Carotid artery ROI positioning method and device

Abstract

The invention relates to a carotid artery ROI positioning method and device, and the method comprises the steps: obtaining a three-dimensional head and carotid artery MRI image, and carrying out the slicing of the three-dimensional head and carotid artery MRI image in a preset direction, and obtaining a slice image sequence in the preset direction; for each preset direction, inputting the slice images in the slice image sequence into a pre-trained classification model for classification to obtain the probability that each slice image belongs to the carotid artery ROI region category; generating a one-dimensional probability vector of the slice image sequence according to the probability of each slice image; performing iterative processing on the one-dimensional probability vector through a sliding window, and determining a range in which a probability mean value meets a preset condition; and determining a carotid artery ROI positioning result of the three-dimensional head and carotid artery MRI image according to the range corresponding to each preset direction. According to the technical scheme, the efficiency and accuracy of carotid artery ROI positioning can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of image processing technology, and in particular to a method and device for locating a carotid artery ROI. Background Art

[0002] Ischemic stroke is characterized by high morbidity, disability and mortality. Rupture of carotid atherosclerotic plaque is an important cause of ischemic stroke. Therefore, early detection of the stability of carotid atherosclerotic plaque can effectively prevent the occurrence of ischemic stroke.

[0003] Imaging is an important means of evaluating carotid artery plaques. At present, high-resolution MR wall imaging of the carotid artery is an examination technology for non-invasive evaluation of carotid artery plaques. It directly displays the structure of the carotid artery wall by suppressing the blood flow signal in the lumen and the fat signal around the blood vessels, and clarifies the lumen morphology and submillimeter plaque characteristics, which are well consistent with histopathology. Due to the huge amount of high-resolution MRI image data, manual judgment is not only time-consuming and labor-intensive, but also easily affected by subjective factors. Therefore, the use of artificial intelligence (AI) technology has become an important means to improve diagnostic efficiency and accuracy, and the first step of AI technology is to accurately locate and segment plaques, thereby providing a reliable basis for further analysis and evaluation. However, carotid artery plaques only occupy a small part of the original high-resolution MRI data of the carotid artery. How to accurately crop the region of interest (Region of Interest) of the carotid artery wall from the complete image is a difficult problem. Plaque localization using ROI (region of interest) has become a key challenge in clinical practice and research.

[0004] In related technologies, carotid artery ROI extraction methods are usually combined with deep learning models for processing. For example, cascade segmentation networks, target detection, etc. are used to extract ROI. Due to the limited size of the data set, uneven data distribution, and imbalanced sample annotation, deep learning-based methods are prone to overfitting, resulting in missed detections or incorrect judgments. In addition, existing methods usually require cascading multiple models or introducing complex target detection structures, which makes the processing flow cumbersome and difficult to efficiently adapt to clinical work needs. Summary of the Invention

[0005] In order to solve the above technical problems, the present disclosure provides a carotid artery ROI positioning method and device.

[0006] In a first aspect, an embodiment of the present disclosure provides a method for locating a carotid artery ROI, comprising:

[0007] Acquire a three-dimensional MRI image of the head and neck arteries, and slice the three-dimensional MRI image of the head and neck arteries in a preset direction to obtain a slice image sequence in the preset direction;

[0008] For each preset direction, the slice images in the slice image sequence are input into a pre-trained classification model for classification, and the probability of each slice image belonging to the carotid artery ROI region category is obtained;

[0009] generating a one-dimensional probability vector of the slice image sequence according to the probability of each slice image;

[0010] Iteratively processing the one-dimensional probability vector by a sliding window to determine a range in which the probability mean satisfies a preset condition;

[0011] The carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is determined according to the range corresponding to each preset direction.

[0012] In a second aspect, an embodiment of the present disclosure provides a carotid artery ROI positioning device, comprising:

[0013] an acquisition module, configured to acquire a three-dimensional MRI image of the head and neck arteries, and slice the three-dimensional MRI image of the head and neck arteries in a preset direction to obtain a sequence of slice images in the preset direction;

[0014] a classification module, configured to input the slice images in the slice image sequence into a pre-trained classification model for classification for each preset direction, and obtain a probability that each slice image belongs to the carotid artery ROI region category;

[0015] A generating module, configured to generate a one-dimensional probability vector of the slice image sequence according to the probability of each slice image;

[0016] a determination module, configured to perform iterative processing on the one-dimensional probability vector through a sliding window to determine a range in which the probability mean satisfies a preset condition;

[0017] The positioning module is used to determine the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image according to the range corresponding to each preset direction.

[0018] The technical solution provided by the embodiments of the present disclosure has the following advantages over the prior art: the probability that the slice images in each direction belong to the carotid artery ROI area category is determined through a classification model, and on the basis of the one-dimensional probability vector of the slice image sequence, a sliding window is used to determine the range in which the probability mean satisfies the preset conditions on the one-dimensional probability vector, and then the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is obtained according to the range of positioning in each direction. Therefore, for the situation where data imbalance causes the model to overfit and miss one side, the method performs ROI positioning through the probability mean of the sliding window on the basis of the one-dimensional probability vector of the slice image sequence, which can accurately locate the left and right carotid artery ROI areas, improves the accuracy of carotid artery ROI positioning, reduces missed detections caused by data imbalance, and improves processing efficiency and carotid artery ROI positioning efficiency compared to the target detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0020] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A schematic flow chart of a method for locating a carotid artery ROI provided in an embodiment of the present disclosure;

[0022] Figure 2 A schematic diagram of a one-dimensional probability vector provided by an embodiment of the present disclosure;

[0023] Figure 3 A schematic diagram of a positioning process provided by an embodiment of the present disclosure;

[0024] Figure 4 A schematic structural diagram of a carotid artery ROI positioning device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0026] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0027] Figure 1 This is a flow chart of a carotid artery ROI positioning method provided in an embodiment of the present disclosure. The method provided in an embodiment of the present disclosure can be performed by a carotid artery ROI positioning device, which can be implemented using software and / or hardware and can be integrated into any electronic device with computing capabilities.

[0028] like Figure 1 As shown, the carotid artery ROI positioning method provided by the embodiment of the present disclosure may include:

[0029] Step 101 : Acquire a three-dimensional MRI image of the head and neck arteries, and slice the three-dimensional MRI image of the head and neck arteries in a preset direction to obtain a slice image sequence in the preset direction.

[0030] In this embodiment, a three-dimensional head and neck artery MRI (Magnetic Resonance Imaging) image is composed of multiple sequence images. Optionally, the three-dimensional head and neck artery MRI image includes four sequence images. The three-dimensional head and neck artery MRI image is preprocessed, and then, for any preset direction, the three-dimensional head and neck artery MRI image is sliced ​​in the preset direction. After slicing, a slice image sequence in the preset direction is obtained. The slice image sequence includes multiple slice images, each slice image being a multi-modal overlapping image, for example, a quad-modal overlapping image.

[0031] As an example, a three-dimensional head and neck artery MRI image includes a T1WI image (T1 weighted image), a T2WI image (T2 weighted image), a T1WI enhanced image and a time of flight (TOF) image. Before slicing the three-dimensional head and neck artery MRI image in a preset direction, preprocessing includes: performing isotropic interpolation processing on the three-dimensional head and neck artery MRI image to make the above four sequence images consistent in spatial resolution.

[0032] Among them, the preset directions can be set as needed. For example, the preset directions include axial, sagittal and coronal. Slices are taken in the three directions of axial, sagittal and coronal to generate two-dimensional image data, providing multi-view input for the deep learning model.

[0033] Optionally, before slicing the three-dimensional head and neck artery MRI images in a preset direction, the three-dimensional head and neck artery MRI images are subjected to quantile normalization processing. The upper limit of the quantile normalization processing is 99 and the lower limit is 1. By applying the normalization method to the three-dimensional head and neck artery MRI images, the consistency of the numerical distribution between different modalities and slices is ensured, and the convergence speed and robustness of the model are improved.

[0034] Step 102: For each preset direction, the slice images in the slice image sequence are input into a pre-trained classification model for classification to obtain the probability that each slice image belongs to the carotid artery ROI region category.

[0035] Step 103: Generate a one-dimensional probability vector of the slice image sequence according to the probability of each slice image.

[0036] In this embodiment, a corresponding classification model is trained for each preset direction. The input of the classification model is a slice image, and the output is a classification result of the slice image. The classification result indicates whether the slice image belongs to the carotid artery ROI area category, and specifically can be the probability that the slice image contains the carotid artery ROI area. Thus, for N slice images in any preset direction, N probabilities are obtained, and the N probabilities generate a one-dimensional probability vector according to the slice order.

[0037] The training process of the classification model includes: obtaining training data, which includes positive samples and negative samples. Positive samples are slices containing carotid artery walls or atherosclerotic plaques, and those that do not are negative samples; based on the training data, the preset classification network is trained to obtain a classification model. Optionally, a pre-trained ResNet is used as the basic classification network to perform classification learning on the generated two-dimensional slice images. The positive samples are any slice images containing carotid artery walls or atherosclerotic plaques. In order to address the imbalance problem of the positive and negative sample ratios, the loss function of the model training adopts the focal loss (FocalLoss) function to enhance the model's learning ability for minority classes and reduce the missed detection rate.

[0038] It should be noted that the implementation method of the above-mentioned classification network is not limited to ResNet, and other deep learning models such as Transformer can also be used for classification, which is not limited here.

[0039] As an example, for the three directions of axial, sagittal and coronal, the classification models corresponding to the axial direction, the classification models corresponding to the sagittal direction, and the classification models corresponding to the coronal direction are trained respectively. Then, for each direction, the slice images of that direction are classified by the corresponding classification model to obtain the probability of each slice image and generate a one-dimensional probability vector for that direction.

[0040] Step 104 , performing iterative processing on the one-dimensional probability vector through a sliding window to determine a range in which the probability mean satisfies a preset condition.

[0041] In this embodiment, a sliding window of a specified size is slid on a one-dimensional probability vector to calculate the probability mean within the window, and the positioning range is determined on the one-dimensional probability vector based on the probability mean within the window and a preset condition, wherein the probability on the one-dimensional probability vector represents the probability that the corresponding slice image belongs to the carotid artery ROI area category. Optionally, the probability mean satisfies the preset condition including the maximum probability mean.

[0042] The positioning process is described below.

[0043] In one embodiment of the present disclosure, iterative processing is performed on a one-dimensional probability vector through a sliding window to determine a range in which the probability mean satisfies a preset condition, including: determining a first range with a maximum probability mean on the one-dimensional probability vector through a first sliding window of a first size; determining a second range with a maximum probability mean within the first range of the one-dimensional probability vector through a second sliding window of a second size; roughly locating the carotid artery region within the second range of the one-dimensional probability vector through a third sliding window of a third size to obtain a rough positioning range of the carotid artery on each side; and determining a range with a maximum probability mean within the rough positioning range of the one-dimensional probability vector through a fourth sliding window of a fourth size. In the carotid artery ROI positioning scenario, carotid artery sample annotation involves marking the plaque portion. However, the location and length of plaques vary between patients. For example, when plaques are distributed in the z-direction on the same side, one patient's plaque length may be 7 while another's may be 80. The uneven distribution of the markings often causes the model to overfit to one carotid artery. Traditional methods would miss detections at this time. This method, based on the anatomical symmetry of the left and right carotid arteries, first uses a longer, non-overlapping sliding window to constrain the range of the left and right carotid arteries, and then uses a smaller sliding window to accurately locate each carotid artery within the coarse positioning range.

[0044] Among them, the first size is larger than the second size, the second size is larger than the third size, and the third size is larger than the fourth size. Optionally, the first size and the second size are determined by statistically analyzing the spatial distribution characteristics of ROIs in a large number of segmentation probability maps to ensure that sufficient potential ROI areas can be covered; then, on the one-dimensional probability vector, a non-overlapping sliding window of the third size is used to roughly locate the left and right carotid artery regions. This process is mainly used to exclude low-confidence areas and quickly narrow the search space. The coarse positioning range can be one or two, and the number of coarse positioning ranges is determined according to the direction; iterative processing, based on the coarse positioning results, further perform a fourth-size sliding window process within the coarse positioning range of each carotid artery. This step can be iteratively performed to gradually optimize the final ROI positioning result according to the context probability local peak information, effectively avoid missing important structural areas, improve positioning accuracy and integrity, and further accurately locate the ROI areas of the left and right carotid arteries, and can reduce redundant candidate areas to ensure the accuracy and reliability of ROI positioning.

[0045] The following is an example of a one-dimensional probability vector diagram. The diagram of a one-dimensional probability vector in a preset direction is as follows Figure 2 It should be noted that, Figure 2 The length of the one-dimensional probability vector and the size of each sliding window are only examples and are not limited here. Figure 2 Mark 22 is a second sliding window of a second size, the first sliding window is not shown, mark 23 is a third sliding window of a third size, mark 24 is a fourth sliding window of a fourth size, and mark 25 is a range determined on the one-dimensional probability vector.

[0046] As an example, taking the X direction as an example, the length of the one-dimensional probability vector is 512, the first dimension is 320, and the second dimension is 240. To obtain carotid artery statistics for a sample, 320 is determined by the range of the statistics for all persons in the sample, 240 is determined based on the length of the statistics for a single person in the sample, the third dimension is 100, and the fourth dimension is 64. A first sliding window is used to determine a first range of 320 in length on the one-dimensional probability vector. A second sliding window is used to determine a second range of 240 in length within the first range. A third sliding window is then used to determine a coarse positioning range of 100 within the second range. Furthermore, two fourth sliding windows are used within each coarse positioning range to determine the ROI positioning range in the X direction. The constraint condition in this step also includes an intersection-over-union ratio between the two fourth sliding windows, e.g., an intersection-over-union ratio between the two fourth sliding windows of less than or equal to 0.5.

[0047] Optionally, the carotid artery region is roughly located within the second range of the one-dimensional probability vector to obtain a coarse positioning range for each carotid artery, including: when the number of coarse positioning ranges corresponding to the preset direction is two, combined with the constraint of the intersection-over-union ratio between the two third sliding windows, two coarse positioning ranges with the largest probability mean within the second range are determined using the two third sliding windows, and the intersection-over-union ratio between the two third sliding windows is greater than a threshold and the threshold is a negative number. The threshold can be determined based on experiments. By setting the intersection-over-union ratio between the two third sliding windows to be greater than a certain negative value, such as -0.25, a certain distance is constrained between the two coarse positioning ranges to avoid including the area between the left and right carotid arteries that is easily misdetected in the coarse positioning range, thereby further improving the accuracy of carotid artery positioning.

[0048] It should be noted that the above-mentioned implementation method of determining the range on the one-dimensional probability vector is improved by adopting the concept of NMS (Non-Maximum Suppression). ROI positioning is achieved by iterative NMS processing on the one-dimensional probability vector through sliding windows of different sizes. Optionally, other strategies can also be introduced for implementation, which is not specifically limited here.

[0049] Step 105 : determining the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image according to the range corresponding to each preset direction.

[0050] In this embodiment, after the ROI positioning range of each preset direction is determined, the ROI positioning range of each preset direction can be converted into a carotid artery ROI positioning result of a three-dimensional head and neck artery MRI image in combination with the preset direction.

[0051] As an example, taking the axial, sagittal and coronary planes as examples, the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is determined according to the range corresponding to each preset direction, including: determining the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image according to the range corresponding to the axial position, the range corresponding to the sagittal position and the range corresponding to the coronary position. In this example, the Cartesian product is calculated by the range corresponding to the axial position, the range corresponding to the sagittal position and the range corresponding to the coronary position to determine the carotid artery ROI positioning result in the three-dimensional head and neck artery MRI image. The processing flow of actual application is, for example Figure 3 As shown, Figure 3 Cartesian product, Hadamard product, and iterative NMS.

[0052] In one embodiment of the present disclosure, after determining the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image, the method further includes: automatically cropping the three-dimensional head and neck artery MRI image according to the carotid artery ROI positioning result to obtain the carotid artery ROI image cropping result; visually displaying the carotid artery ROI image cropping result, and the form of the visual display includes a multimodal overlapping image and a slice comparison image, which is convenient for intuitively verifying the positioning accuracy and medical relevance of the ROI area. In this embodiment, by outputting the cropped high-quality ROI, these ROI areas are accurately positioned and processed and can be used for subsequent medical research or clinical diagnosis, providing doctors with high-precision imaging data support, and assisting in early diagnosis of the disease and treatment decision-making.

[0053] According to the technical solution of the embodiment of the present disclosure, the probability that the slice images in each direction belong to the carotid artery ROI area category is determined by a classification model. On the basis of the one-dimensional probability vector of the slice image sequence, a sliding window is used to determine the range in which the probability mean satisfies the preset conditions on the one-dimensional probability vector, and then the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is obtained according to the range of positioning in each direction. Therefore, for the situation where the data imbalance causes the model to overfit and miss one side, this method performs ROI positioning by the probability mean of the sliding window on the basis of the one-dimensional probability vector of the slice image sequence, and can accurately locate the left and right carotid artery ROI areas. This method improves the accuracy of carotid artery ROI positioning and reduces missed detections due to data imbalance. Compared with target detection methods, it does not require a complex target detection model structure, which improves processing efficiency and the efficiency of carotid artery ROI positioning. This allows for the stable generation of ROI regions covering key carotid artery structures (including the vessel wall and lesion areas) without missed detection, effectively reducing background interference containing a large amount of irrelevant tissue. Compared with traditional methods, this process not only improves positioning accuracy, but also significantly reduces the interference of invalid information in subsequent image analysis, helping to improve the model's performance in downstream tasks (such as segmentation and quantitative analysis). In addition, this method is applicable to multi-sequence data of high-resolution MRI of the head and neck arteries and can be extended to other related clinical positioning tasks, showing good versatility.

[0054] Figure 4 This is a structural diagram of a carotid artery ROI positioning device provided by an embodiment of the present disclosure, as shown in FIG. Figure 4 As shown, the carotid artery ROI positioning device includes: an acquisition module 41 , a classification module 42 , a generation module 43 , a determination module 44 , and a positioning module 45 .

[0055] An acquisition module 41 is configured to acquire a three-dimensional MRI image of the head and neck arteries and slice the three-dimensional MRI image of the head and neck arteries in a preset direction to obtain a sequence of slice images in the preset direction;

[0056] A classification module 42 is configured to input the slice images in the slice image sequence into a pre-trained classification model for classification for each preset direction, and obtain a probability that each slice image belongs to the carotid artery ROI region category;

[0057] A generating module 43 is used to generate a one-dimensional probability vector of the slice image sequence according to the probability of each slice image;

[0058] A determination module 44 is configured to perform iterative processing on the one-dimensional probability vector through a sliding window to determine a range in which the probability mean satisfies a preset condition;

[0059] The positioning module 45 is used to determine the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image according to the range corresponding to each preset direction.

[0060] In one embodiment of the present disclosure, the preset directions include axial, sagittal, and coronal directions, and the positioning module 45 is specifically configured to:

[0061] The carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is determined according to the range corresponding to the axial position, the range corresponding to the sagittal position, and the range corresponding to the coronal position.

[0062] In one embodiment of the present disclosure, the apparatus further comprises:

[0063] A training module is used to obtain training data; the training data includes positive samples and negative samples, and the positive samples are slices containing carotid artery walls or atherosclerotic plaques;

[0064] The preset classification network is trained based on the training data to obtain a classification model.

[0065] In one embodiment of the present disclosure, the loss function for model training adopts a focal loss function.

[0066] In one embodiment of the present disclosure, the determination module 44 is specifically configured to:

[0067] Determining, on the one-dimensional probability vector, a first range having a maximum probability mean by using a first sliding window of a first size;

[0068] Determining a second range having a maximum probability mean within the first range of the one-dimensional probability vector using a second sliding window of a second size, wherein the second size is smaller than the first size;

[0069] Roughly locating the carotid artery region within the second range of the one-dimensional probability vector using a third sliding window of a third size, to obtain a rough positioning range of each carotid artery; the third size is smaller than the second size;

[0070] A range with a maximum probability mean is determined within the coarse positioning range of the one-dimensional probability vector through a fourth sliding window of a fourth size; the fourth size is smaller than the third size.

[0071] In one embodiment of the present disclosure, the determination module 44 is specifically configured to:

[0072] When the number of coarse positioning ranges is two, an intersection-over-union ratio between the two third sliding windows is greater than a threshold and the threshold is a negative number.

[0073] In one embodiment of the present disclosure, the three-dimensional head and neck artery MRI images include T1WI images, T2WI images, T1WI enhanced images, and time-of-flight images, and the apparatus further includes:

[0074] The preprocessing module is used to perform isotropic interpolation processing on three-dimensional head and neck artery MRI images to make the spatial resolution meet the consistency condition.

[0075] In one embodiment of the present disclosure, the apparatus further comprises:

[0076] The normalization module is used to perform quantile normalization processing on three-dimensional head and neck artery MRI images; wherein the upper limit of the quantile normalization processing is 99 and the lower limit is 1.

[0077] In one embodiment of the present disclosure, the apparatus further comprises:

[0078] The display module is used to automatically crop the three-dimensional head and neck artery MRI image according to the carotid artery ROI positioning result to obtain the carotid artery ROI image cropping result;

[0079] The carotid artery ROI image cropping results are visualized; the visualization forms include multimodal overlap images and slice comparison images.

[0080] The carotid artery ROI positioning device provided in the embodiments of the present disclosure can execute any carotid artery ROI positioning method provided in the embodiments of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method. For details not fully described in the embodiments of the present disclosure, please refer to the description of any method embodiment of the present disclosure.

[0081] The present disclosure also provides an electronic device, which includes one or more processors and a memory. The processor can be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and can control other components in the electronic device to perform desired functions. The memory can include one or more computer program products, and the computer program product can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory can include, for example, random access memory (RAM) and / or cache memory (cache), etc. Non-volatile memory can include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions can be stored on the computer-readable storage medium, and the processor can run the program instructions to implement the method of the embodiment of the present disclosure above and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. can also be stored in the computer-readable storage medium.

[0082] In one example, the electronic device may further include an input device and an output device, these components being interconnected via a bus system and / or other forms of connection mechanisms. Furthermore, the input device may include, for example, a keyboard, a mouse, and the like. The output device may output various information to the outside, including determined distance information, direction information, and the like. The output device may include, for example, a display, a speaker, a printer, a communication network, and remote output devices connected thereto. Furthermore, depending on the specific application, the electronic device may further include any other appropriate components, such as a bus, an input / output interface, and the like.

[0083] In addition to the above methods and devices, the embodiments of the present disclosure may also be a computer program product, which includes computer program instructions. When the computer program instructions are executed by a processor, the processor executes any method provided by the embodiments of the present disclosure.

[0084] The computer program product may be written in any combination of one or more programming languages ​​to implement the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's computing device, as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0085] In addition, the embodiments of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the processor is enabled to perform any method provided by the embodiments of the present disclosure.

[0086] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include but is not limited to a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0087] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0088] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for locating a carotid artery ROI, characterized in that: The method comprises: Acquire a three-dimensional MRI image of the head and neck arteries, and slice the three-dimensional MRI image of the head and neck arteries in a preset direction to obtain a slice image sequence in the preset direction; For each preset direction, the slice images in the slice image sequence are input into a pre-trained classification model for classification, and the probability of each slice image belonging to the carotid artery ROI region category is obtained; generating a one-dimensional probability vector of the slice image sequence according to the probability of each slice image; Iteratively processing the one-dimensional probability vector by a sliding window to determine a range in which the probability mean satisfies a preset condition; The carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is determined according to the range corresponding to each preset direction.

2. The method according to claim 1, wherein The preset directions include axial, sagittal, and coronal directions, and determining the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image according to the range corresponding to each preset direction includes: The carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image is determined according to the range corresponding to the axial position, the range corresponding to the sagittal position, and the range corresponding to the coronal position.

3. The method according to claim 1, wherein The method further comprises: Acquire training data; the training data includes positive samples and negative samples, the positive samples are slices containing carotid artery walls or atherosclerotic plaques; Model training is performed on a preset classification network based on the training data to obtain the classification model.

4. The method according to claim 3, wherein The loss function for model training adopts the focal loss function.

5. The method according to claim 1, wherein The iterative processing is performed on the one-dimensional probability vector by using a sliding window to determine a range in which the probability mean satisfies a preset condition, including: Determining, on the one-dimensional probability vector, a first range having a maximum probability mean by using a first sliding window of a first size; Determining, by a second sliding window of a second size, a second range having a maximum probability mean within the first range of the one-dimensional probability vector, wherein the second size is smaller than the first size; Roughly locating the carotid artery region within the second range of the one-dimensional probability vector using a third sliding window of a third size, to obtain a rough positioning range of each carotid artery; wherein the third size is smaller than the second size; A range with a maximum probability mean is determined within the coarse positioning range of the one-dimensional probability vector through a fourth sliding window of a fourth size; the fourth size is smaller than the third size.

6. The method according to claim 5, wherein The roughly locating the carotid artery region within the second range of the one-dimensional probability vector to obtain a rough positioning range of each carotid artery includes: When the number of the coarse positioning ranges is two, the intersection-over-union ratio between the two third sliding windows is greater than a threshold and the threshold is a negative number.

7. The method according to claim 1, wherein The three-dimensional head and neck artery MRI image includes a T1WI image, a T2WI image, a T1WI enhanced image, and a time-of-flight image. Before slicing the three-dimensional head and neck artery MRI image in a preset direction, the method further includes: The three-dimensional head and neck artery MRI image is subjected to isotropic interpolation processing so that the spatial resolution satisfies the consistency condition.

8. The method according to claim 7, wherein Before slicing the three-dimensional head and neck artery MRI image in a preset direction, the method further includes: The three-dimensional head and neck artery MRI image is subjected to quantile normalization processing; wherein the upper limit of the quantile normalization processing is 99 and the lower limit is 1.

9. The method according to claim 1, wherein After determining the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image, the method further includes: Automatically cropping the three-dimensional head and neck artery MRI image according to the carotid artery ROI positioning result to obtain a carotid artery ROI image cropping result; The carotid artery ROI image cropping result is visualized; the visualization display includes a multimodal overlapping image and a slice comparison image.

10. A carotid artery ROI positioning device, characterized in that: include: an acquisition module, configured to acquire a three-dimensional MRI image of the head and neck arteries, and slice the three-dimensional MRI image of the head and neck arteries in a preset direction to obtain a sequence of slice images in the preset direction; a classification module, configured to input the slice images in the slice image sequence into a pre-trained classification model for classification for each preset direction, and obtain a probability that each slice image belongs to the carotid artery ROI region category; A generating module, configured to generate a one-dimensional probability vector of the slice image sequence according to the probability of each slice image; a determination module, configured to perform iterative processing on the one-dimensional probability vector through a sliding window to determine a range in which the probability mean satisfies a preset condition; The positioning module is used to determine the carotid artery ROI positioning result of the three-dimensional head and neck artery MRI image according to the range corresponding to each preset direction.

Images