Aortic dissection rupture risk early warning method based on CTA images

Abstract

This invention relates to the field of medical data processing technology, specifically to a method for early warning of aortic dissection rupture risk based on CTA images. The method includes: determining the abnormal blood pressure trend of a target patient based on the distribution of systolic blood pressure within a preset historical time period, and predicting potential high-pressure time intervals; acquiring CTA images of the target patient, determining early warning indicators for precursory ruptures, the spread of new ruptures, and the correlation between blood pressure and images; and, based on the target patient's systolic blood pressure, overall correlation coefficient, and current CTA images within the current time period, using a pre-trained aortic dissection rupture risk warning network to achieve early warning of aortic dissection rupture risk. This invention combines systolic blood pressure, overall correlation coefficient, and current CTA images within the current time period when using the network, thereby enriching the richness of the network input data and improving the accuracy of aortic dissection rupture risk warning.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, specifically to a method for early warning of aortic dissection rupture risk based on CTA images. Background Technology

[0002] With the development of technology, neural networks are being applied more and more widely. For example, they can be used for early warning of aortic dissection rupture risk. Specifically, based on CTA (Computed Tomography Angiography) images of aortic dissection patients acquired at the current moment, neural networks are used to achieve early warning of aortic dissection rupture risk.

[0003] However, when considering only CTA images of aortic dissection patients acquired at the current moment and using neural networks to achieve early warning of aortic dissection rupture risk, the following technical problems often arise:

[0004] CTA images of aortic dissection patients acquired at the current moment can only show the appearance of the aortic dissection at that moment. However, the risk of aortic dissection rupture is not only related to the appearance of the aortic dissection. Therefore, when making aortic dissection rupture risk warnings, if only CTA images of aortic dissection patients acquired at the current moment are considered, the richness of the data input to the neural network will often be poor, resulting in poor accuracy of aortic dissection rupture risk warnings. Summary of the Invention

[0005] To address the technical problem of poor accuracy in aortic dissection rupture risk warning due to the limited richness of data input to the neural network, this invention proposes an aortic dissection rupture risk warning method based on CTA images.

[0006] In a first aspect, the present invention provides a method for early warning of aortic dissection rupture risk based on CTA images, the method comprising:

[0007] Based on the distribution of systolic blood pressure of the target patient within a preset historical time period, determine the degree of abnormal blood pressure trend of the target patient and predict the possible high blood pressure time interval of the target patient.

[0008] Acquire all historical CTA images of the target patient, acquire the current CTA images corresponding to the target patient based on the possible high-pressure time intervals within the current cycle period, and determine the early warning indicators of the target patient's aortic rupture based on the aortic diameter in the current CTA images.

[0009] Based on the differences between the current CTA images and the ruptures in the previous CTA images, the extent of new rupture spread in the target patient is determined.

[0010] Based on the early warning indicators of premature rupture, the spread of new ruptures, and the trend of abnormal blood pressure, the correlation between blood pressure images and current CTA images is determined. Similarly, the correlation between blood pressure images and historically acquired CTA images of the target patient is determined.

[0011] Based on the correlation of all blood pressure images, the overall correlation coefficient corresponding to the target patient is determined. Based on the target patient's systolic blood pressure, overall correlation coefficient and current CTA images in the current cycle period, the risk warning of aortic dissection rupture is realized through a pre-trained aortic dissection rupture risk warning network.

[0012] In conjunction with the first aspect above, in one possible implementation, determining the blood pressure abnormality trend performance of the target patient based on the distribution of systolic blood pressure within a preset historical time period includes:

[0013] If the systolic blood pressure of the target patient collected within the preset historical time period is greater than or equal to the preset systolic blood pressure threshold, then the collection time corresponding to that systolic blood pressure is determined as the high blood pressure time.

[0014] The preset historical time period is periodically divided to obtain historical periodic time periods. The consecutive high-pressure moments within the historical periodic time period are then used to form historical high-pressure time periods, resulting in a set of historical high-pressure time periods corresponding to each historical periodic time period.

[0015] The total number of all high-pressure moments within the historical cycle period is determined as the representative value of the high-pressure duration corresponding to the historical cycle period, and the average of the representative values ​​of the high-pressure duration corresponding to all historical cycle periods is determined as the stage high-pressure duration value corresponding to the target patient.

[0016] Based on the set of historical high-pressure periods corresponding to all historical cycle periods, the recent continuous concurrent high-pressure performance of the target patient is determined.

[0017] The sustained hypertension performance of the target patient is determined based on the duration of hypertension at the corresponding stage and the recent sustained hypertension performance during the same period.

[0018] Based on the set of historical high-pressure periods corresponding to all historical cycle periods, the short-term high-pressure fluctuations corresponding to the target patient are determined;

[0019] Based on the sustained systolic blood pressure performance and short-term systolic blood pressure fluctuation of the target patient, the abnormal blood pressure trend performance of the target patient is determined.

[0020] In conjunction with the first aspect above, in one possible implementation, determining the recent persistent concurrent hypertension manifestation degree of the target patient based on the set of historical high-pressure periods corresponding to all historical cycle periods includes:

[0021] Each historical high-pressure period in the set of historical high-pressure periods corresponding to each historical cycle period is identified as the target stage, thus obtaining the target stage set corresponding to each historical cycle period.

[0022] The intersection of the target phase sets corresponding to all historical periodic time periods is determined as the recent sustained high-pressure time interval corresponding to the target patient;

[0023] The union of the target stages corresponding to all historical periodic time intervals is determined as the recent hypertension occurrence time interval for the target patient.

[0024] Based on the duration of the recent sustained high blood pressure time interval and the duration of the recent high blood pressure occurrence time interval, the degree of recent sustained concurrent high blood pressure manifestation of the target patient is determined.

[0025] In conjunction with the first aspect above, in one possible implementation, determining the short-term systolic volatility of the target patient based on the set of historical high-pressure periods corresponding to all historical cycle periods includes:

[0026] All historical high-pressure periods in the set of historical high-pressure periods corresponding to all historical cycle periods constitute a historical high-pressure period sequence;

[0027] The mean of all systolic blood pressures collected from the target patients during each historical high-pressure period is determined as the representative systolic blood pressure value for each historical high-pressure period.

[0028] Based on the ratio between the representative systolic blood pressure values ​​corresponding to adjacent historical high-pressure periods in the historical high-pressure period sequence, and the duration between adjacent historical high-pressure periods, the short-term systolic blood pressure fluctuation corresponding to the target patient is determined.

[0029] In conjunction with the first aspect above, in one possible implementation, the possible high-pressure time interval corresponding to the predicted target patient includes:

[0030] The recent sustained high-pressure time interval corresponding to the target patient is determined as the possible high-pressure time interval corresponding to the target patient.

[0031] In conjunction with the first aspect above, in one possible implementation, acquiring the current CTA image corresponding to the target patient based on the possible high-pressure time interval within the current cycle period includes:

[0032] If the possible high-pressure time interval is not empty, then the CTA image of the target patient acquired at a random moment in the possible high-pressure time interval within the current cycle period is determined as the current CTA image corresponding to the target patient.

[0033] If the high-pressure time interval is empty, then the CTA image of the target patient acquired at a random moment within the current cycle period will be determined as the current CTA image corresponding to the target patient.

[0034] In conjunction with the first aspect above, in one possible implementation, determining the early warning index for premature rupture of the target patient based on the aortic diameter in the current CTA image includes:

[0035] Obtain the aortic diameter of all historical patients belonging to the preset aortic dissection category of the target patient;

[0036] The mean aortic diameter of all historical patients under the preset aortic dissection category to which the target patient belongs is determined as the reference diameter index for the target patient.

[0037] Based on the aortic diameter in the current CTA image, the reference diameter index corresponding to the target patient, and the preset abnormal aortic diameter threshold, the early warning index for the precursor rupture of the target patient is determined.

[0038] In conjunction with the first aspect above, in one possible implementation, determining the spread of new ruptures in the target patient based on the difference between the ruptures in the current CTA image and those in the previously acquired CTA image includes:

[0039] The maximum diameter of each breach in the current CTA image is determined as the breach diameter index, thus obtaining the breach diameter index set corresponding to the current CTA image;

[0040] The sum of all the rupture diameter indices in the rupture diameter index set corresponding to the current CTA image is determined as the representative rupture diameter corresponding to the current CTA image. Similarly, the representative rupture diameter corresponding to the CTA image acquired in the previous CTA image is determined.

[0041] The spread of new ruptures in the target patient is determined by the difference between the representative diameter of the rupture in the current CTA image and the representative diameter of the rupture in the previous CTA image.

[0042] In conjunction with the first aspect above, in one possible implementation, determining the correlation between the blood pressure image and the current CTA image based on the early warning indicators of premature rupture, the spread of new ruptures, and the abnormal trend of blood pressure includes:

[0043] The false lumen dilation index for the target patient is determined based on the difference between the false lumen cross-sectional area corresponding to the current CTA image and the false lumen cross-sectional area corresponding to the previously acquired CTA image.

[0044] If CTA images of the target patient were acquired before the current CTA image, the static risk of acute rupture for the target patient is determined based on the new rupture spread and false lumen expansion index corresponding to the target patient.

[0045] If no CTA images of the target patient have been acquired before the current CTA image, the static risk of acute rupture for the target patient is set to a constant of 0.

[0046] Based on the acute rupture static risk level and the early warning index of the precursor rupture corresponding to the target patient, the aortic rupture warning level corresponding to the target patient is determined;

[0047] Based on the aortic rupture warning level and blood pressure abnormality trend performance of the target patient, the blood pressure image correlation degree corresponding to the current CTA image is determined.

[0048] In conjunction with the first aspect above, in one possible implementation, determining the overall correlation coefficient corresponding to the target patient based on the correlation of all blood pressure images includes:

[0049] If the correlation between the CTA image and the blood pressure image of the target patient is greater than or equal to the preset correlation threshold, then the CTA image is determined as the reference CTA image.

[0050] The total number of all CTA images acquired from the target patient is determined as the number of images representing the target patient.

[0051] The ratio of the number of reference CTA images to the number of images represented is determined as the high correlation ratio coefficient corresponding to the target patient;

[0052] Based on the correlation degree of the blood pressure image corresponding to the current CTA image and the high correlation ratio coefficient corresponding to the target patient, the overall correlation coefficient corresponding to the target patient is determined.

[0053] Secondly, the present invention provides an early warning system for the risk of aortic dissection rupture based on CTA images, the system comprising:

[0054] The determination and prediction module is used to determine the degree of abnormal blood pressure trend of the target patient based on the distribution of systolic blood pressure of the target patient within a preset historical time period, and to predict the possible high blood pressure time interval of the target patient.

[0055] The acquisition and determination module is used to acquire all CTA images of the target patient in the past, acquire the current CTA image of the target patient according to the possible high-pressure time interval in the current cycle period, and determine the early warning index of the target patient based on the aortic diameter in the current CTA image.

[0056] The new rupture spread determination module is used to determine the new rupture spread of the target patient based on the difference between the rupture in the current CTA image and the rupture in the previous CTA image.

[0057] The blood pressure image correlation determination module is used to determine the blood pressure image correlation of the current CTA image based on the early warning indicators of prodromal rupture, the spread of new rupture, and the trend of abnormal blood pressure. Similarly, it determines the blood pressure image correlation of the target patient in historically acquired CTA images.

[0058] The identification and early warning module is used to determine the overall correlation coefficient corresponding to the target patient based on the correlation of all blood pressure images, and to realize the early warning of aortic dissection rupture risk based on the target patient's systolic blood pressure, overall correlation coefficient and current CTA images in the current cycle period, through a pre-trained aortic dissection rupture risk early warning network.

[0059] Thirdly, a server is provided, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, causing the device to perform the methods of the first aspect or any possible implementation thereof.

[0060] Fourthly, a computer program product is provided, comprising: computer program code, which, when run on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0061] Fifthly, a computer-readable storage medium is provided that stores computer program code, which, when executed on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0062] The present invention has the following beneficial effects:

[0063] This invention presents a method for early warning of aortic dissection rupture based on CTA images. When using a network, this invention combines systolic blood pressure within the current time period, the overall correlation coefficient, and the current CTA image. This solves the technical problem of poor accuracy in early warning of aortic dissection rupture risk caused by the limited richness of data input to the neural network, thereby enriching the richness of the network input data and improving the accuracy of early warning of aortic dissection rupture risk. Specifically, when using a neural network for early warning of aortic dissection rupture risk, this invention considers not only CTA images but also the patient's systolic blood pressure distribution and quantifies the overall correlation coefficient characterizing the overall correlation between CTA images and the patient's systolic blood pressure. This enriches the input data of the aortic dissection rupture risk warning network to a certain extent, thereby improving the risk warning effect of the aortic dissection rupture risk warning network and ultimately improving the accuracy of early warning of aortic dissection rupture risk. Attached Figure Description

[0064] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0065] Figure 1 This is a flowchart of a method for early warning of aortic dissection rupture based on CTA images according to the present invention;

[0066] Figure 2 This is a schematic diagram of the composition structure of an aortic dissection rupture risk warning system based on CTA images according to the present invention;

[0067] Figure 3 This is a schematic diagram of the structure of a computer device according to the present invention. Detailed Implementation

[0068] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the technical solution proposed according to the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0069] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0070] Aortic dissection rupture is an extremely serious medical emergency, often posing a severe threat to life. When rupture occurs, patients typically experience rapid and severe circulatory problems, such as acute hemorrhagic shock and multiple organ failure. This condition often develops rapidly and fatally, with a very high mortality rate without timely intervention. However, in recent years, advancements in medical imaging technology, especially computed tomography (CTA), have provided high-resolution vascular images, facilitating the accurate identification of the location, size, and morphology of aortic dissections and their potential risk of rupture. This has significantly improved the methods for diagnosing and assessing the risk of aortic dissection.

[0071] When patients have risk factors such as hypertension, the aortic wall is subjected to continuous pressure, leading to intimal damage and further increasing the risk of aortic dissection rupture. However, current diagnosis of aortic dissection primarily relies on CTA imaging analysis. While CTA provides detailed static imaging information, it has limitations: CTA images only reflect the vascular structure at a fixed point in time and cannot perform dynamic blood flow analysis. This restricts a comprehensive understanding of the progression of aortic dissection. Particularly in the assessment of acute aortic dissection, changes in dynamic blood flow characteristics are crucial for judging the evolution of the condition, the risk of rupture, and determining the timing of surgery.

[0072] Thus, the lack of dynamic monitoring capabilities makes it difficult for traditional CTA imaging analysis methods to accurately reflect changes in the patient's condition. As a result, diagnostic results can only be based on static data, which ignores the complexity and rapid changes in hemodynamic characteristics and may affect the accuracy of clinical decision-making.

[0073] In addition to considering CTA images, this invention also considers changes in systolic blood pressure that reflect dynamic blood flow characteristics and quantifies the overall correlation coefficient that characterizes the overall correlation between CTA images and the patient's systolic blood pressure, thereby enriching the richness of network input data and improving the accuracy of aortic dissection rupture risk warning.

[0074] refer to Figure 1 The flowchart illustrates some embodiments of a method for early warning of aortic dissection rupture based on CTA images according to the present invention. This method for early warning of aortic dissection rupture based on CTA images includes the following steps:

[0075] Step S1: Based on the distribution of systolic blood pressure of the target patient within a preset historical time period, determine the degree of abnormal blood pressure trend of the target patient and predict the possible high blood pressure time interval of the target patient.

[0076] The target patient can be an aortic dissection patient undergoing a risk warning for aortic dissection rupture. Aortic dissection (AD), also known as aortic aneurysm dissection, is a serious cardiovascular emergency. When a tear occurs in the intima of the artery wall, blood enters the arterial wall through the tear, forming a hematoma, and further dissecting the intima and media of the aorta, resulting in aortic dissection. The preset historical time period can be a pre-set historical period, the corresponding duration of which can be one month. The next moment after the end of the preset historical time period can be the start moment of the current cycle period. The current cycle period can be the cycle period during which the target patient plans to undergo CTA (Computed Tomography Angiography), the corresponding duration of which can be one day. In this embodiment of the invention, one day is used as a cycle period. For example, the preset historical time period can be the month of May 2025, and the current cycle period can be June 1, 2025. Systolic blood pressure reflects the systolic function of the left ventricle and the elasticity of the aorta and great vessels. The possible high blood pressure time interval can be the time interval during which the target patient frequently experiences abnormal blood pressure.

[0077] It should be noted that aortic dissection is often caused by structural abnormalities and hemodynamic abnormalities in the aortic media itself, and patients with hypertension are often at high risk of developing aortic dissection. When monitoring and analyzing patients' hypertension, systolic blood pressure is the core detection target because the shear stress on the aortic wall during cardiac contraction is positively correlated with systolic blood pressure, and fluctuations in systolic blood pressure can cause fatigue damage to the aortic wall, which is a key factor in inducing acute rupture.

[0078] The risk of aortic dissection rupture is closely related to blood pressure levels. Long-term hypertension can cause continuous pressure on the aortic wall, damaging the intima and increasing the risk of rupture. Therefore, dynamic hemodynamic analysis based on blood flow can be considered for patients with aortic dissection, so that CTA images can be dynamically acquired and further compared and analyzed according to their manifestations.

[0079] The impact of hypertension on aortic dissection rupture is reflected not only in its persistence but also in its fluctuation. This is because a sustained high blood pressure load often leads to the rupture of collagen fibers and the degeneration of elastic fibers in the aortic wall, and accelerates atherosclerosis, making the intima more prone to tearing. Blood pressure fluctuation, especially short-term fluctuations, is closely related to the risk of dissection rupture. When blood pressure fluctuates drastically, the shear force on the aortic wall often changes more significantly, which may induce secondary intimal tearing or false lumen expansion.

[0080] As an example, this step may include the following steps:

[0081] The first step is to determine the systolic blood pressure of the target patient within a preset historical time period as the systolic blood pressure threshold if the systolic blood pressure is greater than or equal to the preset systolic blood pressure threshold.

[0082] The preset systolic blood pressure threshold can be a pre-set threshold, such as 140 mmHg.

[0083] It should be noted that a patient's systolic blood pressure can be obtained through wearable devices. These wearable devices can be smart bracelets, etc.

[0084] The second step is to periodically divide the preset historical time period to obtain historical periodic time periods, and then to form historical high-pressure time periods by combining the continuous high-pressure moments within the historical periodic time periods, thus obtaining the set of historical high-pressure time periods corresponding to each historical periodic time period.

[0085] The historical period can be a period whose duration is equal to that of the current period. For example, if the preset historical period is May 2025, then every day in May 2025 can be recorded as a historical period.

[0086] For example, if there are 10 systolic blood pressure measurement times within a certain historical period, and these 10 systolic blood pressure measurement times are the first time, the second time, the third time, the fourth time, the fifth time, the sixth time, the seventh time, the eighth time, the ninth time, and the tenth time, and the third time, the fourth time, the fifth time, the seventh time, the eighth time, the ninth time, and the tenth time are high pressure times, then the set of historical high pressure times corresponding to the historical period can have two historical high pressure times, and these two historical high pressure times are {the third time, the fourth time, the fifth time} and {the seventh time, the eighth time, the ninth time, the tenth time}, respectively.

[0087] The third step is to determine the total number of all high-pressure moments within the historical cycle period as the representative value of the high-pressure duration corresponding to the historical cycle period, and to determine the average of the representative values ​​of the high-pressure duration corresponding to all historical cycle periods as the stage high-pressure duration value corresponding to the target patient.

[0088] The fourth step, based on the set of historical high-pressure periods corresponding to all historical cycle periods, determines the recent persistent concurrent high-pressure manifestation level for the target patients, which may include the following sub-steps:

[0089] The first sub-step involves identifying the stage corresponding to each historical high-pressure period in the set of historical high-pressure periods corresponding to each historical cycle period as the target stage, thus obtaining the set of target stages corresponding to each historical cycle period.

[0090] The historical high-pressure period can be a stage within a periodic period, representing a time segment within that period. For example, if a historical high-pressure period begins at 11:12:02 on May 2, 2025, and ends at 11:12:16 on May 2, 2025, then the start time of the stage corresponding to that historical high-pressure period can be 11:12:02, and the start time of the stage corresponding to that historical high-pressure period can also be 11:12:16. The number of target stages in the set of target stages corresponding to a historical periodic period can be equal to the number of historical high-pressure periods in the set of historical high-pressure periods corresponding to that historical periodic period.

[0091] The second sub-step involves determining the intersection of the target phase sets corresponding to all historical cycle periods as the recent sustained high-pressure time interval for the aforementioned target patient.

[0092] It should be noted that the recent sustained high-pressure time interval can represent the time during which the target patient frequently experiences high pressure in each historical cycle within a preset historical time period.

[0093] The third sub-step involves determining the union of the target stage sets corresponding to all historical cycle periods as the recent hypertension occurrence time interval for the aforementioned target patients.

[0094] It should be noted that the recent high-tension occurrence time interval can represent the high-tension time that the target patient has experienced in different historical cycle periods within the preset historical time period.

[0095] The fourth sub-step involves determining the degree of recent continuous hypertension in the target patient based on the duration of the aforementioned recent sustained hypertension time interval and the duration of the aforementioned recent hypertension occurrence time interval.

[0096] The fifth step is to determine the sustained hypertension performance of the target patients based on the duration of hypertension in the relevant phase and the recent sustained hypertension performance in the same period.

[0097] For example, the formula for determining the degree of persistent hypertension in a target patient can be:

[0098] ;

[0099] Here, xA represents the degree of persistent high blood pressure in the target patient. It is a normalization function. xB is the duration of hypertension corresponding to the target patient, which is the average of the representative values ​​of hypertension duration for all historical periodic periods within the preset historical time period. It refers to the duration of the recent sustained high-pressure time interval corresponding to the target patient. It refers to the duration of the recent hypertension occurrence time interval for the target patient. It is a pre-set factor greater than 0, mainly used to prevent the denominator from being 0, and it can be 0.0001. It refers to the recent, persistent, and concurrent hypertension symptoms of the target patient.

[0100] It should be noted that a larger xB value often indicates a longer duration of high pressure within different historical periods, and a greater likelihood of persistent high pressure during those periods. A larger xA value often indicates that hypertension is more likely to occur frequently at the same time within different historical periods, and that hypertension is more likely to be persistent across different historical periods. Therefore, a larger xA value often indicates that the target patient is more likely to experience relatively persistent hypertension within the pre-defined historical time period.

[0101] Step 6, based on the set of historical high-pressure periods corresponding to all historical cycle periods, determining the short-term high-pressure fluctuations for the target patients may include the following sub-steps:

[0102] The first sub-step involves constructing a historical high-pressure period sequence from all historical high-pressure periods in the set of historical high-pressure periods corresponding to all historical periodic periods.

[0103] Among them, the historical high-pressure period sequence can be a time series.

[0104] The second sub-step involves determining the average systolic blood pressure of all target patients collected during each historical high-pressure period as the representative systolic blood pressure value for that period.

[0105] The third sub-step involves determining the short-term systolic blood pressure fluctuation for the target patient based on the ratio between the representative values ​​of systolic blood pressure corresponding to adjacent historical high-pressure periods in the aforementioned historical high-pressure period sequence, as well as the duration between adjacent historical high-pressure periods.

[0106] For example, the formula for determining the short-term systolic blood pressure fluctuations corresponding to a target patient can be:

[0107] ;

[0108] Where D represents the short-term systolic blood pressure fluctuation corresponding to the target patient. N represents the number of historical systolic blood pressure periods in the historical systolic blood pressure period sequence. i represents the sequence number of the historical systolic blood pressure period in the historical systolic blood pressure period sequence. It is the representative value of systolic blood pressure corresponding to the i-th historical high-pressure period in the historical high-pressure period sequence. It is the representative value of systolic blood pressure corresponding to the (i+1)th historical high-pressure period in the historical high-pressure period sequence. It is the duration between the i-th historical high-pressure period and the (i+1)-th historical high-pressure period in the historical high-pressure period sequence. It can be equal to the duration between the latest moment of the i-th historical high-pressure period and the earliest moment of the (i+1)-th historical high-pressure period. It is a pre-set factor greater than 0, mainly used to prevent the denominator from being 0, and it can be 0.0001.

[0109] It should be noted that, for adjacent historical high-pressure periods, if the latter is at a higher systolic pressure level than the former, it often indicates a potentially more severe trend of fluctuating growth. A higher value generally indicates that the systolic blood pressure in the (i+1)th historical high-pressure period is more likely to be higher than that in the ith historical high-pressure period, suggesting a more likely upward trend in systolic blood pressure. For patients with frequent short-term fluctuations, there is often a clear discontinuity within their historical high-pressure periods; that is, the smaller the time interval between adjacent historical high-pressure periods, the more frequent the short-term fluctuations between them. The smaller the value of D, the more frequent the short-term fluctuations between the i-th and (i+1)-th historical high-pressure periods tend to be. Therefore, the larger the value of D, the more likely the target patient's systolic blood pressure exhibits frequent short-term fluctuations.

[0110] Step 7: Based on the sustained systolic blood pressure performance and short-term systolic blood pressure fluctuation of the target patients, determine the abnormal blood pressure trend performance of the target patients.

[0111] For example, the formula for determining the blood pressure abnormality trend performance of a target patient can be:

[0112] ;

[0113] Where F represents the degree of abnormal blood pressure trend in the target patient. It is a normalized function. xA represents the sustained high blood pressure performance of the target patient. D represents the short-term high blood pressure fluctuation of the target patient.

[0114] It should be noted that a larger xA value often indicates that the target patient is more likely to have relatively persistent high blood pressure within the preset historical time period. A larger D value often indicates that the target patient's systolic blood pressure is more consistent with short-term, frequent fluctuations. Therefore, a larger F value often indicates that the target patient's systolic blood pressure distribution is relatively more abnormal.

[0115] The eighth step is to determine the recent sustained high-pressure time interval corresponding to the above-mentioned target patients as the possible high-pressure time interval corresponding to the above-mentioned target patients.

[0116] Step S2: Acquire all historical CTA images of the target patient. Based on the possible high-pressure time intervals within the current cycle period, acquire the current CTA images corresponding to the target patient. Based on the aortic diameter in the current CTA images, determine the early warning indicators for the target patient's prodromal rupture.

[0117] It should be noted that the CTA images in the embodiments of the present invention can be CTA images of aortic dissection.

[0118] As an example, this step may include the following steps:

[0119] The first step is to determine the current CTA image of the target patient as the one acquired at a random moment within the potential high-pressure time interval during the current cycle period if the possible high-pressure time interval is not empty.

[0120] The second step is to determine the current CTA image of the target patient as the target patient if the high-pressure time interval is empty.

[0121] The third step is to obtain the aortic diameter of all historical patients under the preset aortic dissection category belonging to the target patient.

[0122] It's important to note that aortic dissection is primarily classified into two systems based on the location and extent of involvement: the Stanford classification and the DeBakey classification. The Stanford classification is further divided into types A and B, while the DeBakey classification is subdivided into types I, II, and III. Different medical strategies apply to different types of aortic dissection, especially for Stanford types A, DeBakey types I, and II, which involve the ascending aorta. While the specific medical strategies differ, the common treatment principle is high surgical urgency. Stanford types B and DeBakey types III, which do not involve the ascending aorta, have relatively lower urgency in their treatment strategies. Therefore, aortic dissection can be categorized into two types based on surgical urgency; that is, there can be two pre-defined categories: an urgent category and a less urgent category. The urgent category includes Stanford types A, DeBakey types I, and II. The less urgent category includes Stanford types B and DeBakey types III. Patients with a history of aortic dissection are eligible. Patients with a predefined aortic dissection category can be those with aortic dissection disease belonging to that category. The aorta often includes different parts such as the false lumen, intima-lamella, true lumen, and tear. The aortic diameter for a historical patient can be obtained as follows: acquire CTA images of the historical patient; segment the aortic intima-lamella, true and false lumen, and tear using image segmentation techniques to obtain segmented images of different parts of the aorta; based on the segmented images of each part of the aorta, a NURBS surface model can be constructed, and a meshed surface can be generated using the Marching Cubes algorithm to obtain the morphological parameters corresponding to each part of the aorta, namely the aortic diameter (maximum diameter) and the cross-sectional area of ​​the false lumen.

[0123] The fourth step is to determine the average aortic diameter of all historical patients under the preset aortic dissection category to which the target patient belongs as the reference diameter index for the target patient.

[0124] The fifth step is to determine the early warning indicators for the aforementioned target patient's aortic rupture based on the current aortic diameter in the CTA images, the reference diameter index corresponding to the target patient, and the preset abnormal aortic diameter threshold.

[0125] The preset abnormal aortic diameter threshold can be a pre-set threshold, which can be 5.5 cm.

[0126] It should be noted that in medicine, an aortic diameter of 5.5 cm is usually considered an abnormal standard, so this value is recorded here as the preset abnormal aortic diameter threshold.

[0127] For example, the formula for determining the early warning index for premature rupture in a target patient can be:

[0128] ;

[0129] G is the early warning indicator for premature rupture corresponding to the target patient. It is a normalization function. It is the diameter of the aorta in the current CTA image. This is a preset threshold for abnormal aortic diameter. H is the reference diameter for the target patient.

[0130] It should be noted that when A larger G value generally indicates a larger aortic diameter in the target patient at the time of examination, and thus a relatively higher risk of impending rupture. A larger H value also generally indicates that the target patient is more likely to be classified as having an emergency condition. Therefore, a larger G value generally indicates a relatively higher risk of impending rupture.

[0131] Step S3: Determine the extent of new rupture spread in the target patient based on the differences between the current CTA image and the rupture in the previous CTA image.

[0132] It should be noted that the CTA image acquired before the current CTA image can be the CTA image of the target patient acquired in the previous acquisition of the current CTA image.

[0133] As an example, this step may include the following steps:

[0134] The first step is to determine the maximum diameter of each breach in the current CTA image as the breach diameter index, thus obtaining the breach diameter index set corresponding to the current CTA image.

[0135] The second step is to determine the cumulative value of all the breach diameter indicators in the breach diameter indicator set corresponding to the current CTA image as the representative breach diameter corresponding to the current CTA image. Similarly, the representative breach diameter corresponding to the previous CTA image is determined.

[0136] It should be noted that the method for obtaining the representative diameter of the rupture corresponding to the previous CTA image can be the same as the method for obtaining the representative diameter of the rupture corresponding to the current CTA image, and will not be repeated here.

[0137] The third step is to determine the spread of new ruptures in the target patients based on the difference between the representative diameter of the rupture in the current CTA image and the representative diameter of the rupture in the previous CTA image.

[0138] For example, the formula for determining the spread of new ulcers in a target patient can be:

[0139] ;

[0140] in, It refers to the spread of new ruptures corresponding to the target patient. It is a normalization function. It is the representative diameter of the rupture corresponding to the current CTA image. It is the representative diameter of the rupture corresponding to the previous CTA image acquired for the current CTA image.

[0141] It should be noted that when The larger the value, the more likely the rupture area of ​​the target patient has expanded, meaning it may have spread beyond the original rupture or a new rupture may have appeared at a different location from the original rupture.

[0142] Step S4: Based on the early warning indicators of premature rupture, the spread of new ruptures, and the trend of abnormal blood pressure, determine the correlation between the current CTA images and the blood pressure images. Similarly, determine the correlation between the target patient and the blood pressure images in the historically acquired CTA images.

[0143] As an example, this step may include the following steps:

[0144] The first step is to determine the false lumen dilation index for the target patient based on the difference between the false lumen cross-sectional area corresponding to the current CTA image and the false lumen cross-sectional area corresponding to the previous CTA image.

[0145] For example, the formula for determining the false lumen dilation index for a target patient can be:

[0146] ;

[0147] Where h is the pseudocavity dilation index corresponding to the target patient. It is a normalization function. It is the cross-sectional area of ​​the false cavity corresponding to the current CTA image. It is the cross-sectional area of ​​the false cavity corresponding to the previous CTA image acquired for the current CTA image.

[0148] It should be noted that the larger the h value, the more likely the false cavity area of ​​the target patient has been expanded.

[0149] The second step, if CTA images of the target patient were acquired before the current CTA images, then the formula for determining the static risk level of acute rupture for the target patient, based on the new rupture spread and false lumen expansion indices, can be:

[0150] ;

[0151] Where k is the static risk level of acute rupture for the target patient. It is a normalization function. is the new wound spread rate for the target patient. h is the false lumen expansion rate for the target patient.

[0152] It's important to note that when the false lumen of aortic dissection expands, it's usually accompanied by changes in blood flow pressure. This can lead to uneven stress on the aortic wall, potentially causing the existing tear to enlarge further or a new tear to form at the dissection site. Simultaneously, the expansion of the false lumen may slow blood flow, promoting thrombus formation within the false lumen. Thrombosis can increase pressure in the false lumen, leading to continuous stress on the wall and exacerbating the tear and worsening the condition. Therefore, increased false lumen expansion and tear propagation often reflect the instability of the patient's condition. Conversely, if these changes are slow or limited, the condition may be relatively stable, indicating milder clinical manifestations and a lower probability of acute complications. Thus, a larger k value often indicates a greater likelihood of expansion in the tear and false lumen areas of the target patient, suggesting a relatively higher risk of acute rupture.

[0153] Third, if no CTA images of the target patient have been acquired before the current CTA images, the static risk of acute rupture corresponding to the target patient is set to a constant of 0.

[0154] The fourth step is to determine the aortic rupture warning level for the target patients based on the acute rupture static risk level and the early warning index of the precursor rupture.

[0155] For example, the formula for determining the aortic rupture warning level for a target patient can be:

[0156] ;

[0157] Where Y represents the aortic rupture warning level for the target patient. It is a normalized function. G is the early warning index for premature rupture corresponding to the target patient. k is the static risk level of acute rupture corresponding to the target patient.

[0158] It should be noted that a larger G value generally indicates a greater risk of premature rupture. A larger k value generally indicates that the rupture area and false lumen area of ​​the target patient are more likely to have expanded, which generally indicates a greater risk of acute rupture. Therefore, a larger Y value generally indicates a greater risk of aortic rupture in the target patient.

[0159] The fifth step is to determine the blood pressure image correlation degree corresponding to the current CTA image based on the aortic rupture warning level and blood pressure abnormality trend performance of the target patients mentioned above.

[0160] For example, the formula for determining the correlation between the current CTA image and the blood pressure image can be:

[0161] ;

[0162] Where P is the correlation degree between the current CTA image and the blood pressure image. It is an exponential function with the natural constant as its base. It is an absolute value function. F represents the degree of abnormal blood pressure trend in the target patient. Y represents the degree of aortic rupture warning in the target patient.

[0163] It should be noted that a larger F value generally indicates a more abnormal systolic blood pressure distribution in the target patient. A larger Y value generally indicates a higher risk of aortic rupture in the target patient. Therefore, a larger P value generally indicates a closer similarity between the target patient's abnormal systolic blood pressure and the aortic rupture risk shown in the current CTA imaging. This often suggests better clarity in the current CTA imaging and, to some extent, greater analytical value in the target patient's systolic blood pressure distribution.

[0164] The sixth step, similarly, is to determine the correlation between the target patient's blood pressure images and the historically acquired CTA images.

[0165] It should be noted that the method for obtaining the correlation between blood pressure images and historically acquired CTA images of the target patient can be the same as the method for obtaining the correlation between blood pressure images and current CTA images, and will not be elaborated here.

[0166] Step S5: Based on the correlation of all blood pressure images, determine the overall correlation coefficient corresponding to the target patient, and based on the target patient's systolic blood pressure, overall correlation coefficient and current CTA images in the current cycle period, realize the early warning of aortic dissection rupture risk through the pre-trained aortic dissection rupture risk network.

[0167] As an example, this step may include the following steps:

[0168] The first step is to determine the CTA image as the reference CTA image if the correlation between the CTA image of the target patient and the blood pressure image is greater than or equal to the preset correlation threshold.

[0169] The preset association threshold can be a pre-set threshold, which can be 0.6.

[0170] The second step is to determine the total number of CTA images collected from the target patient as the number of images representing the target patient.

[0171] The third step is to determine the high correlation ratio coefficient corresponding to the target patients by the ratio of the number of reference CTA images to the number of images represented above.

[0172] The fourth step is to determine the overall correlation coefficient for the target patients based on the correlation coefficient of the blood pressure images corresponding to the current CTA images and the high correlation ratio coefficient corresponding to the target patients.

[0173] For example, the formula for determining the overall correlation coefficient corresponding to the target patient can be:

[0174] ;

[0175] Where W is the overall correlation coefficient corresponding to the target patient. This is the normalization function. m is the number of reference CTA images. M is the number of images representing the target patient. P is the blood pressure image correlation degree corresponding to the current CTA image.

[0176] It should be noted that a larger P value often indicates a greater similarity between the target patient's abnormal systolic blood pressure and the aortic rupture risk shown in the current CTA imaging. This can suggest, to some extent, better clarity of the current CTA images and, to a certain extent, greater analytical value in analyzing the target patient's systolic blood pressure distribution. A larger W value generally indicates a higher correlation between the acquired CTA images and systolic blood pressure distribution in the target patient. Therefore, a larger W value often suggests a greater similarity between the target patient's abnormal systolic blood pressure and the aortic rupture risk shown on the CTA images. This can also indicate better clarity in the acquired CTA images and, to some extent, greater analytical value in the target patient's systolic blood pressure distribution.

[0177] The fifth step involves using the target patient's systolic blood pressure during the current cycle period, along with the overall correlation coefficient and current CTA images, to perform a pre-trained aortic dissection rupture risk warning through a pre-trained aortic dissection rupture risk warning network.

[0178] Among them, the aortic dissection rupture risk warning network can be a CNN (Convolutional Neural Network) used for aortic dissection rupture risk warning.

[0179] For example, the training process of an aortic dissection rupture risk warning network may include the following sub-steps:

[0180] The first sub-step is to construct a CNN as a pre-training network for the risk of aortic dissection rupture.

[0181] The second sub-step involves obtaining the preset time period, overall correlation coefficient, and current CTA images for different aortic dissection patients.

[0182] The preset time period for aortic dissection patients can be the time period during which the patient is scheduled to undergo a CTA examination. For example, the current time period mentioned earlier is the preset time period for the target patient.

[0183] It should be noted that the overall correlation coefficient and the method for acquiring the current CTA images for patients with aortic dissection can be the same as those for the target patients, and will not be elaborated further here.

[0184] The third sub-step involves using the systolic blood pressure of different aortic dissection patients during their corresponding preset time periods, along with the overall correlation coefficient and current CTA images of different aortic dissection patients, as training samples for the aortic dissection rupture risk warning network. The aortic dissection rupture risk level corresponding to different aortic dissection patients is used as the training label for the aortic dissection rupture risk warning network. The aortic dissection rupture risk warning network is then trained using the training samples and training labels to obtain the completed aortic dissection rupture risk warning network.

[0185] It should be noted that the aortic dissection rupture risk level for a patient with aortic dissection can be the aortic dissection rupture risk level assessed by a doctor. A higher value generally indicates a greater likelihood of aortic dissection rupture. Furthermore, the aortic dissection rupture risk warning network, once trained, can assist doctors in assessing aortic dissection rupture to some extent.

[0186] refer to Figure 2 Based on the same inventive concept as the above-described method embodiments, this invention provides a CTA image-based aortic dissection rupture risk warning system. This system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the steps of a CTA image-based aortic dissection rupture risk warning method, specifically including:

[0187] The determination and prediction module 201 is used to determine the blood pressure abnormality trend of the target patient based on the distribution of systolic blood pressure of the target patient within a preset historical time period, and to predict the possible high blood pressure time interval of the target patient.

[0188] The acquisition and determination module 202 is used to acquire all CTA images of the target patient in the past, acquire the current CTA image of the target patient according to the possible high pressure time interval in the current cycle period, and determine the early warning index of the target patient based on the aortic diameter in the current CTA image.

[0189] The new rupture spread determination module 203 is used to determine the new rupture spread of the target patient based on the difference between the rupture in the current CTA image and the rupture in the previous CTA image.

[0190] The blood pressure image correlation determination module 204 is used to determine the blood pressure image correlation of the current CTA image based on the early warning indicators of prodromal rupture, the spread of new rupture and the trend of abnormal blood pressure. Similarly, it determines the blood pressure image correlation of the target patient in the historically acquired CTA images.

[0191] The determination and early warning module 205 is used to determine the overall correlation coefficient corresponding to the target patient based on the correlation of all blood pressure images, and to realize the early warning of aortic dissection rupture risk based on the target patient's systolic blood pressure, overall correlation coefficient and current CTA images in the current cycle period, through a pre-trained aortic dissection rupture risk early warning network.

[0192] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. For example, as shown... Figure 3 As shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute any of the aforementioned methods for early warning of aortic dissection rupture based on CTA images.

[0193] Based on the same inventive concept as the above-described method embodiments, the present invention provides a server, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes any of the above-described aortic dissection rupture risk warning methods based on CTA images.

[0194] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer program product comprising: computer program code, which, when run on a computer, causes the computer to execute any of the above-described methods for early warning of aortic dissection rupture based on CTA images.

[0195] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer-readable storage medium storing computer program code, which, when executed on a computer, causes the computer to perform any of the above-described methods for early warning of aortic dissection rupture based on CTA images.

[0196] In summary, when using neural networks for aortic dissection rupture risk warning, this invention considers not only CTA images but also the patient's systolic blood pressure distribution. Furthermore, it quantifies the overall correlation coefficient, which characterizes the overall correlation between CTA images and the patient's systolic blood pressure. This enriches the input data of the aortic dissection rupture risk warning network to a certain extent, thereby improving the risk warning effect of the aortic dissection rupture risk warning network and ultimately improving the accuracy of aortic dissection rupture risk warning.

[0197] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for early warning of aortic dissection rupture risk based on CTA images, characterized in that, Includes the following steps: Based on the distribution of systolic blood pressure of the target patient within a preset historical time period, determine the degree of abnormal blood pressure trend of the target patient and predict the possible high blood pressure time interval of the target patient. Acquire all historical CTA images of the target patient, acquire the current CTA images corresponding to the target patient based on the possible high-pressure time intervals within the current cycle period, and determine the early warning indicators of the target patient's aortic rupture based on the aortic diameter in the current CTA images. Based on the differences between the current CTA images and the ruptures in the previous CTA images, the extent of new rupture spread in the target patient is determined. Based on the early warning indicators of premature rupture, the spread of new ruptures, and the trend of abnormal blood pressure, the correlation between blood pressure images and current CTA images is determined. Similarly, the correlation between blood pressure images and historically acquired CTA images of the target patient is determined. Based on the correlation of all blood pressure images, the overall correlation coefficient corresponding to the target patient is determined. Based on the target patient's systolic blood pressure, overall correlation coefficient and current CTA images in the current cycle period, the aortic dissection rupture risk warning network, the aortic dissection rupture risk warning is realized. The step of determining the blood pressure abnormality trend performance of the target patient based on the distribution of systolic blood pressure within a preset historical time period includes: If the systolic blood pressure of the target patient collected within the preset historical time period is greater than or equal to the preset systolic blood pressure threshold, then the collection time corresponding to that systolic blood pressure is determined as the high blood pressure time. The preset historical time period is periodically divided to obtain historical periodic time periods. The consecutive high-pressure moments within the historical periodic time period are then used to form historical high-pressure time periods, resulting in a set of historical high-pressure time periods corresponding to each historical periodic time period. The total number of all high-pressure moments within the historical cycle period is determined as the representative value of the high-pressure duration corresponding to the historical cycle period, and the average of the representative values ​​of the high-pressure duration corresponding to all historical cycle periods is determined as the stage high-pressure duration value corresponding to the target patient. Based on the set of historical high-pressure periods corresponding to all historical cycle periods, the recent continuous concurrent high-pressure performance of the target patient is determined. The sustained hypertension performance of the target patient is determined based on the duration of hypertension at the corresponding stage and the recent sustained hypertension performance during the same period. Based on the set of historical high-pressure periods corresponding to all historical cycle periods, the short-term high-pressure fluctuations corresponding to the target patient are determined; Based on the sustained systolic blood pressure performance and short-term systolic blood pressure fluctuation of the target patient, the abnormal blood pressure trend performance of the target patient is determined. The determination of the new rupture spread in the target patient based on the difference between the rupture in the current CTA image and the previously acquired CTA image includes: The maximum diameter of each breach in the current CTA image is determined as the breach diameter index, thus obtaining the breach diameter index set corresponding to the current CTA image; The sum of all the rupture diameter indices in the rupture diameter index set corresponding to the current CTA image is determined as the representative rupture diameter corresponding to the current CTA image. Similarly, the representative rupture diameter corresponding to the CTA image acquired in the previous CTA image is determined. The spread of new ruptures in the target patient is determined by the difference between the representative diameter of the rupture in the current CTA image and the representative diameter of the rupture in the previous CTA image.

2. The method for early warning of aortic dissection rupture based on CTA images according to claim 1, characterized in that, The step of determining the recent persistent concurrent hypertension manifestation level of the target patient based on the set of historical hypertension periods corresponding to all historical cycle periods includes: Each historical high-pressure period in the set of historical high-pressure periods corresponding to each historical cycle period is identified as the target stage, thus obtaining the target stage set corresponding to each historical cycle period. The intersection of the target phase sets corresponding to all historical periodic time periods is determined as the recent sustained high-pressure time interval corresponding to the target patient; The union of the target stages corresponding to all historical periodic time intervals is determined as the recent hypertension occurrence time interval for the target patient. Based on the duration of the recent sustained high blood pressure time interval and the duration of the recent high blood pressure occurrence time interval, the degree of recent sustained concurrent high blood pressure manifestation of the target patient is determined.

3. The method for early warning of aortic dissection rupture based on CTA images according to claim 1, characterized in that, The step of determining the short-term systolic fluctuation of the target patient based on the set of historical systolic periods corresponding to all historical cycle periods includes: All historical high-pressure periods in the set of historical high-pressure periods corresponding to all historical cycle periods constitute a historical high-pressure period sequence; The mean of all systolic blood pressures collected from the target patients during each historical high-pressure period is determined as the representative systolic blood pressure value for each historical high-pressure period. Based on the ratio between the representative systolic blood pressure values ​​corresponding to adjacent historical high-pressure periods in the historical high-pressure period sequence, and the duration between adjacent historical high-pressure periods, the short-term systolic blood pressure fluctuation corresponding to the target patient is determined.

4. The method for early warning of aortic dissection rupture based on CTA images according to claim 2, characterized in that, The possible high-pressure time intervals corresponding to the predicted target patients include: The recent sustained high-pressure time interval corresponding to the target patient is determined as the possible high-pressure time interval corresponding to the target patient.

5. The method for early warning of aortic dissection rupture based on CTA images according to claim 1, characterized in that, The step of acquiring the current CTA images corresponding to the target patient based on the possible high-pressure time intervals within the current cycle period includes: If the possible high-pressure time interval is not empty, then the CTA image of the target patient acquired at a random moment in the possible high-pressure time interval within the current cycle period is determined as the current CTA image corresponding to the target patient. If the high-pressure time interval is empty, then the CTA image of the target patient acquired at a random moment within the current cycle period will be determined as the current CTA image corresponding to the target patient.

6. The method for early warning of aortic dissection rupture based on CTA images according to claim 1, characterized in that, The method for determining the early warning indicators of premature rupture for target patients based on the aortic diameter in the current CTA images includes: Obtain the aortic diameter of all historical patients belonging to the preset aortic dissection category of the target patient; The mean aortic diameter of all historical patients under the preset aortic dissection category to which the target patient belongs is determined as the reference diameter index for the target patient. Based on the aortic diameter in the current CTA image, the reference diameter index corresponding to the target patient, and the preset abnormal aortic diameter threshold, the early warning index for the precursor rupture of the target patient is determined.

7. The method for early warning of aortic dissection rupture based on CTA images according to claim 1, characterized in that, The determination of the correlation between the current CTA image and the blood pressure image is based on the early warning indicators of premature rupture, the spread of new ruptures, and the abnormal trend of blood pressure. The false lumen dilation index for the target patient is determined based on the difference between the false lumen cross-sectional area corresponding to the current CTA image and the false lumen cross-sectional area corresponding to the previously acquired CTA image. If CTA images of the target patient were acquired before the current CTA image, the static risk of acute rupture for the target patient is determined based on the new rupture spread and false lumen expansion index corresponding to the target patient. If no CTA images of the target patient have been acquired before the current CTA image, the static risk of acute rupture for the target patient is set to a constant of 0. Based on the acute rupture static risk level and the early warning index of the precursor rupture corresponding to the target patient, the aortic rupture warning level corresponding to the target patient is determined; Based on the aortic rupture warning level and blood pressure abnormality trend performance of the target patient, the blood pressure image correlation degree corresponding to the current CTA image is determined.

8. The method for early warning of aortic dissection rupture based on CTA images according to claim 1, characterized in that, The determination of the overall correlation coefficient for the target patient based on the correlation of all blood pressure images includes: If the correlation between the CTA image and the blood pressure image of the target patient is greater than or equal to the preset correlation threshold, then the CTA image is determined as the reference CTA image. The total number of all CTA images acquired from the target patient is determined as the number of images representing the target patient. The ratio of the number of reference CTA images to the number of images represented is determined as the high correlation ratio coefficient corresponding to the target patient; Based on the correlation degree of the blood pressure image corresponding to the current CTA image and the high correlation ratio coefficient corresponding to the target patient, the overall correlation coefficient corresponding to the target patient is determined.

Images