is the number of data units that are continuously identified to have a specific target.

[0124] In some embodiments, the specific target identification module 602 is configured to:

[0125] Identify whether the data unit in the valid data has a specific target, and count the number of valid units in real time;

[0126] When a data unit without a specific target is identified, the number of valid units is cleared, and the remaining data units are continuously identified to see whether they have a specific target, and the number of valid units is recounted until the number of valid units reaches a threshold or all data units have been identified.

[0127] In some embodiments, the rupture result determination module 603 is configured to:

[0128] When the number of valid units does not reach the quantity threshold, it is determined that no abdominal aortic aneurysm rupture has occurred;

[0129] When the number of valid units reaches a number threshold, it is determined that abdominal aortic aneurysm rupture occurs.

[0130] In some embodiments, the data unit is a single layer image or a single cuboid unit in the original scan data. Specifically, when the specific target detection model is a two-dimensional detection model, the data unit is a single layer image in the original scan data; when the specific target detection model is a three-dimensional detection model, the data unit is a single cuboid unit in the original scan data.

[0131] In some embodiments, the specific target is a site of retroperitoneal bleeding and / or a site of contrast agent extravasation.

[0132] In some embodiments, the valid data acquisition module 601 is configured to:

[0133] The original scanning data of the abdominal aortic aneurysm is input into the retroperitoneal space recognition model, and the retroperitoneal space recognition model is used to obtain valid data including the retroperitoneal space.

[0134] In some embodiments, the retroperitoneal space recognition model is trained by:

[0135] The retroperitoneal space area is marked in each sample original scan data, and then all the sample original scan data are divided into a training set and a test set;

[0136] Using the training set to train an initial retroperitoneal space recognition model, wherein the retroperitoneal space recognition model is a U-Net model, a nnU-Net model, or a U-Net 3D model;

[0137] After the retroperitoneal space recognition model is trained using the training set, the retroperitoneal space recognition model is tested using the test set.

[0138] In some embodiments, the specific target identification module 602 is configured to: input the data unit in the valid data into the specific target detection model, and use the target detection model to determine whether the data unit contains a specific target.

[0139] In some embodiments, the specific target detection model is trained by:

[0140] Mark a specific target in each sample valid data containing the retroperitoneal space, and then divide all the sample valid data into a training set and a test set;

[0141] Using the training set to train an initial specific target detection model, where the target detection model includes a convolution module, an RPN module, and a Roi Head module;

[0142] After the specific target detection model is trained using the training set, the specific target detection model is tested using the test set.

[0143] Combination Figure 7As shown, an embodiment of the present disclosure provides an electronic device 700, which includes a processor 701 and a memory 702. Optionally, the electronic device 700 may also include a communication interface 703 and a bus 704. The processor 701, the communication interface 703, and the memory 702 may communicate with each other through the bus 704. The communication interface 703 may be used for information transmission. The processor 701 may call the logic instructions in the memory 702 to execute the abdominal aortic aneurysm rupture detection method of the above embodiment.

[0144] In addition, the logic instructions in the memory 702 described above may be implemented in the form of software functional units and when sold or used as independent products, may be stored in a computer-readable storage medium.

[0145] The memory 702 is a computer-readable storage medium that can be used to store software programs and computer executable programs, such as program instructions / modules corresponding to the method in the embodiment of the present disclosure. The processor 701 executes the function application and data processing by running the program instructions / modules stored in the memory 702, that is, the abdominal aortic aneurysm rupture detection method in the above embodiment is implemented.

[0146] The memory 702 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 702 may include a high-speed random access memory and may also include a non-volatile memory.

[0147] An embodiment of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to execute the above-mentioned abdominal aortic aneurysm rupture detection method.

[0148] The technical solution of the embodiment of the present disclosure can be embodied in the form of a software product, which is stored in a storage medium and includes one or more instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiment of the present disclosure. The aforementioned storage medium may be a non-transient storage medium, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, and other media that can store program codes.

[0149] The above description and the accompanying drawings fully illustrate the embodiments of the present disclosure so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process and other changes. The embodiments represent only possible changes. Unless explicitly required, separate components and functions are optional, and the order of operation may vary. The parts and features of some embodiments may be included in or replace the parts and features of other embodiments. Moreover, the words used in this application are only used to describe the embodiments and are not used to limit the claims. As used in the description of the embodiments and the claims, unless the context clearly indicates, the singular forms of "a", "an" and "the" are intended to include plural forms as well. Similarly, the term "and / or" as used in this application refers to any and all possible combinations of listings containing one or more associated ones. In addition, when used in the present application, the term "comprise" and its variants "comprises" and / or comprising refer to the presence of stated features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof. In the absence of further restrictions, the elements defined by the sentence "comprising a ..." do not exclude the presence of other identical elements in the process, method or device comprising the elements. In this article, each embodiment may focus on the differences from other embodiments, and the same and similar parts between the various embodiments may refer to each other. For the methods, products, etc. disclosed in the embodiments, if they correspond to the method part disclosed in the embodiments, then the relevant parts can refer to the description of the method part.

[0150] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software may depend on the specific application and design constraints of the technical solution. The technicians may use different methods for each specific application to implement the described functions, but such implementations should not be considered to exceed the scope of the embodiments of the present disclosure. The technicians may clearly understand that, for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above may refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0151] In the embodiments disclosed herein, the disclosed methods and products (including but not limited to devices, equipment, etc.) can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units can be only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to implement this embodiment. In addition, each functional unit in the embodiment of the present disclosure may be integrated in a processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit.

[0152] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

Claims

1. A method for detecting abdominal aortic aneurysm rupture, characterized in that: include: Acquire valid data including the retroperitoneal space from the original scan data of the abdominal aortic aneurysm, wherein the valid data includes a plurality of data units; Identify whether a data unit in valid data has a specific target, and count the number of valid units of data units that have the specific target; The presence of abdominal aortic aneurysm rupture was determined based on the number of effective units.

2. The method according to claim 1, characterized in that The number of valid units is the number of data units that are continuously identified to have a specific target.

3. The method according to claim 2, characterized in that Identify whether a data unit in valid data has a specific target, and count the number of valid units of data units with the specific target, including: Identify whether the data unit in the valid data has a specific target, and count the number of valid units in real time; When a data unit without a specific target is identified, the number of valid units is cleared, and the remaining data units are continuously identified to see whether they have a specific target, and the number of valid units is recounted until the number of valid units reaches a threshold or all data units have been identified.

4. The method according to claim 1, characterized in that The presence of abdominal aortic aneurysm rupture is determined based on the number of effective units, including: When the number of valid units does not reach the quantity threshold, it is determined that no abdominal aortic aneurysm rupture has occurred; When the number of valid units reaches a number threshold, it is determined that abdominal aortic aneurysm rupture occurs.

5. The method according to claim 1, characterized in that: The data unit is a single layer image or a single cuboid unit in the original scan data.

6. The method according to claim 1, characterized in that: Specific targets are sites of retroperitoneal bleeding and / or sites of contrast extravasation.

7. The method according to any one of claims 1 to 6, characterized in that In the original scan data of abdominal aortic aneurysm, valid data including the retroperitoneal space is obtained, including: The original scan data of the abdominal aortic aneurysm is input into the retroperitoneal space recognition model, and the retroperitoneal space recognition model is used to obtain valid data containing the retroperitoneal space; the retroperitoneal space recognition model is trained in the following way: the retroperitoneal space area is marked in each sample original scan data, and then all the sample original scan data are divided into a training set and a test set; the training set is used to train the initial retroperitoneal space recognition model; After the retroperitoneal space recognition model is trained using the training set, the retroperitoneal space recognition model is tested using the test set.

8. The method according to any one of claims 1 to 6, characterized in that Identify whether a data unit in valid data has a specific target, including: Input the data unit in the valid data into the specific target detection model, and use the target detection model to determine whether the data unit has a specific target; The specific target detection model is trained in the following way: a specific target is marked in each sample valid data containing the retroperitoneal space, and then all sample valid data are divided into a training set and a test set; the training set is used to train the initial specific target detection model, wherein the target detection model includes a convolution module, an RPN module and a Roi Head module; after the specific target detection model is trained using the training set, the specific target detection model is tested using the test set.

9. An abdominal aortic aneurysm rupture detection device, characterized in that: include: An effective data acquisition module is configured to acquire effective data including the retroperitoneal space in the original scan data of the abdominal aortic aneurysm, wherein the effective data includes a plurality of data units; A specific target identification module is configured to identify whether a data unit in the valid data has a specific target, and count the number of valid units of the data unit with the specific target; The rupture result determination module is configured to determine whether the abdominal aortic aneurysm rupture occurs according to the number of valid units.

10. An electronic device comprising a processor and a memory storing program instructions, characterized in that: The processor is configured to execute the abdominal aortic aneurysm rupture detection method according to any one of claims 1 to 8 when running the program instructions.

Images