Acute stage cerebral infarction detection system and method

Abstract

The invention relates to the technical field of disease analysis, and provides an acute-stage cerebral infarction detection system and method.The method comprises the steps that basic information, physical sign data and blood sample data of a brain patient are collected, and data integration is conducted to obtain integrated data; a brain scanning image is obtained, and a high-definition brain image is obtained after artifacts are removed; recognizing image semantics, performing image segmentation, evaluating a segmentation level, and obtaining a target image when a preset level is met; identifying brain lesion features by using the target image; an angiogram is collected, the brain blood vessel caliber is calculated, and the brain blood perfusion condition is analyzed; constructing a blood vessel network by using the angiogram, and analyzing a blood vessel variation degree; based on the blood vessel caliber, the blood perfusion condition and the blood vessel variation degree, analyzing the brain blood vessel lesion degree; collecting an electroencephalogram, and analyzing brain activity characteristics; and performing acute-stage cerebral infarction analysis based on the integrated data, the brain lesion features, the brain vascular lesion degree and the brain activity features. The invention aims to improve the rationality of acute cerebral infarction detection.

Description

Technical Field

[0001] The present invention relates to the technical field of disease analysis, and particularly to a detection system and method for acute cerebral infarction. Background Art

[0002] Acute cerebral infarction refers to a specific period after the onset of cerebral infarction, usually within several hours to several days after the onset. Cerebral infarction is caused by the obstruction of blood vessels in the brain, resulting in local cerebral tissue ischemia, hypoxia, necrosis and dysfunction. In the acute phase, the damage of brain cells is still progressing. If timely diagnosis and treatment can be carried out at this stage, it is possible to save the brain tissue that has not been completely necrotic, reduce the degree of neurological deficit, and improve the prognosis of patients. The symptoms of acute cerebral infarction may include sudden weakness of one limb, numbness, slurred speech, deviation of the mouth angle, blurred vision, dizziness, headache, etc. Its pathogenesis is relatively complex, and common causes include atherosclerosis, thrombosis, embolism, etc.

[0003] At present, traditional detection methods for acute cerebral infarction mainly rely on clinical symptom observation, neuroimaging examinations (such as CT, MRI, etc.) and laboratory index detection. Although these methods can help diagnose acute cerebral infarction to a certain extent, they have obvious limitations. Among them, clinical symptom observation is subjective and prone to missed diagnosis when the symptoms are atypical; although neuroimaging examinations have high accuracy, the equipment is expensive and the examination time is long, which may delay the treatment opportunity; laboratory index detection may be affected by various factors, and the specificity and sensitivity are not ideal enough, resulting in unreasonable detection of acute cerebral infarction. Summary of the Invention

[0004] The present invention provides a detection system and method for acute cerebral infarction, and its main purpose is to improve the rationality of acute cerebral infarction detection.

[0005] To achieve the above object, a detection system for acute cerebral infarction provided by the present invention includes: a data integration module, a brain feature recognition module, a contrast analysis module, a brain lesion analysis module, and a disease analysis module.

[0006] The data integration module is used to collect the basic information, physical sign data and blood sample data of brain patients, and integrate the basic information, the physical sign data and the blood sample data to obtain integrated data.

[0007] A brain feature recognition module, which is used to obtain brain scan images of a brain patient, remove artifacts from the brain scan images to obtain high-definition brain images, recognize the image semantics of the high-definition brain images, segment the high-definition brain images based on the image semantics to obtain segmented images, evaluate the segmentation level of the segmented images, and obtain target images when the segmentation level meets a preset level, and use the target images to recognize the brain lesion features of the brain patient;

[0008] An angiography analysis module, which is used to collect angiograms of a brain patient, calculate the brain blood vessel caliber of the brain patient according to the angiograms, and analyze the cerebral blood perfusion status of the brain patient according to the angiograms;

[0009] A brain lesion analysis module, which is used to analyze the degree of blood vessel variation of the brain patient by using the angiograms, and analyze the degree of brain blood vessel lesions of the brain patient based on the blood vessel caliber, the cerebral blood perfusion status and the degree of blood vessel variation;

[0010] A disease analysis module, which is used to collect the electroencephalogram of the brain patient, analyze the brain activity characteristics of the brain patient based on the electroencephalogram, and perform acute cerebral infarction analysis on the brain patient based on the integrated data, the brain lesion characteristics, the degree of brain blood vessel lesions and the brain activity characteristics to obtain an analysis report.

[0011] Optionally, the data integration of the basic information, the physical sign data and the blood sample data to obtain integrated data includes:

[0012] Perform data preprocessing on the basic information, the physical sign data and the blood sample data to obtain preprocessed data;

[0013] Perform text processing on the preprocessed data to obtain text data;

[0014] Construct a data relationship table of the text data;

[0015] Use the data relationship table to perform data integration on the text data to obtain integrated data.

[0016] Optionally, the removing artifacts from the brain scan images to obtain high-definition brain images includes:

[0017] Perform filtering processing on the brain scan images to obtain filtered images;

[0018] Adjust the brightness and contrast of the filtered images to obtain adjusted images;

[0019] Perform image reconstruction on the adjusted images to obtain high-definition brain images.

[0020] Optionally, the image segmentation of the high-definition brain image based on the image semantics to obtain a segmented image includes:

[0021] Performing corner detection on the high-definition brain image based on the image semantics to obtain detected corners;

[0022] Converting the high-definition brain image into a grayscale image;

[0023] Performing non-maximum suppression on the grayscale image to obtain a suppressed image;

[0024] Performing image segmentation on the suppressed image using the detected corners to obtain a segmented image.

[0025] Optionally, the evaluation of the segmentation level of the segmented image includes:

[0026] Calculating the mean intersection over union of the segmented image;

[0027] Evaluating the segmentation level of the segmented image according to the mean intersection over union.

[0028] Optionally, the identification of the brain lesion characteristics of the brain patient using the target image includes:

[0029] Identifying the feature points of the target image;

[0030] Querying the feature descriptors corresponding to the feature points;

[0031] Performing feature matching on the target image based on the feature descriptors to obtain matching image information;

[0032] Determining the target information of the matching image information using the feature points;

[0033] Identifying the brain lesion characteristics of the brain patient based on the target information.

[0034] Optionally, the calculation of the brain blood vessel caliber of the brain patient according to the angiogram includes:

[0035] Calculating the pixel of the blood vessel area of the angiogram and calculating the blood vessel area of the brain patient based on the pixel of the blood vessel area;

[0036] Calculating the blood vessel length of the brain patient based on the angiogram and calculating the blood vessel caliber of the brain patient using the blood vessel area and the blood vessel length.

[0037] Optionally, the analysis of the brain blood perfusion status of the brain patient according to the angiogram includes:

[0038] Query the pixel of the vascular region in the angiogram;

[0039] Based on the pixel of the vascular region, calculate the area around the blood vessels of the brain patient;

[0040] Based on the pixel of the vascular region, calculate the area of the vascular region pixel;

[0041] Based on the area around the blood vessels and the area of the vascular region pixel, calculate the vascular perfusion index of the brain patient;

[0042] Determine the cerebral blood perfusion status according to the vascular perfusion index.

[0043] Optionally, analyzing the degree of vascular variation of the brain patient using the angiogram includes:

[0044] Using the angiogram to construct the vascular network of the brain patient;

[0045] Identify the structural features of the vascular network;

[0046] Based on the structural features, analyze the degree of vascular variation of the brain patient.

[0047] A method for detecting acute cerebral infarction, the method includes:

[0048] Collect the basic information, physical sign data and blood sample data of the brain patient, integrate the basic information, the physical sign data and the blood sample data to obtain integrated data;

[0049] Obtain the brain scan image of the brain patient, remove artifacts from the brain scan image to obtain a high-definition brain image, identify the image semantics of the high-definition brain image, based on the image semantics, segment the high-definition brain image to obtain a segmented image, evaluate the segmentation level of the segmented image, when the segmentation level meets the preset level, obtain a target image, and use the target image to identify the brain lesion characteristics of the brain patient;

[0050] Collect the angiogram of the brain patient, according to the angiogram, calculate the caliber of the cerebral blood vessels of the brain patient, and analyze the cerebral blood perfusion status of the brain patient according to the angiogram;

[0051] Use the angiogram to analyze the degree of vascular variation of the brain patient, and based on the vascular caliber, the cerebral blood perfusion status and the degree of vascular variation, analyze the degree of cerebral vascular lesion of the brain patient;

[0052] Collect the electroencephalogram of the brain patient, analyze the brain activity characteristics of the brain patient based on the electroencephalogram, and perform acute cerebral infarction analysis on the brain patient based on the integrated data, the brain lesion characteristics, the degree of cerebral vascular lesion, and the brain activity characteristics to obtain an analysis report.

[0053] In the embodiments of the present invention, by collecting the basic information, physical sign data, and blood sample data of brain patients, the changes in the user's physical sign data can be monitored, and then the progress of the disease or the treatment effect can be evaluated. In the embodiments of the present invention, by identifying the image semantics of the high-definition brain images, the manifestations and mechanisms of brain diseases can be deeply understood, which can help doctors more accurately identify the disease characteristics and improve the diagnostic accuracy. In the embodiments of the present invention, by collecting the angiograms of brain patients, doctors can be helped to identify cerebrovascular abnormalities, such as aneurysms, vascular stenosis or occlusion, etc. The angiograms can be collected by digital subtraction angiography technology. In the embodiments of the present invention, by using the angiograms and analyzing the degree of vascular variation of the brain patients, it can be used to evaluate whether there is a risk of blockage in the blood vessels. In the embodiments of the present invention, by collecting the electroencephalogram of the brain patient, the spontaneous and rhythmic electrical activities of the brain cell groups of the brain patient can be understood, which helps doctors understand the functional state and neural activities of the patient's brain. Therefore, an acute cerebral infarction detection system and method provided by the embodiments of the present invention can improve the safety of acute cerebral infarction detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a functional module diagram of an acute cerebral infarction detection system provided by an embodiment of the present invention;

[0055] Figure 2 It is a schematic flowchart of an acute cerebral infarction detection method provided by an embodiment of the present invention.

[0056] The realization, functional characteristics, and advantages of the object of the present invention will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0058] In addition, the step timing in the following method embodiments is only an example and is not strictly limited.

[0059] In fact, the server devices deployed by the acute cerebral infarction detection system may consist of one or more devices. The above-mentioned acute cerebral infarction detection system can be implemented as: a business instance, a virtual machine, or a hardware device. For example, the acute cerebral infarction detection system can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, the system can be understood as a software deployed on a cloud node, which is used to provide acute cerebral infarction detection services for each client. Alternatively, the acute cerebral infarction detection system can also be implemented as a virtual machine deployed on one or more devices in a cloud node. An application software for managing each client is installed in the virtual machine. Or, the acute cerebral infarction detection system can also be implemented as a server composed of many identical or different types of hardware devices, and one or more hardware devices are set to provide acute cerebral infarction detection services for each client.

[0060] In terms of implementation form, the acute cerebral infarction detection system and the client adapt to each other. That is, if the acute cerebral infarction detection system is an application installed on a cloud service platform, then the client is a client that establishes a communication connection with this application; or if the acute cerebral infarction detection system is implemented as a website, then the client is implemented as a web page; or if the acute cerebral infarction detection system is implemented as a cloud service platform, then the client is implemented as a small program in an instant messaging application.

[0061] Refer to Figure 1 As shown, it is a functional module diagram of the acute cerebral infarction detection system provided by an embodiment of the present invention.

[0062] The acute cerebral infarction detection system 100 described in the present invention can be set in a cloud server. In terms of implementation form, it can be used as one or more service devices, or can be installed as an application on the cloud (such as the server of the power grid information operation and maintenance support decision-making system, server cluster, etc.), or can also be developed as a website. According to the functions implemented, the acute cerebral infarction detection system 100 includes a data integration module 101, a brain feature recognition module 102, a contrast analysis module 103, a brain lesion analysis module 104, and a disease analysis module 105.

[0063] In the embodiments of the present invention, in the tracking based on the detection of acute cerebral infarction, each of the above modules can be independently implemented and called by other modules. Here, the call can be understood as that a certain module can be connected to multiple modules of another type and provide corresponding services for the multiple modules it is connected to. In the acute cerebral infarction detection system provided by the embodiments of the present invention, without modifying the program code, the applicable range of the acute cerebral infarction detection architecture can be adjusted by adding modules and directly calling them, realizing cluster-level expansion, so as to achieve the purpose of quickly and flexibly expanding the acute cerebral infarction detection system. In practical applications, the above modules can be set in the same device or different devices, or can be set in virtual devices, such as service instances in a cloud server.

[0064] Next, specific embodiments will be used to illustrate each component and the specific working process of the acute cerebral infarction detection system respectively.

[0065] The data integration module 101 is used to collect the basic information, physical sign data and blood sample data of brain patients, and integrate the basic information, the physical sign data and the blood sample data to obtain integrated data.

[0066] In the embodiments of the present invention, by collecting the basic information, physical sign data and blood sample data of brain patients, the change of the user's physical sign data can be monitored, and then the progress of the disease or the treatment effect can be evaluated. Among them, the basic information refers to information such as the patient's age, gender, medical history and symptom onset time, the physical sign data refers to characteristics such as the patient's blood pressure, heart rate and body temperature, and the blood sample data includes data such as blood routine, coagulation function and blood lipid index.

[0067] Furthermore, in the embodiments of the present invention, by integrating the basic information, the physical sign data and the blood sample data to obtain integrated data, data from different sources can be integrated, providing a more comprehensive and integrated view for users, helping users better understand the business situation, make more informed decisions, and thus improving the quality and accuracy of the data.

[0068] As an embodiment of the present invention, integrating the basic information, the physical sign data and the blood sample data to obtain integrated data includes: performing data preprocessing on the basic information, the physical sign data and the blood sample data to obtain preprocessed data, performing text processing on the preprocessed data to obtain text data, constructing a data relationship table for the text data, and using the data relationship table to integrate the text data to obtain integrated data.

[0069] Among them, the data relationship table refers to a tabular form used to describe the relationship between data, which can be constructed by using an excel tool.

[0070] Optionally, the data preprocessing of the basic information, the physical sign data, and the blood sample data to obtain preprocessed data refers to a series of operations performed on the original data before data processing, such as data standardization, type conversion, etc. The preprocessing can be performed using MySQL tools. The text processing of the preprocessed data to obtain text data refers to extracting useful information from the data. The text data can be segmented into words, and the words can be translated using a bag-of-words model to obtain text data. The data integration of the text data using the data relationship table to obtain integrated data is achieved by filling the text data into the relationship data table to obtain a filled data table. When the detection result in the filled data table is normal, integrated data is obtained.

[0071] The brain feature recognition module 102 is configured to obtain a brain scan image of a brain patient, remove artifacts from the brain scan image to obtain a high-definition brain image, identify the image semantics of the high-definition brain image, based on the image semantics, segment the high-definition brain image to obtain a segmented image, evaluate the segmentation level of the segmented image, and when the segmentation level meets a preset level, obtain a target image. The brain lesion features of the brain patient are identified using the target image.

[0072] In an embodiment of the present invention, obtaining the brain scan image of the brain patient can help users or doctors analyze brain diseases more accurately. The brain scan image can be collected using a magnetic resonance imaging device.

[0073] Furthermore, in an embodiment of the present invention, removing artifacts from the brain scan image to obtain a high-definition brain image can remove noise in the image, improve image clarity, and thereby enhance the accuracy of image analysis.

[0074] As an embodiment of the present invention, removing artifacts from the brain scan image to obtain a high-definition brain image includes: performing a filtering process on the brain scan image to obtain a filtered image, adjusting the brightness and contrast of the filtered image to obtain an adjusted image, and performing image reconstruction on the adjusted image to obtain a high-definition brain image.

[0075] Optionally, filtering the brain scan image to obtain a filtered image means removing the noise influence in the image, which can be processed by a filter. Adjusting the brightness and contrast of the filtered image to obtain an adjusted image means improving the image quality by adjusting the contrast and brightness parameters of the image, which can be achieved by adjusting the device parameters of the camera acquisition device. Suppose there is a photo with an original brightness value range of 50 - 200 and an average brightness of about 125. Now we want to make it brighter and increase the contrast. We can increase the brightness to 150, which will make the whole image brighter. Then, we can use contrast stretching or histogram equalization to increase the contrast. Suppose we choose contrast stretching and stretch the bright part to 200 and the dark part to 50. This will make the bright part brighter and the dark part darker, thus enhancing the contrast of the image. The brightness range of the adjusted image becomes 50 - 200, and the average brightness is about 150, and the contrast is enhanced. It should be noted that the actual adjustment may require repeated experiments and fine-tuning according to the specific situation of the image.

[0076] Performing image reconstruction on the adjusted image to obtain a high-definition brain image can be achieved through TV regularization technology.

[0077] Furthermore, by recognizing the image semantics of the high-definition brain image in the embodiments of the present invention, the manifestations and mechanisms of brain diseases can be deeply understood, which can further help doctors more accurately identify disease characteristics and improve the diagnostic accuracy. The image semantics of the high-definition brain image can be recognized by a convolutional neural network.

[0078] Even further, by segmenting the high-definition brain image based on the image semantics in the embodiments of the present invention to obtain a segmented image, the target analysis region in the image can be extracted, thereby improving the analysis accuracy of the target region of the image.

[0079] As an embodiment of the present invention, segmenting the high-definition brain image based on the image semantics to obtain a segmented image includes: detecting corner points of the high-definition brain image based on the image semantics to obtain detected corner points, converting the high-definition brain image into a grayscale image, performing non-maximum suppression on the grayscale image to obtain a suppressed picture, and using the detected corner points to segment the suppressed picture to obtain a segmented image.

[0080] Optionally, for the corner detection of the high-definition brain image based on the image semantics, the detected corners can query the feature points of the high-definition brain image through the image semantics, and the detection is performed based on the feature points using the sift algorithm. The conversion of the high-definition brain image into a grayscale image can be achieved through the binary method. The non-maximum suppression of the grayscale image to obtain the suppressed image can be achieved through the threshold method. The use of the detected corners to perform image segmentation on the suppressed image to obtain the segmented image can be achieved by connecting the detected corners to obtain a connected region and segmenting the connected region and retaining it.

[0081] It can be understood from the embodiment of the present invention by evaluating the segmentation level of the segmented image that when the image is segmented, the degree of detail retention of the image can be understood, and excessive segmentation and loss of image details can be avoided.

[0082] As an embodiment of the present invention, the evaluation of the segmentation level of the segmented image includes: calculating the mean intersection over union (mIoU) of the segmented image by using the following:

[0083]

[0084] where, represents the mean intersection over union (mIoU), represents the number of pixels in the segmented image with the true value of class i and predicted as class j, represents the number of classes in the segmented image, represents the number of correctly predicted pixels in the segmented image, represents the number of incorrectly predicted pixels in the segmented image,

[0085] Evaluate the segmentation level of the segmented image according to the mean intersection over union (mIoU).

[0086] Among them, the mean intersection over union (mIoU) is a method for evaluating the accuracy of image segmentation. Evaluating the segmentation level of the segmented image according to the mean intersection over union (mIoU) can be obtained through the calculation result of the mean intersection over union (mIoU). If an image is segmented into multiple regions, the segmentation result is compared with the true annotation. For each region, the intersection and union with the true annotation are calculated. The intersection represents the common part of the segmentation result and the true annotation, while the union represents the sum of the two. The mean intersection over union (mIoU) is the average of the intersections and unions of all regions. A higher mean intersection over union (mIoU) value indicates that the segmentation result is closer to the true annotation and the segmentation is more accurate.

[0087] It should be noted that when the segmentation level meets the preset level, obtaining the target image means that the processing effect of the segmented image meets the requirements. The mean intersection over union (mIoU) can be set to 0.9, or it can also be set according to the real-time application scenario.

[0088] Further, in the embodiment of the present invention, by using the target image to identify the brain lesion characteristics of the brain patient, other clinical information can be combined to predict the risk of the patient developing a specific brain disease in the future.

[0089] As an embodiment of the present invention, using the target image to identify the brain lesion characteristics of the brain patient includes: identifying the feature points of the target image, querying the feature descriptors corresponding to the feature points, based on the feature descriptors, performing feature matching on the target image to obtain matching picture information, using the feature points to determine the target information of the matching picture information, and based on the target information, identifying the brain lesion characteristics of the brain patient.

[0090] Wherein, the feature descriptor refers to a special code, which is a method used in the computer field to describe the characteristics of pictures.

[0091] Optionally, querying the feature descriptors corresponding to the feature points is calculated through the descriptor calculation function in the computer vision library (such as opencv), performing feature matching on the target image based on the feature descriptors to obtain matching picture information is obtained by identifying the picture information of the feature points through the feature descriptors and then using the picture information for matching, using the feature points to determine the target information of the matching picture information is performed by matching the feature points through the particularity or uniqueness of the feature points. If the matching is successful, it indicates that the information around the feature points is the same, and then the information of the matching picture is determined. Identifying the picture features of the target image based on the target information is obtained by compiling the picture based on the target information through feature engineering.

[0092] The angiography analysis module 103 is used to collect the angiogram of the brain patient, calculate the brain blood vessel diameter of the brain patient according to the angiogram, and analyze the brain blood perfusion status of the brain patient according to the angiogram.

[0093] In the embodiment of the present invention, collecting the angiogram of the brain patient can help doctors identify cerebrovascular abnormalities, such as aneurysms, vascular stenosis or occlusion, etc. The angiogram can be collected through digital subtraction angiography technology.

[0094] Further, in the embodiment of the present invention, calculating the brain blood vessel diameter of the brain patient according to the angiogram can help doctors evaluate the brain blood supply situation of the brain patient.

[0095] As an embodiment of the present invention, calculating the brain blood vessel diameter of the brain patient according to the angiogram includes: calculating the pixel of the blood vessel area of the angiogram, and calculating the blood vessel area of the brain patient based on the pixel of the blood vessel area by using the following formula:

[0096]

[0097] Among them, represents the blood vessel area, n represents the total number of blood vessels, represents the pixels of blood vessels among the pixels of the blood vessel area, represents the point coordinates of the pixels among the pixels of the blood vessel area;

[0098] Based on the angiogram, calculate the blood vessel length of the brain patient, and calculate the blood vessel caliber of the brain patient by using the following formula based on the blood vessel area and the blood vessel length:

[0099]

[0100] Among them, represents the blood vessel caliber, represents the blood vessel area, represents the blood vessel length.

[0101] Optionally, calculating the blood vessel length of the brain patient based on the angiogram can identify the center line of the blood vessels in the angiogram through the ITK-SNAP tool and measure the length of the blood vessels along the center line.

[0102] Furthermore, through analyzing the cerebral blood perfusion status of the brain patient according to the angiogram in the embodiment of the present invention, the dynamic changes of cerebral blood flow of the brain patient can be understood, including information such as blood vessel constriction and dilation, blood flow velocity and direction, etc., which is helpful for monitoring the blood flow state and possible abnormalities.

[0103] As an embodiment of the present invention, analyzing the cerebral blood perfusion status of the brain patient according to the angiogram includes: querying the pixels of the blood vessel area of the angiogram and calculating the peripheral area of the blood vessels of the brain patient by using the following formula based on the pixels of the blood vessel area:

[0104]

[0105] Among them, represents the peripheral area of the blood vessels, represents the pixels at the edge of the blood vessels among the pixels of the blood vessel area, n represents the total number of blood vessels, represents the point coordinates of the pixels among the pixels of the blood vessel area;

[0106] Calculate the area of the area of the pixels of the blood vessel area by using the following formula based on the pixels of the blood vessel area:

[0107]

[0108] Among them, Represents the area of the region, Represents the number of pixels in the blood vessel region, Represents the total number of blood vessels, Represents the point coordinates of the pixels in the blood vessel region pixels;

[0109] Based on the blood vessel peripheral area and the combination of the blood vessel peripheral area and the region area, calculate the blood vessel perfusion index of the brain patient according to the following formula:

[0110]

[0111] Wherein, Represents the blood vessel perfusion index, Represents the blood vessel peripheral area, Represents the area of the region;

[0112] Determine the cerebral blood perfusion status according to the blood vessel perfusion index.

[0113] Optionally, the determining the cerebral blood perfusion status according to the blood vessel perfusion index is calculated by comparing the calculated value of the blood vessel perfusion index with the blood flow of normal tissues. A higher perfusion index indicates a higher blood perfusion status, and a lower perfusion index indicates a reduction in blood perfusion.

[0114] The brain lesion analysis module 104 is used to analyze the degree of blood vessel variation of the brain patient by using the angiogram, and analyze the degree of brain blood vessel lesion of the brain patient based on the blood vessel caliber, the cerebral blood perfusion status and the degree of blood vessel variation.

[0115] In the embodiment of the present invention, analyzing the degree of blood vessel variation of the brain patient by using the angiogram can be used to evaluate whether there is a risk of blockage in the blood vessel.

[0116] As an embodiment of the present invention, the analyzing the degree of blood vessel variation of the brain patient by using the angiogram includes: using the angiogram to construct a blood vessel network of the brain patient, identifying the structural characteristics of the blood vessel network, and analyzing the degree of blood vessel variation of the brain patient based on the structural characteristics.

[0117] Wherein, the blood vessel network refers to a complex network composed of different types of blood vessels in the blood vessels of the patient's brain, and can be constructed by using computer image processing technology.

[0118] Optionally, the structural features of the blood vessel network are identified by determining the length, diameter, bifurcation points, and connection methods of the blood vessels in the blood vessel network as the structural features of the blood vessel network. Based on the structural features, the degree of blood vessel variation in the brain patient is analyzed by determining whether the branching pattern of the blood vessels is abnormal, such as reduced branching, increased branching, or abnormal morphology, observing whether the course of the blood vessels is tortuous, twisted, or abnormally angled, and evaluating the dysplasia of the blood vessels, such as too small blood vessel diameter or weak blood vessel wall structure.

[0119] In an embodiment of the present invention, by analyzing the degree of cerebral blood vessel lesions in the brain patient based on the blood vessel caliber, the cerebral blood perfusion status, and the degree of blood vessel variation, it can help the user determine the type and severity of the cerebrovascular disease of the brain patient.

[0120] Optionally, the process of analyzing the degree of cerebral blood vessel lesions in the brain patient based on the blood vessel caliber, the cerebral blood perfusion status, and the degree of blood vessel variation is as follows: Combine the measurement results of the blood vessel caliber with the blood perfusion parameters. For example, severe blood vessel stenosis (significant reduction in caliber) accompanied by low perfusion (decrease in CBF) in the lesion area indicates that the blood vessel lesion has caused a significant reduction in blood flow, and the degree of the lesion is severe. Consider the influence of the type and degree of blood vessel variation on blood perfusion. Complex blood vessel variations may lead to hemodynamic changes, further aggravating the cerebral ischemia condition. For the case of multiple blood vessel lesions, comprehensively evaluate the situation of each blood vessel and their interactions. For example, stenosis or variation of multiple blood vessels may synergistically lead to more extensive and severe cerebral blood vessel lesions.

[0121] The disease analysis module 105 is configured to collect the electroencephalogram of the brain patient, analyze the brain activity characteristics of the brain patient based on the electroencephalogram, and perform acute cerebral infarction analysis on the brain patient based on the integrated data, the brain lesion characteristics, the degree of cerebral blood vessel lesions, and the brain activity characteristics to obtain an analysis report.

[0122] In an embodiment of the present invention, by collecting the electroencephalogram of the brain patient, the spontaneous and rhythmic electrical activities of the brain cell groups of the brain patient can be understood, which helps the doctor understand the functional state and neural activities of the patient's brain.

[0123] In an embodiment of the present invention, by analyzing the brain activity characteristics of the brain patient based on the electroencephalogram, the brain signal activity of the brain patient can be understood, which is convenient for analyzing whether there is a brain abnormality in the brain patient.

[0124] As an embodiment of the present invention, analyzing the brain activity characteristics of the brain patient based on the electroencephalogram includes: constructing a spectrogram of the electroencephalogram, extracting the electrical wave frequencies in different frequency ranges in the spectrogram, determining the spectral characteristics of the spectrogram based on the electrical wave frequencies, querying the amplitude, frequency, and time-domain characteristics of the spectrogram to determine the waveform pattern of the spectrogram, identifying the spatial distribution pattern of the electroencephalogram signals in the spectrogram, determining the topological distribution of the spectrogram based on the spatial distribution pattern, and analyzing the brain activity characteristics of the brain patient based on the spectral characteristics, the waveform pattern, and the topological distribution.

[0125] Among them, the spectrogram refers to an image with frequency as the abscissa and signal amplitude or power as the ordinate, representing the distribution of the signal at different frequencies, which can be constructed by combining the spectral data of the electroencephalogram through Java software. The electrical wave frequency refers to the number of times the electromagnetic wave completes periodic changes per unit time, usually expressed in Hertz (Hz). The amplitude refers to the intensity or energy of the signal at different frequencies. The frequency refers to the frequency values of each component. The time-domain characteristic is used to reflect the change of the signal over time. The amplitude, the frequency, and the time-domain characteristic are obtained by observing the spectral curve of the spectrogram.

[0126] Optionally, determining the spectral characteristics of the spectrogram based on the electrical wave frequencies is obtained by analyzing the frequencies and intensities of the alpha waves and beta waves of the electrical wave frequencies. Analyzing the brain activity characteristics of the brain patient based on the spectral characteristics, the waveform pattern, and the topological distribution is obtained by using a deep learning model to compare and analyze the spectral characteristics, the waveform pattern, and the topological distribution with the collected normal brain activity characteristics. Identifying the spatial distribution pattern of the electroencephalogram signals in the spectrogram is analyzed by the positions of the electrodes on the scalp when collecting the brain avatar of the patient.

[0127] Furthermore, in the embodiment of the present invention, by analyzing the acute cerebral infarction of the brain patient based on the integrated data, the brain lesion characteristics, the degree of brain vascular lesions, and the brain activity characteristics, the analysis report obtained can provide doctors with more comprehensive and accurate information to help confirm whether it is acute cerebral infarction and determine the specific type and location of the infarction.

[0128] Optionally, the acute cerebral infarction analysis of the brain patient based on the integrated data, the brain lesion characteristics, the degree of cerebral vascular lesions, and the brain activity characteristics to obtain an analysis report can be analyzed using the dynamic contrast method. For example, for a patient with acute cerebral infarction, an infarction focus is found in the left basal ganglia area by MRI, the angiography shows a 60% stenosis of the ipsilateral middle cerebral artery, and the electroencephalogram shows an increase in slow waves in the left cerebral hemisphere. Combining these information, it is judged that the vascular stenosis is the main cause of the infarction, and the change in the electroencephalogram is consistent with the location of the infarction focus. At the same time, combined with the patient's history of hypertension for many years and the sudden right-sided limb hemiplegia symptoms, the diagnosis is further clarified and the corresponding treatment plan is formulated.

[0129] In the embodiment of the present invention, by collecting the basic information, physical sign data, and blood sample data of the brain patient, the change of the user's physical sign data can be monitored, and then the progress of the disease or the treatment effect can be evaluated. In the embodiment of the present invention, by identifying the image semantics of the high-definition brain image, the manifestations and mechanisms of brain diseases can be deeply understood, and then it helps doctors to more accurately identify the disease characteristics and improve the diagnostic accuracy. In the embodiment of the present invention, by collecting the angiogram of the brain patient, it can help doctors identify cerebrovascular abnormalities, such as aneurysms, vascular stenosis or occlusion, etc. The angiogram can be collected by digital subtraction angiography technology. In the embodiment of the present invention, by using the angiogram and analyzing the degree of vascular variation of the brain patient, it can be used to evaluate whether there is a risk of blockage in the blood vessel. In the embodiment of the present invention, by collecting the electroencephalogram of the brain patient, the spontaneous and rhythmic electrical activities of the brain cell groups of the brain patient can be understood, which helps doctors understand the functional state and neural activities of the patient's brain. Therefore, an acute cerebral infarction detection system and method provided by the embodiment of the present invention can improve the safety of acute cerebral infarction detection.

[0130] Refer to Figure 2 As shown, it is a schematic flowchart of an acute cerebral infarction detection method provided by an embodiment of the present invention. In this embodiment, the acute cerebral infarction detection method includes:

[0131] Collect the basic information, physical sign data, and blood sample data of the brain patient, and perform data integration on the basic information, the physical sign data, and the blood sample data to obtain integrated data;

[0132] Obtain the brain scan image of the brain patient, remove the artifacts from the brain scan image to obtain a high-definition brain image, identify the image semantics of the high-definition brain image, based on the image semantics, perform image segmentation on the high-definition brain image to obtain a segmented image, evaluate the segmentation level of the segmented image, and when the segmentation level meets the preset level, obtain a target image, and use the target image to identify the brain lesion characteristics of the brain patient;

[0133] Collect angiograms of brain patients, calculate the caliber of the brain blood vessels of the brain patients according to the angiograms, and analyze the cerebral blood perfusion status of the brain patients according to the angiograms;

[0134] Use the angiograms to analyze the degree of vascular variation of the brain patients, and analyze the degree of brain vascular lesions of the brain patients based on the blood vessel caliber, the cerebral blood perfusion status, and the degree of vascular variation;

[0135] Collect the electroencephalograms of the brain patients, analyze the brain activity characteristics of the brain patients based on the electroencephalograms, and perform acute cerebral infarction analysis on the brain patients based on the integrated data, the brain lesion characteristics, the degree of brain vascular lesions, and the brain activity characteristics to obtain an analysis report.

[0136] In several embodiments provided by the present invention, it should be understood that the provided systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation.

[0137] In addition, in each embodiment of the present invention, the functional modules can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware, or in the form of a hardware plus a software functional module.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An acute cerebral infarction detection system, characterized in that: The monitoring system includes: a data integration module, a brain feature recognition module, an imaging analysis module, a brain lesion analysis module, and a disease analysis module. A data integration module, used for collecting basic information, vital sign data and blood sample data of brain patients, and integrating the basic information, the vital sign data and the blood sample data to obtain integrated data; A brain feature recognition module is used to obtain a brain scan image of a brain patient, remove artifacts from the brain scan image to obtain a high-definition brain image, identify the image semantics of the high-definition brain image, perform image segmentation on the high-definition brain image based on the image semantics to obtain a segmented image, evaluate the segmentation level of the segmented image, obtain a target image when the segmentation level meets a preset level, and use the target image to identify the brain lesion characteristics of the brain patient; Angiography analysis module, used for collecting angiography images of brain patients, calculating the caliber of brain blood vessels of the brain patients according to the angiography images, and analyzing the brain blood perfusion status of the brain patients according to the angiography images; A brain lesion analysis module, used to analyze the degree of vascular variation of the brain patient using the angiography image, and to analyze the degree of cerebral vascular lesions of the brain patient based on the vascular caliber, the brain blood perfusion status and the degree of vascular variation; The disease analysis module is used to collect the electroencephalogram (EEG) of the brain patient, analyze the brain activity characteristics of the brain patient based on the EEG, and perform acute cerebral infarction analysis on the brain patient based on the integrated data, the brain lesion characteristics, the degree of brain vascular lesions and the brain activity characteristics to obtain an analysis report.

2. An acute cerebral infarction detection system as claimed in claim 1, characterized in that: The step of integrating the basic information, the vital sign data and the blood sample data to obtain integrated data includes: Performing data preprocessing on the basic information, the vital sign data and the blood sample data to obtain preprocessed data; Performing text processing on the preprocessed data to obtain text data; Constructing a data relationship table of the text data; The text data is integrated using the data relationship table to obtain integrated data.

3. The acute cerebral infarction detection system according to claim 1, characterized in that: The step of removing artifacts from the brain scan image to obtain a high-definition brain image includes: Performing filtering processing on the brain scan image to obtain a filtered image; Adjusting the brightness and contrast of the filtered image to obtain an adjusted image; The adjusted image is reconstructed to obtain a high-definition brain image.

4. The acute cerebral infarction detection system according to claim 1, characterized in that: The step of performing image segmentation on the high-definition brain image based on the image semantics to obtain a segmented image includes: performing corner point detection on the high-definition brain image based on the image semantics to obtain detected corner points; converting the high-definition brain image into a grayscale image; Performing non-maximum suppression on the grayscale image to obtain a suppressed image; The suppressed image is segmented using the detected corner points to obtain a segmented image.

5. The acute cerebral infarction detection system according to claim 1, characterized in that: The evaluating the segmentation level of the segmented image comprises: Calculating the mean intersection and ratio of the segmented image; The segmentation level of the segmented image is evaluated according to the mean intersection union ratio.

6. The acute cerebral infarction detection system according to claim 1, characterized in that: The step of using the target image to identify the brain lesion characteristics of the brain patient includes: Identifying feature points of the target image; Querying a feature descriptor corresponding to the feature point; Based on the feature descriptor, feature matching is performed on the target image to obtain matching picture information; Determining target information of the matching picture information using the feature points; The brain lesion characteristics of the brain patient are identified based on the target information.

7. The acute cerebral infarction detection system according to claim 1, characterized in that: The step of calculating the caliber of the brain blood vessels of the brain patient according to the angiography image comprises: Calculating blood vessel region pixels of the angiography image, and calculating the blood vessel area of ​​the brain patient based on the blood vessel region pixels; The length of the blood vessels of the brain patient is calculated based on the angiography image, and the caliber of the blood vessels of the brain patient is calculated based on the blood vessel area and the blood vessel length formula.

8. The acute cerebral infarction detection system according to claim 1, characterized in that: The analyzing the cerebral blood perfusion status of the brain patient according to the angiography includes: querying blood vessel region pixels of the angiography image; Calculating the blood vessel peripheral area of ​​the brain patient based on the blood vessel region pixels; Calculating the area of ​​the blood vessel region pixels based on the blood vessel region pixels; Calculating the vascular perfusion index of the brain patient based on the blood vessel perimeter area and the blood vessel perimeter area combined with the regional area; The cerebral blood perfusion status is determined according to the vascular perfusion index.

9. The acute cerebral infarction detection system according to claim 1, characterized in that: The method of analyzing the degree of vascular variation of the brain patient by using the angiography image includes: constructing a vascular network of the brain patient using the angiography image; identifying structural features of the vascular network; Based on the structural features, the degree of vascular variation in the brain patient is analyzed.

10. A method for detecting acute cerebral infarction, characterized in that: The method comprises: Collecting basic information, vital sign data and blood sample data of the brain patient, and integrating the basic information, the vital sign data and the blood sample data to obtain integrated data; Acquire a brain scan image of a brain patient, remove artifacts from the brain scan image to obtain a high-definition brain image, identify the image semantics of the high-definition brain image, perform image segmentation on the high-definition brain image based on the image semantics to obtain a segmented image, evaluate the segmentation level of the segmented image, obtain a target image when the segmentation level meets a preset level, and use the target image to identify brain lesion characteristics of the brain patient; Acquiring angiography images of a brain patient, calculating the caliber of the brain blood vessels of the brain patient according to the angiography images, and analyzing the brain blood perfusion status of the brain patient according to the angiography images; Analyzing the degree of vascular variation of the brain patient using the angiography image, and analyzing the degree of cerebral vascular lesions of the brain patient based on the vascular caliber, the cerebral blood perfusion status and the degree of vascular variation; Collect the electroencephalogram (EEG) of the brain patient, analyze the brain activity characteristics of the brain patient based on the EEG, perform acute cerebral infarction analysis on the brain patient based on the integrated data, the brain lesion characteristics, the degree of brain vascular lesions and the brain activity characteristics, and obtain an analysis report.

Images