Pulmonary arterial hypertension patient early prognosis prediction method combined with iconography analysis

Abstract

The invention relates to the technical field of pulmonary artery risk assessment, in particular to a pulmonary arterial hypertension patient early prognosis prediction method combined with iconography analysis. The method comprises the following steps: dividing an analysis period of each patient into systole periods, and obtaining different motion behavior sets according to the area difference of standard right ventricle images in different systole periods of each patient and the position difference of the standard right ventricle images in the corresponding systole period; selecting an abnormal behavior set of the patients according to the area difference of the standard right ventricle images belonging to the same movement behavior set of each patient and other patients in the corresponding cardiac contraction cycles; and performing prognosis prediction on the patient according to the abnormal degree difference and the number difference of the standard right ventricle images in the abnormal behavior set of each patient and other patients. The condition that different patients are in the same motion state of the heart is analyzed, and the prognosis accuracy of the pulmonary arterial hypertension patients is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of pulmonary artery risk assessment, and in particular to a method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis. Background Art

[0002] Pulmonary hypertension typically progresses slowly, but as the disease progresses, the risk of right heart failure increases. Failure to intervene promptly can lead to serious complications. Early prognosis prediction allows for tailored treatment plans to slow disease progression.

[0003] Existing imaging methods for predicting the prognosis of patients with pulmonary hypertension use cardiac morphology from multiple patients at a single moment. Based on the differences in right ventricular dilation across these patients, the severity of each patient's pulmonary hypertension is analyzed to predict the patient's prognosis. However, the right ventricle deforms differently under different states of cardiac dilation. This results in multiple patients experiencing different cardiac dilation patterns at the same moment, making it impossible to measure the severity of pulmonary hypertension abnormalities. Consequently, the accuracy of prognostic predictions for patients with pulmonary hypertension is low. Summary of the Invention

[0004] In order to solve the technical problem that the accuracy of prognosis prediction for patients with pulmonary hypertension is low due to differences in right ventricular function under different cardiac expansion states, the present invention aims to provide a method for early prognosis prediction for patients with pulmonary hypertension combined with imaging analysis. The technical solution adopted is as follows:

[0005] The present invention proposes a method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis, the method comprising:

[0006] Obtain standard right ventricular images of different patients at each moment during the analysis period;

[0007] The analysis period for each patient was divided into different cardiac contraction cycles;

[0008] The standard right ventricular images of each patient in the analysis period are divided into different motion behavior sets according to the area difference of the right ventricular region in the standard right ventricular images in different cardiac contraction cycles of each patient and the position difference of the standard right ventricular images in the cardiac contraction cycle of the patient;

[0009] The abnormal behavior set of each patient is selected based on the difference in the area of ​​the right ventricular region in the standard right ventricular images of each patient and the rest of the patients belonging to the same movement behavior set during the corresponding cardiac contraction cycles;

[0010] The prognosis of each patient is predicted based on the difference in the degree and number of abnormalities in the standard right ventricular images between each patient and the rest of the patients.

[0011] Furthermore, dividing the analysis period of each patient into different cardiac contraction cycles includes:

[0012] For each patient, the standard right ventricular images of the patient during the analysis period were arranged in time sequence to obtain an image sequence;

[0013] The first element in the image sequence corresponds to the standard right ventricle image as the cycle start image, and the third element corresponds to the standard right ventricle image as the analysis image to be updated. The expansion similarity between the cycle start image and the analysis image to be updated is calculated. If the expansion similarity does not meet a preset condition, the next standard right ventricle image after the analysis image to be updated is used as the new analysis image to be updated.

[0014] When the expansion similarity meets the preset conditions, n standard right ventricular images are selected continuously from the cycle start image and the analysis image to be updated, respectively, to form the start image sequence and the analysis image sequence in sequence; it is judged whether the expansion similarity of the standard right ventricular images with the same subscript in the start image sequence and the analysis image sequence both meet the preset conditions, if not, the next adjacent standard right ventricular image of the analysis image to be updated is used as the new analysis image to be updated; if so, the corresponding moments of the cycle start image and the previous standard right ventricular image of the analysis image to be updated are respectively used as the start time and end time of the first cardiac contraction cycle, and the analysis image to be updated is used as the new cycle start image, the new analysis image to be updated is determined, and it is judged whether the expansion similarity of the standard right ventricular images with the same subscript in the start image sequence of the new cycle start image and the analysis image sequence of the new analysis image to be updated both meet the preset conditions, the image sequence is traversed, and the analysis period is divided into different cardiac contraction cycles.

[0015] Furthermore, the method for obtaining the expansion similarity includes:

[0016] For each standard right ventricle image, at the edge of the right ventricle region in the standard right ventricle image, the absolute value of the difference between the horizontal coordinates of the pixel coordinates of the two edge pixels that are farthest apart in the horizontal direction is used as the width of the right ventricle region, and the absolute value of the difference between the vertical coordinates of the pixel coordinates of the two edge pixels that are farthest apart in the vertical direction is used as the length of the right ventricle region; the sum of the width and the length is recorded as the local dilation value of the standard right ventricle image;

[0017] Obtaining an expansion index of the standard right ventricular image according to the total number of pixels in the right ventricular region of the standard right ventricular image and the local expansion value; wherein both the total number of pixels in the right ventricular region and the local expansion value are positively correlated with the expansion index;

[0018] The absolute value of the difference between the expansion index of the cycle start image and the analysis image to be updated is negatively correlated and normalized to obtain the expansion similarity between the cycle start image and the analysis image to be updated.

[0019] Furthermore, the standard right ventricular image of each patient during the analysis period is divided into different motion behavior sets, including:

[0020] The time interval between each moment of each cardiac contraction cycle and the moment corresponding to the maximum value of the dilation index among all moments is recorded as the target dilation duration at each moment in each cardiac contraction cycle; the proportion of the target dilation duration in each cardiac contraction cycle is recorded as the dilation ratio at each moment in each cardiac contraction cycle;

[0021] For each patient, one cardiac contraction cycle of the patient is randomly selected as a reference cycle, and a negative correlation mapping is performed on the absolute value of the difference between the expansion ratio at each moment in the reference cycle and each moment in each of the remaining cardiac contraction cycles. The product of the mapping result and the expansion similarity is used as the behavior matching coefficient between each moment in the reference cycle and each moment in each of the remaining cardiac contraction cycles.

[0022] A movement behavior set of the patient is formed by the standard right ventricular image at each moment in the reference cycle and the standard right ventricular image at the moment corresponding to the maximum value of the behavior matching coefficient between each moment and all moments in each other cardiac contraction cycle.

[0023] Furthermore, selecting the abnormal behavior set of each patient includes:

[0024] A patient is randomly selected as a target patient, and a control period of each cardiac contraction cycle of the target patient and a motion control moment of each moment in each cardiac contraction cycle within each control period are determined;

[0025] Calculating the absolute value of the difference between the mean value of the dilation index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient and the mean value of the dilation index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient and the motion control moment in all control cycles of the target patient as the local abnormality index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient; normalizing the mean value of the local abnormality index of all standard right ventricular images in each motion behavior set of the target patient to obtain a comprehensive abnormality index for each motion behavior set of the target patient;

[0026] Based on the comprehensive abnormality index, an abnormal behavior set is selected from all motor behavior sets of the target patient.

[0027] Furthermore, determining the control period of each cardiac contraction cycle of the target patient, and the motion control moment of each moment in each cardiac contraction cycle within each control period, includes:

[0028] The time interval between the start time of each cardiac contraction cycle of each patient and the start time of the analysis period is used as a position index of the corresponding cardiac contraction cycle; a reference period for each cardiac contraction cycle of the target patient is determined; the reference period is the cardiac contraction cycle corresponding to the minimum absolute value of the difference between the position indexes of all cardiac contraction cycles of the target patient and each cardiac contraction cycle of the target patient;

[0029] Obtain a comprehensive dilation value for each movement behavior set; calculate the comprehensive dilation value of the movement behavior set at each moment in each cardiac contraction cycle of the target patient, and the absolute value of the difference between the comprehensive dilation value of the movement behavior set at all moments in each control cycle of each cardiac contraction cycle of the target patient, and select the moment corresponding to the minimum absolute value of the difference from each control cycle, and record it as the movement control moment for each moment in each cardiac contraction cycle of the target patient in each control cycle.

[0030] Furthermore, the prognosis prediction for each patient is performed based on the difference in the degree of abnormality and the number of abnormalities in the standard right ventricular images in the abnormal behavior set between each patient and the rest of the patients, including:

[0031] Recording the corresponding moment of the standard right ventricular image in all abnormal behavior sets of each patient as the abnormal moment; obtaining basic right ventricular images of different cardiac ultrasound sections of each patient at each abnormal moment; performing three-dimensional reconstruction on the basic right ventricular images of all cardiac ultrasound sections at each abnormal moment to obtain a three-dimensional model of the right ventricle at each abnormal moment; selecting the maximum value of the section areas obtained by cutting the three-dimensional right ventricular model at different horizontal planes and recording it as the abnormal section value at each abnormal moment; and taking the average of the abnormal section values ​​of all abnormal moments of each patient as the overall abnormal section value of each patient;

[0032] Obtaining an expansion abnormality index for each patient based on the total number of abnormal behavior sets for each patient, the overall value of the abnormal section, and the minimum value of the comprehensive abnormality index of all abnormal behavior sets; the total number of abnormal behavior sets, the overall value of the abnormal section, and the minimum value are all positively correlated with the expansion abnormality index;

[0033] The difference between the abnormal expansion index of each patient and the mean of the abnormal expansion index of the remaining patients is normalized to obtain the prognostic evaluation value of each patient.

[0034] Furthermore, n is equal to a result of rounding up half of the number of images between the cycle start image and the analysis image to be updated.

[0035] Furthermore, the preset condition is that the expansion similarity is greater than a preset similarity threshold.

[0036] Furthermore, the comprehensive dilation value is equal to the average of the dilation indices of the standard right ventricular images in each movement behavior set.

[0037] The present invention has the following beneficial effects:

[0038] In an embodiment of the present invention, the analysis period of the patient is accurately divided into different cardiac contraction cycles. In order to facilitate the analysis of abnormal deformation of the right ventricle, standard right ventricular images of the patient's heart in the same operating state during all cardiac contraction cycles are extracted to obtain the patient's movement behavior set; by the difference in the area of ​​the right ventricular region in the standard right ventricular images of each patient and the other patients belonging to the same movement behavior set during the corresponding cardiac contraction cycles, the abnormal expansion of each patient and the other patients at the moment of the same cardiac movement behavior is analyzed, and the patient's abnormal behavior set is screened; the number of abnormal behavior sets and the degree of abnormality of the standard right ventricular images in the abnormal behavior set can both reflect the abnormal expansion of the patient's right ventricle during the analysis period, and the difference in the abnormal expansion of the right ventricle between each patient and the other patients can measure the degree of abnormal expansion of the right ventricle of each patient, thereby predicting the prognosis of the patient, and effectively improving the accuracy of predicting the prognosis of patients with pulmonary hypertension. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0040] Figure 1 A flowchart of the steps of a method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis provided by one embodiment of the present invention;

[0041] Figure 2 A flowchart of a method for obtaining an abnormal behavior set provided by one embodiment of the present invention;

[0042] Figure 3 A flowchart of a method for obtaining a prognostic evaluation value provided by one embodiment of the present invention;

[0043] Figure 4 A schematic diagram of a computer device for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0044] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effectiveness of a method for predicting the early prognosis of patients with pulmonary arterial hypertension using combined imaging analysis, according to the present invention. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0045] Unless defined otherwise, 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 belongs.

[0046] The specific scheme of the method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis provided by the present invention is described in detail below with reference to the accompanying drawings.

[0047] Example 1:

[0048] This invention proposes a method for predicting the early prognosis of patients with pulmonary hypertension by combining imaging analysis. Figure 1 , which shows a flowchart of the steps of a method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis provided by one embodiment of the present invention, the method comprising:

[0049] Step S1: Acquire standard right ventricular images of different patients at each moment within the analysis period.

[0050] Patients initially diagnosed with pulmonary hypertension within the hospital were analyzed. Within each analysis period, cardiac ultrasound examinations were performed using a three-dimensional ultrasound diagnostic device to obtain basic echocardiographic images from different cardiac ultrasound sections. These sections included the apical four-chamber view, the parasternal left ventricular long-axis view, the parasternal right ventricular inflow tract long-axis view, and the subxiphoid four-chamber view. Patients were placed in a supine position and maintained calm breathing throughout the examination.

[0051] Pulmonary hypertension can lead to excessive resistance in the right ventricle, gradually causing right ventricular hypertrophy. As the disease progresses, it can cause right ventricular enlargement, ultimately leading to right ventricular failure. To predict the prognosis of patients with pulmonary hypertension, it is necessary to focus on the right ventricular region in baseline echocardiograms. In an embodiment of the present invention, each baseline echocardiogram is input into a trained neural network, which outputs a baseline echocardiogram with the right ventricular region labeled, referred to as a baseline right ventricular image.

[0052] Since the apical four-chamber view can clearly display the four chambers of the heart and their internal structures macroscopically, and is convenient for observing and evaluating the functional status of the right ventricle, the basic right ventricular image of the apical four-chamber view of each patient at each moment during the analysis period is recorded as the standard right ventricular image at each moment.

[0053] In one implementation of the embodiment of the present invention, the duration of the analysis period is set to 15 minutes.

[0054] In one implementation of the embodiment of the present invention, the frequency of the three-dimensional ultrasonic diagnostic apparatus is set to 3 MHz.

[0055] Step S2: Divide the analysis period of each patient into different cardiac contraction cycles.

[0056] When the patient is in a calm state, the expansion and change trends of the heart in different contraction cycles are relatively similar, but there may be differences in different contraction cycles. Based on the similarity of the expansion and change trends of the heart in different contraction cycles, the patient's analysis period can be accurately divided into different cardiac contraction cycles.

[0057] Preferably, in some possible implementations of the embodiments of the present invention, the method for acquiring the cardiac contraction cycle includes: for each patient, arranging the patient's standard right ventricular images within the analysis period in time sequence to obtain an image sequence; using the standard right ventricular image corresponding to the first element in the image sequence as the cycle start image, and using the standard right ventricular image corresponding to the third element as the analysis image to be updated, calculating the expansion similarity between the cycle start image and the analysis image to be updated, and when the expansion similarity does not meet the preset condition, using the next standard right ventricular image of the analysis image to be updated as the new analysis image to be updated; when the expansion similarity meets the preset condition, continuously selecting n standard right ventricular images from the cycle start image and the analysis image to be updated, respectively, to form the start image sequence, the analysis image sequence, and the analysis image sequence in sequence. sequence; determine whether the expansion similarity of the standard right ventricular image with the same subscript in the start image sequence and the analysis image sequence meets the preset conditions, if not, the adjacent next standard right ventricular image of the analysis image to be updated is used as the new analysis image to be updated; if so, the corresponding moments of the cycle start image and the previous standard right ventricular image of the analysis image to be updated are respectively used as the start time and end time of the first cardiac contraction cycle, and the analysis image to be updated is used as the new cycle start image, the new analysis image to be updated is determined, and it is determined whether the expansion similarity of the standard right ventricular image with the same subscript in the start image sequence of the new cycle start image and the analysis image sequence of the new analysis image to be updated meets the preset conditions, traverse the image sequence, and divide the analysis period into different cardiac contraction cycles.

[0058] In some possible implementations of the present invention, a method for obtaining dilation similarity includes: for each standard right ventricular image, at the edge of the right ventricular region in the standard right ventricular image, taking the absolute value of the difference between the horizontal coordinates of the pixel coordinates of the two edge pixels farthest apart in the horizontal direction as the width of the right ventricular region, and taking the absolute value of the difference between the vertical coordinates of the pixel coordinates of the two edge pixels farthest apart in the vertical direction as the length of the right ventricular region; recording the sum of the width and the length as the local dilation value of the standard right ventricular image; obtaining a dilation index of the standard right ventricular image based on the total number of pixels in the right ventricular region in the standard right ventricular image and the local dilation value; the total number of pixels in the right ventricular region and the local dilation value are both positively correlated with the dilation index; and negatively correlating and normalizing the absolute value of the difference between the dilation indexes of the cycle start image and the analysis image to be updated to obtain the dilation similarity between the cycle start image and the analysis image to be updated. Pixel coordinates are well known to those skilled in the art and are not described in detail here.

[0059] The area of ​​the right ventricular region reflects the degree of right ventricular dilation. In this embodiment, the length and width of the right ventricular region indirectly reflect the area of ​​the right ventricular region, while the total number of pixels within the right ventricular region directly reflects the area of ​​the right ventricular region. Combining these two factors makes right ventricular region area analysis more accurate. Since the greater the degree of right ventricular dilation, the larger the area of ​​the right ventricular region, the local dilation value and the total number of pixels within the right ventricular region are positively correlated with the dilation index.

[0060] In a specific implementation of the embodiment of the present invention, the expansion index KZ of the standard right ventricle image is expressed by the formula:

[0061] KZ=Norm((w+h)×Num)

[0062] Where w is the width of the right ventricular region in the standard right ventricular image; h is the length of the right ventricular region in the standard right ventricular image; w+h is the local dilation value of the standard right ventricular image; Num is the total number of pixels in the right ventricular region in the standard right ventricular image; Norm is the normalization function.

[0063] This embodiment uses an exponential function with a natural constant as its base to negatively correlate and normalize the absolute difference between the dilation index of the cycle start image and the analysis image to be updated. Specifically, the method involves first taking the inverse of the absolute difference between the dilation index of the cycle start image and the analysis image to be updated, and using this inverse as the exponent of the exponential function with a natural constant as its base, thereby achieving negative correlation and normalization. Other embodiments may also use other normalization methods and negative correlation mapping methods, which are not limited here.

[0064] In a patient's resting state, cardiac dilation trends are relatively regular across different systolic cycles. Specifically, the degree of dilation at the same location within different systolic cycles is highly similar. The dilation index measures the degree of right ventricular dilation. For any two moments, if the degree of dilation in standard right ventricular images at multiple consecutive moments after the two moments is highly similar, indicating that the dilation trends of some standard right ventricular images after the two moments are relatively similar, the likelihood that the period between the two moments is a single systolic cycle is greater.

[0065] As an example, the image sequence PX = (a1, a2, a3, a4, a5, a6, a7, a8, a9, a10), where a1, ..., a10 represent the right atrium images of the patient during the analysis period. Take a1 as the cycle start image and a3 as the analysis image to be updated. If the expansion similarity KS between a1 and a3 is 1,3≤preset similarity threshold Y, indicating that the expansion similarity between a1 and a3 is low, and the corresponding moment of a3 is not the start moment of the next cardiac contraction cycle of the cardiac contraction cycle of a1, so a4 is selected and then analyzed with a1 for expansion similarity.

[0066] If the expansion similarity KS of a1 and a3 1,3 >Y, indicating that the corresponding moments a1 and a3 may be the start moments of two adjacent cardiac contraction cycles, and the starting image sequence of a1 (a2) and the analysis image sequence of a3 (a4) are determined. The expansion similarity KS of the standard right ventricular images of the elements with the same subscript in the starting image sequence (a2) and the analysis image sequence (a4), i.e., a2 and a4, is determined. 2,4 Are they both less than or equal to Y? If not, it means that the similarity of the expansion change trends of the two adjacent cardiac contraction cycles is low, then the period between the corresponding moments a1 and a2 is not a cardiac contraction cycle, and a4 is used as a new analysis image to be updated; if yes, it means that the similarity of the expansion change trends of the two adjacent cardiac contraction cycles is high, then the period between the corresponding moments a1 and a2 is the first cardiac contraction cycle.

[0067] Take a3 as the new cycle start image and a5 as the new analysis image to be updated. Assume that the expansion similarity of a3 with a5 to a7 is less than Y, and the expansion similarity of a3 with a8 is KS 3,8 >Y, the new image to be updated is a8; the starting image sequence of a3 (a4, a5) and the analysis image sequence of a8 (a9, a10), if the expansion similarity KS of a4 and a9 4,9 , the expansion similarity KS of a5 and a10 5,10 If both are greater than Y, then the period between a3 and a7 is the second cardiac contraction cycle. Only a10 remains and no further analysis is required.

[0068] It should be noted that the preset condition is that the expansion similarity is greater than a preset similarity threshold; in the embodiment of the present invention, the subscript of the new image to be updated for analysis is greater than the new cycle start image, and there is one element between the two images.

[0069] In one implementation of the embodiment of the present invention, the preset similarity threshold Y is set to 0.8.

[0070] In one implementation of the embodiment of the present invention, n is equal to a result of rounding up half of the number of images between the cycle start image and the analysis image to be updated.

[0071] Step S3: Divide the standard right ventricular image of each patient in the analysis period into different motion behavior sets based on the area difference of the right ventricular region in the standard right ventricular image in different cardiac contraction cycles of each patient and the position difference of the standard right ventricular image in the cardiac contraction cycle.

[0072] The duration of different cardiac contraction cycles of patients varies. In order to facilitate the analysis of abnormal deformation of the right ventricle, the patient's cardiac movement during the analysis period is divided into different movement behaviors.

[0073] Because the expansion and change trends of the patient's right ventricle in different cardiac contraction cycles are relatively similar, the right ventricle at the same position in different cardiac contraction cycles may be in the same movement behavior; the area difference of the right ventricular region in the standard right ventricular images in different cardiac contraction cycles and the position difference of the standard right ventricular image in its cardiac contraction cycle respectively reflect the similarity of the degree of right ventricular expansion and the position similarity of the standard right ventricular image in the cardiac contraction cycle, and the standard right ventricular images in different cardiac contraction cycles belonging to the same right ventricular movement behavior are divided to obtain a movement behavior set.

[0074] Preferably, in some possible implementations of the present invention, the method for obtaining a motion behavior set includes: recording the time interval between each moment of each cardiac contraction cycle and the moment corresponding to the maximum value of the expansion index among all moments as the target expansion duration at each moment in each cardiac contraction cycle; taking the proportion of the target expansion duration in each cardiac contraction cycle as the expansion ratio at each moment in each cardiac contraction cycle; for each patient, any cardiac contraction cycle of the patient is recorded as a reference cycle, and each moment in the reference cycle is negatively correlated with the expansion ratio of each moment in each of the remaining cardiac contraction cycles, and the product of the mapping result and the expansion similarity is used as the behavior matching coefficient of each moment in the reference cycle with each moment in each of the remaining cardiac contraction cycles; a motion behavior set of the patient is composed of the standard right ventricular image at each moment in the reference cycle and the standard right ventricular image at the moment corresponding to the maximum value of the behavior matching coefficient of each moment with all of the remaining cardiac contraction cycles.

[0075] Because the duration of different cardiac contraction cycles varies, there is an error in measuring the position of each moment in the cardiac contraction cycle by the time interval between the start of the cardiac contraction cycle and each moment. Each cardiac contraction and relaxation process constitutes a cardiac contraction cycle. The right ventricle has only one moment corresponding to its maximum dilation within the cardiac contraction cycle, and its position is determined by the duration of the cardiac contraction cycle. Measuring the position of each moment in the cardiac contraction cycle by measuring the time interval between each moment in the cardiac contraction cycle and the moment corresponding to the maximum dilation index among all moments can reduce the possibility of deviation in the position of each moment in the cardiac contraction cycle.

[0076] In a specific implementation of the embodiment of the present invention, the behavior matching coefficient is expressed as follows:

[0077] u (jz,t),(a,t′)=KS (jz,t),(a,t′) ×exp(-|BT jz,t -BT a,t′ |)

[0078] Where u (jz,t),(a,t′) The t-th moment in the patient's baseline cycle and the t-th moment in the remaining a-th cardiac contraction cycle ′ The behavioral matching coefficient at each moment; jz is the patient's baseline period; KS (jz,t),(a,t′) The standard right ventricular image at the tth moment in the patient's baseline cycle and the standard right ventricular image at the tth moment in the remaining a-th cardiac contraction cycle ′ The expansion similarity of the standard right ventricular image at the moment; BT jz,t is the dilation ratio at the tth moment in the patient's baseline cycle; BT a,t′ The tth cardiac contraction cycle in the ath cardiac contraction cycle of the patient except the baseline cycle ′ The expansion ratio at each moment; || is the absolute value function; exp is the exponential function with the natural constant as the base. It should be noted that the expansion ratio shows the position of each moment in the cardiac contraction cycle; if |BT jz,t -BT a,t′ The smaller the KS (jz,t),(a,t′) The larger the value, the closer the t-th moment is to the t-th moment. ′ The greater the probability that the right ventricle is in the same position in the cardiac contraction cycle at the tth moment, and the greater the similarity of the expansion degree of the right ventricle at the two moments, the closer the tth moment is to the tth moment. ′ The greater the possibility that the same motion behavior of the right ventricle at each moment is, the greater the behavior matching coefficient u (jz,t),(a,t′) The bigger.

[0079] The maximum value of the behavior matching coefficient between each moment in the reference cycle and all moments in the remaining cardiac contraction cycles corresponds to the same movement behavior moment of the right ventricle.

[0080] In one implementation of the embodiment of the present invention, the first cardiac contraction cycle of each patient is set as the reference cycle.

[0081] Step S4: selecting an abnormal behavior set for each patient based on the area difference of the right ventricular region in the standard right ventricular image of each patient and the rest of the patients belonging to the same movement behavior set in the corresponding cardiac contraction cycles.

[0082] By analyzing the area difference of the right ventricular region in the standard right ventricular images of each patient and the other patients belonging to the same motion behavior set in the corresponding cardiac contraction cycle, the expansion abnormality of each patient and the other patients at the moment of the same cardiac motion behavior is analyzed to screen the abnormal behavior set.

[0083] See also Figure 2, which shows a flowchart of a method for obtaining an abnormal behavior set provided by one embodiment of the present invention, the method comprising:

[0084] Step S410: Select any patient as a target patient, determine a control period for each cardiac contraction cycle of the target patient, and a motion control moment for each moment in each cardiac contraction cycle within each control period.

[0085] Different patients have different starting times of their analysis periods and different degrees of right ventricular contraction and expansion, which results in errors in determining the corresponding systolic cycles of different patients by following the order of the systolic cycles within the analysis period, and thus reduces the accuracy of the analysis of abnormal right ventricular movement and expansion. In an embodiment of the present invention, the method for obtaining the control cycle is as follows: the time interval between the starting time of each systolic cycle of each patient and the starting time of the analysis period is used as the position index of the corresponding systolic cycle; the control cycle of each systolic cycle of the target patient is determined; the control cycle is the systolic cycle corresponding to the minimum absolute value of the difference between the position index of each systolic cycle of all the patients to which it belongs and each systolic cycle of the target patient. Each systolic cycle of the target patient and its control cycle are corresponding cycles.

[0086] Obtain a comprehensive dilation value for each movement behavior set; calculate the comprehensive dilation value of the movement behavior set at each moment in each cardiac contraction cycle of the target patient, and the absolute value of the difference between the comprehensive dilation value of the movement behavior set at all moments in each control cycle of each cardiac contraction cycle of the target patient, and select the moment corresponding to the minimum absolute value of the difference from each control cycle, and record it as the movement control moment for each moment in each cardiac contraction cycle of the target patient in each control cycle.

[0087] The composite dilation value reflects the overall degree of dilation during each systolic cycle of a patient's right ventricle and is used to determine whether the right ventricular motion of different patients is homogeneous. The comparison of each moment within each systolic cycle of the target patient and the control moment within each control cycle indicates the moment when the right ventricle of the target patient and the other patients exhibited the same motion behavior during the corresponding systolic cycles.

[0088] In one implementation of the embodiment of the present invention, the comprehensive dilation value is equal to the average dilation index of the standard right ventricle image in each movement behavior set.

[0089] Step S420: Calculate the absolute value of the difference between the mean of the expansion index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient and the motion control moment in all control cycles as the local abnormality index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient; normalize the mean of the local abnormality index of all standard right ventricular images in each motion behavior set of the target patient to obtain the comprehensive abnormality index of each motion behavior set of the target patient.

[0090] The mean dilation index of the standard right ventricular image at each moment in each systolic cycle of the target patient, across all control cycles, represents the normal dilation level of the target patient's right ventricle at each moment in each systolic cycle. The greater the difference between the mean dilation level and the target patient's dilation level at each moment in each systolic cycle, the more abnormal the right ventricular dilation at each moment in each systolic cycle, and the larger the local abnormality index. The comprehensive abnormality index measures the overall level of dilation abnormality of the standard right ventricular image for each motion behavior set; a larger comprehensive abnormality index indicates a greater degree of abnormality in the corresponding right ventricular motion behavior for each motion behavior set.

[0091] It should be noted that, in the embodiment of the present invention, the Norm function is used for normalization processing, and other normalization methods may also be selected, which are not limited here.

[0092] Step S430: Based on the comprehensive abnormality index, an abnormal behavior set is selected from all movement behavior sets of the target patient.

[0093] In an embodiment of the present invention, for the comprehensive abnormality indexes of all motor behavior sets of the target patient, the motor behavior set corresponding to the comprehensive abnormality index greater than a preset abnormality threshold is recorded as the abnormal behavior set of the target patient.

[0094] In one implementation of the embodiment of the present invention, the preset abnormality threshold is set to 0.5.

[0095] It should be noted that the screening method for the abnormal behavior set of all patients is the same as the screening method for the abnormal behavior set of the target patient.

[0096] Step S5: Prognosis prediction is performed for each patient based on the difference in the degree of abnormality and the number of standard right ventricular images in the abnormal behavior set between each patient and the rest of the patients.

[0097] The number of abnormal behavior sets in a patient and the degree of abnormality of the standard right ventricular images within the abnormal behavior sets can both reflect the abnormal dilatation of the patient's right ventricle during the analysis period. The difference in the abnormal dilatation of the right ventricle between each patient and the rest of the patients can measure the degree of abnormal dilatation of the right ventricle of each patient, and thus predict the prognosis of the patient.

[0098] See also Figure 3 , which shows a flowchart of a method for obtaining a prognostic evaluation value provided by one embodiment of the present invention, the method comprising:

[0099] Step S510: Record the corresponding moment of the standard right ventricular image in all abnormal behavior sets of each patient as the abnormal moment; obtain the basic right ventricular images of different cardiac ultrasound sections of each patient at each abnormal moment; perform three-dimensional reconstruction on the basic right ventricular images of all cardiac ultrasound sections at each abnormal moment to obtain the three-dimensional model of the right ventricle at each abnormal moment; select the maximum value of the section area obtained by cutting the three-dimensional model of the right ventricle at different horizontal planes, and record it as the abnormal section value at each abnormal moment; and take the average of the abnormal section values ​​of all abnormal moments of each patient as the overall abnormal section value of each patient.

[0100] The baseline right ventricular images from all cardiac ultrasound sections at each abnormal moment are input into 3D reconstruction software such as MeshLab for 3D reconstruction. This results in a 3D model of the right ventricle at the corresponding abnormal moment. The 3D model represents the 3D structure of the right ventricle in an abnormal state of right ventricular dilation. Pulmonary hypertension causes right ventricular dilation, which increases the cross-sectional area of ​​the 3D model in the horizontal plane. A larger maximum cross-sectional area of ​​the 3D model obtained from different horizontal planes at all abnormal moments indicates a greater degree of right ventricular dilation at all abnormal moments and a higher likelihood of abnormal right ventricular dilation during the analysis period.

[0101] Step S520: Obtain the expansion abnormality index of each patient based on the total number of abnormal behavior sets of each patient, the overall value of the abnormal section, and the minimum value of the comprehensive abnormality index of all abnormal behavior sets.

[0102] A greater number of abnormal behavior sets indicates a greater likelihood of abnormal right ventricular motion within a single cardiac cycle, leading to a greater degree of abnormal right ventricular dilation within that cycle. A greater minimum value for the combined abnormality index across all abnormal behavior sets indicates a greater degree of abnormality across all abnormal behaviors. A greater overall abnormal section value indicates a greater likelihood of abnormal right ventricular dilation during the analysis period. Therefore, the total number of abnormal behavior sets, the overall abnormal section value, and the minimum abnormal section value are all positively correlated with the dilation abnormality index.

[0103] In an embodiment of the present invention, the product of the total number of abnormal behavior sets of each patient, the overall value of the abnormal section and the minimum value of the comprehensive abnormal index of the abnormal behavior set is used as the expansion abnormality index of each patient.

[0104] In the embodiment of the present invention, other basic mathematical operations can also be used to construct the correlation between the total number of abnormal behavior sets, the overall value of the abnormal section, and the minimum value and the expanded abnormality index in the comprehensive abnormality index of the abnormal behavior set, such as the sum value, which is not limited or elaborated here.

[0105] Step S530: normalize the difference between the abnormal expansion index of each patient and the mean of the abnormal expansion indexes of the remaining patients to obtain the prognostic evaluation value of each patient.

[0106] The prognostic evaluation value of each patient is expressed as follows:

[0107]

[0108] Where, P i is the prognostic evaluation value of the i-th patient; Q i is the expansion abnormality index of the i-th patient; is the mean of the abnormal expansion index of the rest of the patients except the i-th patient; Norm is the normalization function. It should be noted that patients with pulmonary hypertension usually have a poor prognosis due to right heart failure, that is, the larger the abnormal expansion index, the worse the prognosis of the patient. The normal dilatation level of the right ventricle represents the prognosis of patients with pulmonary hypertension. The larger the value is, the more serious the prognosis of the i-th patient is. i The greater the prognostic evaluation value.

[0109] The greater the prognostic evaluation value, the worse the prognosis of the patient, thus providing medical staff with a reference for the early prognosis of patients with pulmonary hypertension.

[0110] So far, the present invention is completed.

[0111] Example 2:

[0112] The present invention also proposes a computer device schematic diagram of an early prognosis prediction device for patients with pulmonary hypertension combined with imaging analysis, please refer to Figure 4 The computer device includes a memory 601, a processor 602, and a computer program 603 stored in the memory 601 and running on the processor 602, wherein when the processor 602 executes the computer program 603, the computer device can execute any one of the above-mentioned combined imaging analysis methods for predicting the early prognosis of patients with pulmonary arterial hypertension.

[0113] In addition, an embodiment of the present application also protects a device, which may include a memory and a processor, wherein the memory stores executable program code, and the processor is used to call and execute the executable program code to perform an early prognosis prediction method for patients with pulmonary hypertension combined with imaging analysis provided in an embodiment of the present application.

[0114] In this embodiment, the device can be divided into functional modules based on the above-described method examples. For example, each functional module can be mapped to a specific functional module, or two or more functions can be integrated into a single processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and represents only a logical functional division. In actual implementation, other division methods may be used.

[0115] In the case of dividing the modules into modules corresponding to their functions, the device may further include a communication module, a signal analysis module, a complexity analysis module, a positioning module, etc. It should be noted that all relevant contents of the various steps involved in the above method embodiment can be referred to the functional description of the corresponding functional modules and will not be repeated here.

[0116] It should be understood that the device provided in this embodiment is used to execute the above-mentioned method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis, and thus can achieve the same effect as the above-mentioned implementation method.

[0117] In the case of an integrated unit, the device may include a processing module and a storage module. When the device is applied to a device, the processing module may be used to control and manage the operation of the device. The storage module may be used to support the device in executing mutual program codes, etc.

[0118] The processing module may be a processor or controller that implements or executes various exemplary logic blocks, modules, and circuits disclosed herein. The processor may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processing (DSP) and a microprocessor, and the storage module may be a memory.

[0119] Example 3:

[0120] This embodiment also provides a computer-readable storage medium, which stores computer program code. When the computer program code runs on a computer, the computer executes the above-mentioned related method steps to implement a method for predicting the early prognosis of patients with pulmonary arterial hypertension combined with imaging analysis provided in the above embodiment.

[0121] Example 4:

[0122] This embodiment also provides a computer program product. When the computer program product is run on a computer, it enables the computer to execute the above-mentioned related steps to implement the early prognosis prediction method for patients with pulmonary hypertension combined with imaging analysis provided by the above embodiment.

[0123] Among them, the device, computer-readable storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0124] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0125] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0126] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis, characterized in that: The method includes: Obtain standard right ventricular images of different patients at each moment during the analysis period; The analysis period for each patient was divided into different cardiac contraction cycles; The standard right ventricular images of each patient in the analysis period are divided into different motion behavior sets according to the area difference of the right ventricular region in the standard right ventricular images in different cardiac contraction cycles of each patient and the position difference of the standard right ventricular images in the cardiac contraction cycle of the patient; The abnormal behavior set of each patient is selected based on the difference in the area of ​​the right ventricular region in the standard right ventricular images of each patient and the rest of the patients belonging to the same movement behavior set during the corresponding cardiac contraction cycles; The prognosis of each patient is predicted based on the difference in the degree and number of abnormalities in the standard right ventricular images between each patient and the rest of the patients.

2. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 1, characterized in that: The analysis period of each patient is divided into different cardiac contraction cycles, including: For each patient, the standard right ventricular images of the patient during the analysis period were arranged in time sequence to obtain an image sequence; The first element in the image sequence corresponds to the standard right ventricle image as the cycle start image, and the third element corresponds to the standard right ventricle image as the analysis image to be updated. The expansion similarity between the cycle start image and the analysis image to be updated is calculated. If the expansion similarity does not meet a preset condition, the next standard right ventricle image after the analysis image to be updated is used as the new analysis image to be updated. When the expansion similarity meets the preset conditions, n standard right ventricular images are selected continuously from the cycle start image and the analysis image to be updated, respectively, to form the start image sequence and the analysis image sequence in sequence; it is judged whether the expansion similarity of the standard right ventricular images with the same subscript in the start image sequence and the analysis image sequence both meet the preset conditions, if not, the next adjacent standard right ventricular image of the analysis image to be updated is used as the new analysis image to be updated; if so, the corresponding moments of the cycle start image and the previous standard right ventricular image of the analysis image to be updated are respectively used as the start time and end time of the first cardiac contraction cycle, and the analysis image to be updated is used as the new cycle start image, the new analysis image to be updated is determined, and it is judged whether the expansion similarity of the standard right ventricular images with the same subscript in the start image sequence of the new cycle start image and the analysis image sequence of the new analysis image to be updated both meet the preset conditions, the image sequence is traversed, and the analysis period is divided into different cardiac contraction cycles.

3. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 2, characterized in that: The method for obtaining the expansion similarity includes: For each standard right ventricle image, at the edge of the right ventricle region in the standard right ventricle image, the absolute value of the difference between the horizontal coordinates of the pixel coordinates of the two edge pixels that are farthest apart in the horizontal direction is used as the width of the right ventricle region, and the absolute value of the difference between the vertical coordinates of the pixel coordinates of the two edge pixels that are farthest apart in the vertical direction is used as the length of the right ventricle region; the sum of the width and the length is recorded as the local dilation value of the standard right ventricle image; Obtaining an expansion index of the standard right ventricular image according to the total number of pixels in the right ventricular region of the standard right ventricular image and the local expansion value; wherein both the total number of pixels in the right ventricular region and the local expansion value are positively correlated with the expansion index; The absolute value of the difference between the expansion index of the cycle start image and the analysis image to be updated is negatively correlated and normalized to obtain the expansion similarity between the cycle start image and the analysis image to be updated.

4. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 3, characterized in that: The standard right ventricular images of each patient during the analysis period are divided into different motion behavior sets, including: The time interval between each moment of each cardiac contraction cycle and the moment corresponding to the maximum value of the dilation index among all moments is recorded as the target dilation duration at each moment in each cardiac contraction cycle; the proportion of the target dilation duration in each cardiac contraction cycle is recorded as the dilation ratio at each moment in each cardiac contraction cycle; For each patient, one cardiac contraction cycle of the patient is randomly selected as a reference cycle, and a negative correlation mapping is performed on the absolute value of the difference between the expansion ratio at each moment in the reference cycle and each moment in each of the remaining cardiac contraction cycles. The product of the mapping result and the expansion similarity is used as the behavior matching coefficient between each moment in the reference cycle and each moment in each of the remaining cardiac contraction cycles. A movement behavior set of the patient is formed by the standard right ventricular image at each moment in the reference cycle and the standard right ventricular image at the moment corresponding to the maximum value of the behavior matching coefficient between each moment and all moments in each other cardiac contraction cycle.

5. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 3, characterized in that: The selecting of abnormal behavior sets of each patient includes: A patient is randomly selected as a target patient, and a control period of each cardiac contraction cycle of the target patient and a motion control moment of each moment in each cardiac contraction cycle within each control period are determined; Calculating the absolute value of the difference between the mean value of the dilation index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient and the mean value of the dilation index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient and the motion control moment in all control cycles of the target patient as the local abnormality index of the standard right ventricular image at each moment in each cardiac contraction cycle of the target patient; normalizing the mean value of the local abnormality index of all standard right ventricular images in each motion behavior set of the target patient to obtain a comprehensive abnormality index for each motion behavior set of the target patient; Based on the comprehensive abnormality index, an abnormal behavior set is selected from all motor behavior sets of the target patient.

6. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 5, characterized in that: Determining the control period of each cardiac contraction cycle of the target patient, and the motion control moment of each moment in each cardiac contraction cycle within each control period, comprises: The time interval between the start time of each cardiac contraction cycle of each patient and the start time of the analysis period is used as a position index of the corresponding cardiac contraction cycle; a reference period for each cardiac contraction cycle of the target patient is determined; the reference period is the cardiac contraction cycle corresponding to the minimum absolute value of the difference between the position indexes of all cardiac contraction cycles of the target patient and each cardiac contraction cycle of the target patient; Obtain a comprehensive dilation value for each movement behavior set; calculate the comprehensive dilation value of the movement behavior set at each moment in each cardiac contraction cycle of the target patient, and the absolute value of the difference between the comprehensive dilation value of the movement behavior set at all moments in each control cycle of each cardiac contraction cycle of the target patient, and select the moment corresponding to the minimum absolute value of the difference from each control cycle, and record it as the movement control moment for each moment in each cardiac contraction cycle of the target patient in each control cycle.

7. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 5, characterized in that: The prognosis prediction for each patient is performed based on the difference in the degree of abnormality and the number of abnormalities in the standard right ventricular images within the abnormal behavior set between each patient and the rest of the patients, including: Recording the corresponding moment of the standard right ventricular image in all abnormal behavior sets of each patient as the abnormal moment; obtaining basic right ventricular images of different cardiac ultrasound sections of each patient at each abnormal moment; performing three-dimensional reconstruction on the basic right ventricular images of all cardiac ultrasound sections at each abnormal moment to obtain a three-dimensional model of the right ventricle at each abnormal moment; selecting the maximum value of the section area obtained by cutting the three-dimensional right ventricular model at different horizontal planes and recording it as the abnormal section value at each abnormal moment; and taking the average of the abnormal section values ​​of all abnormal moments of each patient as the overall abnormal section value of each patient; Obtaining an expansion abnormality index for each patient based on the total number of abnormal behavior sets for each patient, the overall value of the abnormal section, and the minimum value of the comprehensive abnormality index of all abnormal behavior sets; the total number of abnormal behavior sets, the overall value of the abnormal section, and the minimum value are all positively correlated with the expansion abnormality index; The difference between the abnormal expansion index of each patient and the mean of the abnormal expansion index of the remaining patients is normalized to obtain the prognostic evaluation value of each patient.

8. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 2, characterized in that: The n is equal to the result of rounding up half of the number of images between the cycle start image and the analysis image to be updated.

9. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 2, characterized in that: The preset condition is that the expansion similarity is greater than a preset similarity threshold.

10. The method for predicting the early prognosis of patients with pulmonary hypertension combined with imaging analysis according to claim 6, characterized in that: The comprehensive expansion value is equal to the average of the expansion indexes of the standard right ventricle images in each movement behavior set.

Images