Quantitative assessment method for myocardial stress in breast cancer chemotherapy

Abstract

The invention relates to a breast cancer chemotherapy myocardial stress quantitative evaluation method, which comprises the following steps of: acquiring an electromechanical response signal of a myocardial tissue of a chemotherapy patient in a pulsation period, extracting an electromechanical coupling delay parameter of a local tissue, and identifying an initial region of a chemotherapy-induced myocardial stress reaction; for the identified stress starting region and the adjacent cardiac muscle tissue, acquiring the transcellular membrane response time difference, and calculating a stress conduction lag difference index for characterizing the chemotherapy-induced asymmetric stress conduction characteristic; controllable micro-load intervention is carried out on a patient, and a strain reaction curve of a stress starting area and peripheral tissues after the load is applied is monitored; based on the response lag time, the strain peak change rate and the dynamic trend of recovering to the baseline in the load response process, the stress response elastic coefficient of the myocardial tissue is obtained through inversion and is used for quantifying the stress level of the myocardial after breast cancer chemotherapy; tiny changes of myocardial functions are quantified, and a trend judgment basis is provided.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for evaluating myocardial stress induced by cancer chemotherapy, in particular, a method for quantitatively evaluating myocardial stress induced by breast cancer chemotherapy. BACKGROUND

[0002] Currently, for the risk monitoring of early myocardial injury after breast cancer chemotherapy, the existing technology such as the piRNA marker for diagnosing breast cancer chemotherapy-induced cardiac injury, kit and application of China patent CN116694761A, the breast cancer chemotherapy myocardial injury evaluation method mainly relies on the expression change of the molecular biomarker piRNA (such as piR-hsa-31238) in the blood plasma as a diagnostic means, although this method has certain diagnostic value and research basis, but still has the following significant deficiencies and limitations. First, this method is essentially a static molecular level detection, mainly based on the expression change of a certain piRNA molecule in the blood sample, lacking real-time and dynamic quantitative evaluation ability of the actual functional state of myocardial tissue, unable to reflect the mechanical, electrophysiological and structural integration response of myocardium under physiological load or drug stress, which limits its wide applicability in accurately evaluating the influence of chemotherapy on the physiological function of heart; secondly, the expression of piRNA is greatly interfered by many systemic factors, such as inflammation, immune response, other combined chronic disease states, which may affect the concentration of piRNA in peripheral blood plasma, causing false positive or false negative results, so its specificity and stability are insufficient, and it cannot be used as a single indicator for clinical decision-making.

[0003] In addition, this method lacks the ability to describe the differences in local myocardial regions, cannot identify whether there is regional dysfunction, intercellular conduction barrier or stress diffusion asymmetry, and cannot reveal the spatial propagation mechanism in the process of myocardial microenvironment change induced by chemotherapy, while cardiac injury, especially drug-induced myocardial toxicity, often presents a dynamic process from focal electromechanical coupling abnormality to overall functional compensation imbalance, and the existing piRNA marker strategy cannot capture the key turning points in this process; in addition, this scheme cannot solve the problem of longitudinal comparison of individual patients, that is, it does not continuously track the stress response trend of the same patient before and after chemotherapy or between different cycles, nor does it introduce a time series pairing mechanism for changes in chemotherapy dose, so there is a significant blind area in efficacy evaluation and individual early warning; from the technical implementation point of view, the existing piRNA detection needs to rely on high-throughput chip, RT-PCR and other experimental equipment, the process is relatively complex, the detection period is long, and it does not have real-time response capability, making it difficult to achieve high-frequency monitoring in actual hospitalization or follow-up management. In addition, this method does not combine the coupling relationship between myocardial electrophysiological activity and mechanical function, and does not use the widely available non-invasive imaging or physiological signal means (such as electrocardiogram, ultrasonic strain imaging, etc.) to establish a dynamic response model, so it cannot realize comprehensive interpretation under multi-source data fusion, and the information dimension is single. SUMMARY

[0004] The purpose of the present application is to provide a breast cancer chemotherapy myocardial stress quantitative evaluation method, so as to solve some of the problems and deficiencies pointed out in the background art.

[0005] The present application solves the above technical problems by adopting the following technical solution: a breast cancer chemotherapy myocardial stress quantitative evaluation method, comprising: obtaining the electromechanical response signal of the myocardial tissue of the chemotherapy patient in the cardiac cycle, extracting the electromechanical coupling delay parameter of the local tissue, and identifying the starting area of the chemotherapy-induced myocardial stress response; collecting the response time difference between the cell membranes of the identified stress starting area and the adjacent myocardial tissue, and calculating the stress conduction lag difference index to represent the asymmetric stress conduction characteristics induced by chemotherapy; A controllable micro-load intervention is performed on the patient, and the strain response curve of the stress starting area and the surrounding tissue after the load is applied is monitored; based on the response lag time, the strain peak value change rate and the dynamic trend of recovery to the baseline in the load response process, the stress response elastic coefficient of the myocardial tissue is obtained by inversion, which is used to quantify the stress level of the myocardium after breast cancer chemotherapy.

[0006] Further, the collection of the electromechanical response signal includes simultaneously recording the local depolarization signal and the synchronous mechanical strain rate signal of the myocardial tissue to form a cardiac cycle data pair for synchronous comparative analysis; the electromechanical coupling delay parameter is based on the adaptive adjustment of the reference point according to the waveform form in the cardiac cycle to adapt to the alignment error caused by the change of electrocardiogram or strain form after chemotherapy.

[0007] Further, the identification of the stress starting area includes that the electromechanical coupling delay deviates from the basic threshold in three consecutive cardiac cycles, and the region has spatial boundary stability; a one-dimensional lag chain is established based on the intercellular conduction path radially spreading from the center of the starting area to the adjacent area, which is used to mark the conduction directionality; wherein the stress conduction lag difference index includes the bidirectional conduction delay difference ratio between the local adjacent myocardial units, so as to reveal the potential asymmetric diffusion trend.

[0008] Further, the micro-load intervention is in the form of stage-by-stage increasing load, and the disturbance is applied step by step at multiple increasing levels to obtain the stress threshold shift point of the strain response; wherein the strain rate variation difference in the time window of the starting area and the peripheral area is included in the response collection of the load intervention, which is used to evaluate the consistency of the load response.

[0009] Further, the strain response curve after the load response is comprehensively analyzed by calculating the peak lag, the acceleration starting point offset and the recovery trend covariance; wherein the inversion of the stress response elastic coefficient includes two-dimensional mapping of the response slope and the recovery speed curve under the condition of increasing load, and identifying the uncompensated interval.

[0010] Furthermore, the electromechanical delay data, intercellular hysteresis chain data, and load response data are simultaneously entered into the joint feature space for multidimensional dynamic comparison; and the stress response elastic coefficient generated in each assessment is subjected to longitudinal time pairing analysis with historical records to determine the trend of myocardial function changes; wherein the time nodes for collecting electromechanical response signals are selected within 48 hours before and after the first, third, and last chemotherapy cycles to capture the inflection point of stress response initiation. A dynamic mapping function constructed by combining electromechanical delay (EMD), intercellular hysteresis chain (ICDL), and load response (LRD) data is used to continuously assess the stress response elastic trend of the myocardium during breast cancer chemotherapy. Based on the joint feature space structure of multi-modal data, a nonlinear integral and mapping function is constructed to dynamically invert the trend of myocardial function changes and capture the inflection point of stress response. in: Current moment The trend judgment result of the myocardial stress evolution function value is used for subsequent function descent point detection; The integral variable represents the historical response of past time segments; The first-order dynamic disturbance amplitude of the strain response curve under load response is derived from LRD data; Intercellular hysteresis chain length over time rate of change (i.e. The implicit mapping originates from the ICDL data structure; Electromechanical delay at time The asymmetric perturbation factor is defined as the difference between the local maximum and minimum values, which originates from the EMD data distribution. The time decay factor is used to suppress the influence of old data on system settings or adaptive training and learning. Introduce a time-distance weighting factor to ensure that the most recent moment has a greater impact on the time-sensitive mapping kernel; by The main function continuously calculates the stress evolution trend of the myocardium at key nodes in different chemotherapy cycles (1st, 3rd, and 48 hours before and after the last chemotherapy cycle); this function has four core variable control dimensions: response intensity, propagation path, initiation delay, and historical decay weight; denominator structure Used to stabilize the nonlinear amplification effect of abnormal delay peaks on the overall trend; when If the stress inflection point is identified successfully, and the stress level continues to decline. The above formula constructs a nonlinear integral function from the perspective of physiological dynamic systems, at each time step... The output is a scalar value reflecting the current stress evolution trend of the myocardium; from the perspective of mathematical modeling, to ensure the function has historical sensitivity and local response capability, an integral structure with decaying memory is selected, with a time-varying variable representing all past response inputs: wherein, represents an input item of a certain stress function, is a weight kernel function used to adjust the influence of history on the present; combined with the physiological time characteristics, a typical asymmetric weight kernel, i.e., an exponential decay form, is adopted , wherein controls the degree of attenuation of the long-term historical response; defines the internal driving function , in fact, the evolution of myocardial stress is essentially driven by three core dimensions: 1. Load strain disturbance intensity : local strain mutation triggered by active or passive load, representing the real mechanical response of muscle fibers; 2. Cell interstitial lag propagation speed : reflects how stress is conducted and diffused between cells, representing the diffusion breadth or synchronism; 3. Heterogeneity degree of electromechanical delay : represents the phase asymmetry between regional electrical-mechanical responses, which increases at the early stage of stress; Among the three, the first two can be regarded as driving items that positively promote stress evolution, while the third can be regarded as a system stability or internal resistance inhibitor (i.e., the higher the heterogeneity, the worse the trend stability); therefore, the combination is as follows: wherein, the denominator adopts form to ensure that even , the function will not be singular, but when the delay is asymmetric, this term will exponentially inhibit the influence of stress evolution; the numerator uses the product of strain disturbance amplitude and interstitial propagation rate, combining local intensity and propagation speed to form a mechanical-time coupling expression; The entire formula maintains dimensional consistency (scalar output) and has continuous differentiability in mathematics; Finally, the above combination is brought into the initial integral framework to obtain the main function expression, which can be understood as the instantaneous myocardial stress trend projection of all multi-modal signal fusion within a certain period of time (from to the current ), the numerical value of which can be used to quantify whether the myocardial elasticity trend is in a downward state, a decompensation state, or a repair state; its first derivative (such as ) can be used to identify the turning point of myocardial function, combined with the time node you originally defined, the trend turning point can be continuously tracked during chemotherapy.

[0011] Further, the construction of the joint feature space adopts a spatial alignment strategy of myocardial cells in each data source to make the anatomical region consistent between electromechanical delay data and intercellular lag chain data; the alignment includes a time gradient alignment process across data sources, and detects myocardial abnormal response acceleration based on the time derivative change rate as an early stress signal.

[0012] Further, if a response waveform morphology fault mutation occurs in the alignment result, the region is marked as a stress non-continuous region, which is used to infer the conduction barrier between myocardial cells; in the alignment process, the stress cells far away from the starting point in the data points respond earlier than the adjacent regions.

[0013] Further, in the longitudinal time pairing analysis, the first derivative of the stress response elastic coefficient change amplitude of each historical cycle of the patient is used as a functional change sensitivity criterion; wherein the longitudinal time pairing analysis introduces a backtracking segmentation strategy to divide the historical data according to the chemotherapy stage to identify the dose turning point.

[0014] Further, the respiratory state and sympathetic activation score in the 48-hour data segment before and after the time node are included as context reference signals for calibrating the myocardial stress baseline state; the time node data acquisition includes three-dimensional myocardial region directional scanning triggered by electrocardio, so that each cycle is collected at the same cardiac phase; the judgment of the stress response starting inflection point takes the mutation of the average length change rate of the lag chain as the main discriminant criterion.

[0015] The beneficial effects of the present application are: by extracting the electromechanical coupling delay parameters of myocardium and combining the intercellular lag chain to construct the conduction directionality, the anatomical region where the stress response first occurs can be located before the traditional imaging or biochemical indicators show obvious abnormalities, thereby providing a forward warning for cardiotoxicity. By incremental micro-load intervention and strain response curve analysis, the present application introduces the stress response elastic coefficient and constructs a joint feature space, which can continuously track the dynamic changes of myocardial mechanical function, and uses the derivative change rate to identify the downward trend of function, thereby providing a scientific basis for dynamic treatment evaluation and dose adjustment. By identifying the mismatch relationship between response slope and recovery speed through two-dimensional mapping, the critical point of the transition of myocardium from compensation to decompensation can be detected, thereby providing an intervention opportunity for adjusting the chemotherapy regimen (such as drug reduction and introduction of cardioprotective drugs).

[0016] By spatial alignment of myocardial signals, cross-modal time gradient synchronization, and acquisition of three-dimensional images at the same cardiac phase, the method establishes a high-consistency analysis platform at the data fusion level, significantly reduces false positives or missed detections caused by rhythm changes and signal drift. The sympathetic activation score and respiratory state parameters are recorded simultaneously in the acquisition window as background parameters introduced into the model calibration process, which can effectively eliminate false myocardial fluctuations caused by physiological states such as tension and sleep deprivation, and improve the stability and applicability of the results. Through longitudinal time pairing analysis combined with chemotherapy cycle division strategy, the invention not only can observe the myocardial trend trajectory of individuals during the entire treatment process, but also can identify the dose turning point at a specific chemotherapy stage, providing functional evidence for individualized dose optimization. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 Main flow of the breast cancer chemotherapy myocardial stress quantitative evaluation method of the present application.

[0018] Figure 2 Functional relationship diagram for the breast cancer chemotherapy myocardial stress quantitative evaluation method of the present application.

[0019] Figure 3 Multi-modal integrated trend analysis flowchart for the breast cancer chemotherapy myocardial stress quantitative evaluation method of the present application.

[0020] Figure 4 Simple diagram of chemotherapy cycle myocardial monitoring and risk identification for Ms. Wang in Example 1 of the present application.

[0021] Figure 5 Simple diagram of multi-cycle chemotherapy myocardial stress trend analysis and risk identification for Ms. Wang in Example 2 of the present application. DETAILED DESCRIPTION

[0022] The specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0023] In conjunction with the accompanying Figure 1The application provides a breast cancer chemotherapy myocardial stress quantitative evaluation method. In the breast cancer chemotherapy process, subclinical level functional damage of myocardial tissue of a patient is caused by drug toxicity, which is manifested as a change in myocardial stress state. First, an electrocardiogram monitoring device and a high-frame-rate cardiac ultrasound / magnetic resonance imaging system are arranged to obtain synchronous electrical activity signals and corresponding mechanical contraction strain response signals of myocardial tissue of the patient within a single beat cycle. According to a difference between a starting time of a QRS wave segment in the electrical signal and a peak time of the mechanical strain response, an electromechanical coupling delay parameter of each myocardial unit is calculated. The parameter is spatially mapped into a standard AHA left ventricular partition model for block comparison, so as to screen out a myocardial region with significant coupling delay abnormality, which is defined as a starting region of a chemotherapy-induced stress response. Subsequently, an intercellular response path network of a spatially adjacent myocardial region around the starting region is constructed. By collecting activation time sequences of different intercellular units within the same beat cycle, and combining cell spacing measurement and conduction time difference, a propagation speed gradient and a direction change amount of a transmembrane response are calculated, and then a transmembrane response time difference matrix is obtained. A stress conduction lag difference index is constructed based on the matrix. The index is used to quantitatively describe asymmetric and directional diffusion characteristics of myocardial stress propagation in space, including lag difference absolute value mean, main propagation direction vector offset angle, and a ratio of a maximum lag difference to a minimum lag difference. Through these quantitative indexes, a damage degree of electromechanical-structural coupling characteristics of myocardial tissue under the chemotherapy background can be comprehensively described, and early identification of a myocardial functional abnormality diffusion path and a severity can be realized. The analysis process is aligned with a standard myocardial segment model, which is convenient for horizontal case comparison and longitudinal trend tracking.

[0024] On the basis of completing stress initiation zone recognition and intercellular lag chain analysis, in order to further obtain the dynamic response ability of the real stress state of myocardium, the patient is intervened controllably by adopting exogenous micro-load disturbance, and the intervention process is preferably carried out in the quiet state of the patient. The slight mechanical or pharmacological load enhancement of the heart is realized by regulating the intrathoracic pressure change in the respiratory cycle, applying low-dose positive inotropic drugs or guiding passive limb movement, and the real-time monitoring system of cardiac function is started at each stage before, during and after the load application, and the strain response curve of the stress initiation zone and the surrounding myocardial tissue identified previously is recorded, wherein the strain response curve takes the deformation rate of the local tissue per unit time as the vertical axis, and takes the load action time as the reference time axis, to form a complete displacement-time-load dynamic sequence. The collected strain curve is analyzed, and three key dynamic parameters are extracted: one is the response lag time, that is, the time difference between the load application point and the initiation of myocardial strain response, which reflects the sensing and starting speed of the myocardium; the second is the strain peak change rate, that is, the sudden increase amplitude of the strain peak value under the load compared with the baseline state, which reflects the short-time mechanical stretching and contracting ability of the tissue under unit load; the third is the time and trend curve slope of returning to the baseline, which is used to measure the self-regulating ability of the myocardium from the load activated state to the steady state. The above three parameters jointly constitute the elastic response feature group of the region under micro-load disturbance, and the stress response elastic coefficient (Stress Responsivity Coefficient, SRC) of the target myocardial region is further obtained by calculating the time sequence mapping relationship constructed by the three parameters. The coefficient can be defined as the elastic response intensity divided by the normalized product of response lag and recovery time, which is used to quantitatively measure the stress response ability and mechanical recovery ability of the tissue unit under the chemotherapy background. The higher the value of the coefficient is, the better the stress response and functional elasticity of the tissue are. If the coefficient continuously decreases in multiple cycles, it can be prompted that the myocardial region is entering the potential subclinical stage of functional decline. Through this inversion mechanism, without relying on static LVEF and other rough indicators, the functional changes induced by early myocardial toxicity during chemotherapy can be captured, and physiological basis can be provided for subsequent individualized intervention and drug regimen adjustment.

[0025] Combined with the drawings Figure 2To ensure accurate quantification of myocardial stress state after chemotherapy, firstly, the electromechanical response signals of myocardial tissue were synchronously collected when the patient was in a stable resting state, including the depolarization electrical signal and the corresponding mechanical strain rate signal of the local myocardial unit in the same beating cycle. The depolarization signal was captured by a multi-channel electrode array arranged on the surface of a specific segment of the heart to record the changes in myocardial membrane potential. The mechanical strain rate signal was recorded by cardiac ultrasound tissue velocity imaging (TVI) or magnetic resonance tomography (MRItagging) technology to record the instantaneous deformation rate of the corresponding myocardial tissue unit. The two types of signals were synchronously collected at a high time resolution to form a beating level data pair, which served as the basis for subsequent electromechanical coupling analysis. Due to the structural changes in the waveform morphology of myocardial electrical activity and the mechanical contraction characteristics induced by different stages of breast cancer chemotherapy, to avoid delayed misjudgment caused by inaccurate standard reference point setting, a fixed trigger point strategy was not used when extracting the electromechanical coupling delay parameter. Instead, a dynamic reference point was set based on the actual morphological changes of the signal waveform in each beating cycle. In the electrical signal aspect, the potential drop point corresponding to the local maximum negative slope was selected as the depolarization starting point. In the strain curve aspect, the maximum value of the mechanical strain rate or the first phase turning point was identified as the response starting time. Then, the electromechanical coupling delay of the myocardial unit was calculated by the time difference between the two reference points. The above reference point recognition mechanism is updated in real time by introducing an adaptive waveform analysis algorithm, so that the parameter extraction can be compatible with the nonlinear drift or waveform distortion of the electrical activity and strain curve in different chemotherapy stages, improving the robustness and accuracy of signal alignment under abnormal conditions. The delay parameter is not only used for preliminary identification of abnormal coupling areas, but also serves as a core variable input for joint feature space construction, further participating in the comprehensive evaluation process of myocardial function evolution trend together with the lag chain data and load response data.

[0026] After obtaining the electromechanical response signals in the cardiac muscle tissue and extracting the electromechanical coupling delay parameters of each cardiac muscle segment, in order to realize the positioning and identification of the earliest occurrence area of the stress response induced by chemotherapy, a double judgment mechanism of time sequence continuity stability and spatial boundary stability is introduced. Specifically, in the effective cardiac cycle of the patient for three times, the electromechanical coupling delay value of each cardiac muscle segment is calculated, and compared with the delay reference value of the patient before chemotherapy. If the delay of a certain segment exceeds the set basic threshold (such as deviation > 15 ms) and the direction is consistent in three cycles, and at the same time the region presents the distribution characteristics of boundary closure and fixed position on the standard segment model of left ventricle, it is determined that the region is the stress starting region; then taking the gravity center of the stress starting region as the center point, the intercellular response path is expanded to its adjacent segment according to the geometric radial structure, a one-dimensional lag chain with spatial adjacent myocardial units as nodes and signal response time as edge weight is constructed, which records the space-time sequence of signal conduction of myocardial tissue after stress occurs, forms a stress diffusion path with time progressive property, and can be used to infer how the starting stimulus spreads to the surrounding tissue along the muscle bundle; on the basis of the above lag chain, the bidirectional conduction delay difference between each pair of adjacent myocardial units is further calculated, that is, the ratio between the activation time difference from unit A to unit B and the activation time difference from B to A, and defined as local bidirectional delay difference ratio. When the ratio deviates from 1 significantly and exists continuously in the link, it indicates that there is obvious asymmetric diffusion phenomenon in the path, that is, there is difference in the conduction speed of stress signal in different directions. Such asymmetry usually corresponds to structural obstacles, impaired functional electrical activity conduction or local compensatory imbalance area induced by chemotherapy. Through the above lag chain construction and ratio analysis mechanism, not only the spatial confirmation of the stress starting point can be realized, but also the directionality, integrity and conduction stability of its diffusion to the surrounding area can be quantitatively described, which provides high spatiotemporal resolution data support for subsequent elastic response analysis and myocardial function trend modeling.

[0027] To further evaluate its dynamic load-bearing capacity and mechanical adaptability under different stress intensity, a phased incremental micro-load intervention strategy is designed, that is, in the resting state of patients, respiratory regulation, low-dose positive inotropic drug injection or non-invasive chest wall vibration and other means are used to gradually increase the mechanical or metabolic load applied to the heart according to the preset multi-level load grade. After each level of load, high-frame-rate strain rate acquisition of cardiac tissue is performed at a fixed time interval, especially focusing on the strain response of the identified stress initiation area and its corresponding peripheral tissue. During the process of gradually increasing load, the system records the strain peak, strain rate initiation delay and regression rate of the target area under different load levels, and by setting a load response change threshold, such as the strain peak not changing under the continuous two levels of load but jumping under the next level of load, or the strain rate initiation delay significantly increasing at a certain level, the stress threshold transition point of the region is identified. This point represents the turning critical point of the transition from linear response to nonlinear compensation state, indicating that the local myocardium has approached the edge of functional exhaustion. In addition, the strain rate variation difference of the stress initiation area and its surrounding myocardial segments within the same time window is also collected under each level of load, which is used as an index to evaluate the consistency of local load response. If the variation difference is greater than the set standard, such as the local strain response amplitude difference exceeding 15% or the response initiation time deviation exceeding 25ms, the system prompts that there is a mechanical response desynchronization phenomenon in the region caused by myocardial toxic micro-injury. Through the above phased incremental intervention and spatial consistency analysis mechanism, this method not only obtains the nonlinear mechanical characteristics of the tissue under different loads, but also identifies the myocardial region with functional disorders in advance, providing real-time data support and individualized response window judgment for heart protection and drug adjustment strategies during chemotherapy.

[0028] On the basis of implementing the incremental micro-load intervention in the implementation phase and obtaining the myocardial strain rate response data under each load level, in order to more accurately evaluate the dynamic elastic properties of myocardial tissue under external disturbance conditions, the method constructs a comprehensive analysis mechanism for the strain response curve after load response. First, the complete strain response curve after each level of load is extracted, and time domain structure quantization is performed on the curve, including three key dynamic indicators: one is the peak lag, that is, the time interval between the load application point and the strain peak, which reflects the organizational reaction delay; the second is the acceleration starting point offset, which is defined as the time difference between the load starting point and the first appearance point of the strain curve rising section acceleration, which is used to capture the activation threshold delay of myocardial from static to active response state; the third is the recovery trend covariance, which is calculated by calculating the covariance between the slope sequence from the peak point to the recovery to the baseline and the original baseline fitting trend, which quantifies the coupling degree of the organization recovery power and the consistency of the steady trend, the above three indicators are extracted and standardized at each load level and input to the subsequent stress elasticity modeling unit; on this basis, a stress response elasticity coefficient inversion algorithm is further designed, the core idea of which is to map the strain response slope under each level of load and the corresponding recovery speed (i.e. the average slope of the strain recovery section) to a two-dimensional coordinate, construct a response-recovery two-dimensional feature plane, and draw the data point trajectory under different load levels in the feature plane. The trajectory changes can directly reflect the evolution trend of the elastic state of the organization, if the response slope continuously decreases after a certain level of load and the recovery speed slows down at the same time, that is, the two-dimensional feature trajectory shifts to the left lower quadrant, it is identified that the load level is in the non-compensation interval, which indicates that the myocardium cannot effectively resist the mechanical disturbance produced by the load stimulation through the spontaneous mechanism. The point is the critical clinical threshold before the organization enters the functional exhaustion; on the contrary, if the trajectory remains in the right upper quadrant or presents a linear trend, it indicates that the myocardium still maintains good elastic response and compensation ability under the condition of chemotherapy intervention; the stress response elasticity coefficient can finally be fitted to form a single quantitative indicator by weighting the trajectory slope and the trajectory fluctuation amplitude, which is used as the continuous output result of myocardial health level, and can be paired with the previous record for time analysis under multiple chemotherapy cycles, to realize the early identification and quantitative tracking of the functional deterioration trend.

[0029] Combined with the attached Figure 3, in order to realize the dynamic and quantitative evaluation of the myocardial stress state of breast cancer patients during chemotherapy, the method first collects the multi-modal physiological response data of the myocardial tissue of the patient within 48 hours before and after the first, third and last chemotherapy of the patient, which includes electrical-mechanical coupling delay data (EMD), intercellular conduction delay line data (ICDL) and strain response data (LRD) under exogenous micro-load stimulation, the above three types of data are synchronously input into the joint feature space construction module to form a three-dimensional dynamic physiological response matrix, and then multi-dimensional dynamic comparison analysis is carried out, and a kind of nonlinear integral mapping function is constructed through mathematical modeling to continuously map the myocardial function trend. The electrical-mechanical delay heterogeneity, electrical activation diffusion speed and mechanical load response ability are fused and modeled to construct the following core function: Wherein, represents the myocardial stress evolution function value at the current time , which is the main variable of the trend output result, and is used for subsequent downward trend detection; is a historical time variable, which is used for integral rolling; represents the strain disturbance intensity of the myocardial tissue under exogenous load stimulation, which is the function mapping result of LRD data; represents the instantaneous change rate of the intercellular lag chain length at time , which can be analogized as nonlinear fitting representation, which is derived from the dynamic structure of ICDL; is the spatial asymmetric disturbance factor of electrical-mechanical coupling delay at the time point, which is usually represented as the difference between the maximum value and the minimum value of EMD of the segment in the current beating cycle; is a time decay coefficient, which is used to adjust the decay influence of historical data on the current state, and is usually obtained by system adaptive optimization; the exponential decay kernel ensures that the function is more sensitive to recent data and the contribution of long-term data gradually decays, thereby simulating the rapid response characteristics of myocardium to recent stress. The construction basis of the function is the general historical integral response model: Wherein is the stress driving function of the myocardium at a certain time in the past, is an asymmetric time weight kernel, which is set to , which embodies the high attention weight of recent signals, and the driving function part is constructed as follows: The denominator in the function structure adopts form can avoid singularity caused by , while it can suppress trend amplification when the electromechanical delay asymmetry increases sharply, reflecting the system stability feedback logic; The molecular part is the product of strain disturbance and intercellular conduction velocity, representing the comprehensive results of mechanical response strength and propagation breadth of myocardium to external disturbance under unit load input. Bring into the historical response integral structure, that is, form the final stress evolution trend function , the continuous derivative of the function is: , which can be used as a real-time monitoring index of functional decline. When the derivative is continuously negative and the value continues to decline beyond the set threshold, the system determines that the myocardium enters the stress evolution decline phase at the current time, that is, it identifies the trend inflection point or the starting point of potential functional decompensation. This mechanism effectively solves the problem of direct observation of myocardial stress state during breast cancer chemotherapy by integrating multi-modal signals, time dynamic window and nonlinear integral kernel, and is especially suitable for trend prediction and intervention suggestion generation when LVEF is normal but subclinical stress indicators are abnormal in the early stage.

[0030] To realize the collaborative analysis of multi-modal physiological signals in spatial and temporal dimensions, after completing the electromechanical delay data (EMD), intercellular delay chain data (ICDL) and load response data (LRD) collection, first perform the construction operation of joint feature space, specifically: according to the standard myocardial anatomical partition model (such as AHA 17 segment method), the myocardial three-dimensional structure is discretized into multiple functional units, and then the original signals from different data sources are spatially relocated and synchronously mapped, among which the stress characteristic signal points in each type of data are redistributed to the corresponding myocardial segment based on the basic mapping of cardiac magnetic resonance imaging (MRI), echocardiogram (Echo) or CT, so that each delay value in the EMD signal and each conduction chain path in the ICDL signal are directed to the same anatomical structure, so that the signals from different acquisition channels and different physical mechanisms are spatially consistent in the joint feature space. Align, effectively avoid analysis errors and judgment deviations caused by sampling area offset; after completing the spatial alignment, the system performs cross-source time gradient comparison analysis on the time series of each signal, that is, the EMD time series, the ICDL path propagation sequence and the strain rate response sequence in the LRD are jointly flattened in a time window sliding manner, and the first order change rate and the second order derivative change trend of the time derivative are calculated at each time point, thereby constructing a continuous dynamic acceleration spectrum plane, wherein the first order derivative change rate reflects the response speed difference of different signals, and the second order derivative corresponds to the response acceleration of the myocardial unit under the disturbance condition, which is used to find the weak abnormal trend that cannot be detected by traditional amplitude indicators; In particular, the system sets a specific threshold to identify the signal point group of positive mutation of acceleration, that is, when the time derivative of a myocardial unit changes by more than a certain change rate (for example, the original value changes by more than 25% and the second order derivative is positive) within 5ms before and after a certain time point, it is determined that the region has an early abnormal response trend; Through the above spatial alignment and time gradient comparison mechanism, the stress characteristics under different signal sources can be collaboratively analyzed in the same anatomical background and the same time coordinate system, breaking through the correlation failure problem caused by data heterogeneity in the past, and is particularly suitable for early identification of subclinical myocardial dysfunction caused by chemotherapy, and provides accurate, structured and continuous underlying data support for stress trend modeling, response elasticity prediction and risk warning.

[0031] After constructing the joint feature space based on the fusion of electromechanical delay data (EMD), intercellular delay chain data (ICDL), and load response data (LRD), and performing time gradient comparison on each data source, the system further performs morphological continuity analysis and conduction timing anomaly identification on the obtained multi-modal comparison results. Specifically, it includes two key judgment logics: one is waveform morphological mutation detection, and the other is stress response timing anomaly point identification. First, on a continuous myocardial segment, according to the comparison results, the strain response waveform of each region within the same load intervention period is extracted, standardized, and arranged in a spatial sequence to form a set of spatially continuous response curve groups. The system uses the difference function to compare the slopes, amplitudes, and peak times of adjacent segment waveforms in multiple dimensions. When it detects that the curve similarity between a certain segment and its adjacent segments significantly decreases, and the waveform has mutations, discontinuities, or reversals at key feature points (such as peak position, acceleration point, and recovery segment), it can be determined that the region is a response waveform morphological fault zone. The system marks it as a stress discontinuity zone and infers that there is an electrical-mechanical signal conduction disorder or local structural barrier, such as microfiber rupture, connective tissue replacement, or perfusion imbalance caused by chemotherapy, in that area. This judgment is important for identifying potential conduction block areas. Second, on the propagation chain with the same stress starting point as the reference source, the system compares the activation response times of each unit. When it detects that a myocardial unit far from the stress starting point has an activation time earlier than its adjacent regions with closer geographical locations, and this early response phenomenon repeatedly occurs in multiple cycles, the system determines that there is a remote early response phenomenon at that point and marks the unit as an abnormal propagation node, indicating that there is a jumping activation, bypass channel, or abnormal reverse propagation phenomenon in the conduction process of this path. Combined with the geometric distance between this data point and the stress starting area and the ICDL path structure, it can further infer that there is an implicit bypass or atypical diffusion mechanism in the conduction path. Through the collaborative identification of spatial waveform faults and timing response abnormal points in the comparison results, the sensitivity to local pathological conditions of the myocardium is improved, and the ability to infer potential structural barrier areas and abnormal functional activation paths from electrical-mechanical response behavior is achieved, providing a new technical means and judgment basis for precise monitoring and mechanism inference of chemotherapy-related myocardial toxicity micro-damage areas.

[0032] After completing the extraction and mapping of the stress response elasticity coefficient (Stress Responsivity Coefficient, SRC) at each time point, to realize the dynamic identification and critical point judgment of myocardial function changes in the whole process of chemotherapy cycles for breast cancer patients, the method constructs a longitudinal time pairing analysis module. Based on the stress response elasticity coefficient data generated by each historical acquisition cycle of the patient, the module calculates the first-order derivative of the change amplitude between adjacent cycles through mathematical differential operation, i.e., for any two consecutive time points And elasticity coefficient Perform the calculation of the first-order difference or numerical derivative: This first derivative reflects the speed and direction of changes in myocardial function. A larger absolute value indicates a more dramatic change in functional status at that stage. Therefore, this method sets the absolute value of this derivative as a sensitivity criterion for functional changes, used in subsequent trend assessment to mark rapid deterioration, stable fluctuations, or potential recovery stages. Furthermore, to further improve the stage-specificity of the analysis and identify the correspondence between chemotherapy dose and functional inflection points, this method introduces a retrospective segmentation strategy in the longitudinal analysis. Based on chemotherapy cycles, the overall treatment process is segmented according to drug formulation, dosage level, and treatment goals, for example, using the pre-chemotherapy induction stage, the intermediate maintenance stage, and the final consolidation stage as units. The longitudinal time axis is divided into multiple functional evolution sub-intervals. Then, within each sub-interval, the local derivative change trend and curve slope change rate of the stress response elastic coefficient are calculated, and a change trajectory diagram within the stage is established. The derivative slope is compared across sub-intervals. When the system detects a significant trend reversal between a certain segment, i.e., a change in slope direction or a sudden change in rate of change, the point is marked as a dose response reversal point, indicating that the point is a potential chemotherapy-induced functional reversal period, which may correspond to the cumulative effect of chemotherapy dose, drug toxicity threshold, or imbalance node of myocardial compensation mechanism. The identification result will be input into the trend warning engine to assist doctors in assessing whether it is necessary to adjust the subsequent chemotherapy regimen or implement cardioprotective intervention measures.

[0033] To improve the physiological consistency and time-domain accuracy of the quantitative analysis of myocardial stress state, the method sets key time nodes in the data acquisition link, i.e. 48 hours before and after the first, third and last chemotherapy as a unified signal acquisition window, and introduces multi-dimensional context physiological variables as reference baseline signals in this time period, including respiratory status parameters and sympathetic activation scores. Respiratory status is composed of thoracic respiratory motion amplitude and respiratory rate, which can be derived through chest belt pressure sensor or lung volume change image. Sympathetic activation score is generated based on heart rate variability (HRV) parameters, electrodermal activity (EDA) and systolic blood pressure variation (SBV) joint modeling. This score comprehensively reflects the patient's nervous excitement state at the time of acquisition, and is used to assess whether there is a risk of interference of underlying myocardial tension or excessive sympathetic activation on stress signals. All stress signal results will be weighted and calibrated according to the context score to eliminate non-pathological fluctuation factors; At the same time, in order to ensure that the time series data has high phase consistency between different patients and different cycles, an ECG trigger labeling mechanism is used to control the three-dimensional myocardial region scanning time point. Through real-time R-wave recognition, image acquisition is limited to the same cardiac cycle phase (such as early systole or end diastole), and the image frame is phase-locked, so that the sampling of all strain, lag, delay and other parameters has a consistent time starting point, so that the same phase corresponding analysis can be carried out when comparing across cycles, eliminating the phase drift interference caused by heart rhythm differences; In terms of myocardial stress starting point identification, the average length change rate of the lag chain is introduced as the main criterion for inflection point identification. Specifically, in consecutive cardiac cycles, the length of the intercellular lag chain formed by the outward conduction path of the stress starting region is calculated and averaged, and then the change rate of the average length in consecutive cycles is observed on the longitudinal time axis. When the change rate significantly jumps or drops in a certain period, it indicates that the stress propagation ability of the region has changed abruptly, suggesting the starting inflection point of the stress response after chemotherapy intervention. Especially when the change rate reaches a set discrimination threshold (such as more than 200% of the increase amplitude of the previous cycle), the system marks this time point as a functional trend turning point. This judgment can be combined with indicators such as elastic coefficient change and load response change to form a multi-factor early warning algorithm, ultimately achieving sensitive capture and prediction of chemotherapy-induced myocardial function deterioration.

[0034] Example 1: In combination with the accompanying Figure 4In this embodiment, the patient Ms. Wang, 45 years old, was diagnosed as HER2-positive early breast cancer, received AC-T chemotherapy regimen containing anthracycline (doxorubicin + cyclophosphamide → docetaxel), a total of six cycles, cycle interval is three weeks, the clinical doctor's order requires myocardial function monitoring 48 hours before the first, third and last chemotherapy, for the assessment of potential cardiac toxicity risk. In each specified acquisition window of Ms. Wang, the acquisition of electrocardiogram recording and tissue velocity imaging (TVI) was implemented synchronously, and the 6th segment (left ventricular middle anterior wall) was selected as the target monitoring area in the standard left ventricular AHA partition model. During acquisition, a three-channel high-density skin-attached electrode array was used to record the patient's local myocardial electrical depolarization waveform, and at the same time, the mechanical strain rate of the same segment was recorded using the TVI function of cardiac ultrasound, with a frame rate of 120Hz, and the sampling duration was about 8 seconds, covering at least 10 complete cardiac cycles. Taking the QRS complex starting point in the electrocardiogram signal as the reference, the earliest negative slope or strain acceleration inflection point in the strain rate waveform was extracted as the initial response point, forming a beat level data pair. Due to the waveform drift or abnormal morphology of myocardial electrical activity and mechanical response after chemotherapy, in order to avoid errors caused by the use of fixed reference points, a morphology adaptive strategy was introduced, searching for the slope extremum point or maximum derivative point in each beat cycle as the individualized reference site, so as to accurately align different cycle signals. For example, in the first acquisition before chemotherapy of Ms. Wang, the QRS starting point time of the electrical signal was 5.40 seconds, and the strain rate peak time was 5.56 seconds, and the initial calculation of the electromechanical coupling delay was 0.16 seconds; in the 48 hours before the third chemotherapy, due to the appearance of local T wave fusion, the identification of QRS starting point was difficult, the algorithm of the present invention took the maximum negative derivative point (5.42 seconds) of the electrocardiogram signal as the starting point, and the strain signal was delayed to 5.61 seconds due to the decrease of local tissue compliance induced by chemotherapy, and the calculated delay parameter was 0.19 seconds; while in the last chemotherapy, the electrical signal appeared irregular rhythm fluctuation, and the strain response signal also showed obvious attenuation, the identification of QRS phase starting point was 6.03 seconds, the strain response acceleration starting point was 6.22 seconds, and the delay was as high as 0.19 seconds, and the subsequent comparative analysis showed that Ms. Wang showed preliminary signs of electromechanical decoupling in this area under stress load. Finally, based on the delay data obtained in each cycle, the system can generate the electromechanical coupling trend curve of the myocardial segment and compare it with the reference threshold, to identify the risk area of functional early damage in advance, and provide high-quality basic index support for subsequent load response test and trend warning module.

[0035] After completing the electromechanical response signal acquisition and delay parameter extraction before and after the third chemotherapy, the next stage is entered: identification of stress initiation area and multi-level micro-load intervention analysis process, to capture potential functional abnormal areas and quantitatively evaluate the trend of their stress response characteristics.

[0036] The system first performs a sliding calculation analysis of the electrical-mechanical coupling delay in the left ventricular anterior wall (segment 6) of Ms. Wang within 48 hours before her third chemotherapy, and finds that the delay values of three consecutive cycles are 0.19s, 0.21s and 0.23s, all of which are more than 30% higher than the baseline value (0.14s) before chemotherapy, exceeding the set functional deviation threshold (defined as baseline ± 2σ), so the system preliminarily marks this area as a potential stress initiation area. Subsequently, according to the myocardial anatomical grid model, the system expands from segment 6 to adjacent segments 5 (basal anterior wall) and 7 (apical septal wall), and forms a one-dimensional delay chain by analyzing the intercellular electrical-mechanical delay conduction path: the delay in the conduction direction of segment 6→segment 5 is 0.11s, and the reverse is 0.07s; the bidirectional delay of segment 6→segment 7 is 0.14s and 0.09s respectively, and the system calculates the asymmetric delay ratio of the delay chain as , which indicates that there is a significant directional imbalance, further supporting that segment 6 is a potential stress propagation source.

[0037] Entering the micro-load intervention phase, Ms. Wang is subjected to a phased increase in respiratory synchronous load under resting conditions, with load levels divided into L1 (light), L2 (moderate), and L3 (medium-high) three levels, each level lasting for 15 seconds with a 10-second recovery interval. Under each level of load, the system records the strain rate response of segment 6 and its adjacent segments (5 and 7). In the L1 stage, the strain peak value appears within 2 seconds after the load and the difference between regions is not large (average strain rate difference <6%), while in the L2 stage, segment 6 appears a response lag (peak delay to 3.2 seconds) and the strain rate decreases by 15%, and in the L3 stage, the difference further expands, with a strain response delay of 4.5 seconds, and the maximum strain rate variation difference between segment 6 and segment 5 reaches 23%, so the system identifies that the stress threshold shift point occurs between L2 and L3, and marks segment 6 as a load desynchronization area.

[0038] Next, the system performs in-depth analysis of the strain response curve after load response, extracting peak lag (the time delay between load application and peak strain), acceleration initiation offset (the difference between the initial rise point of the strain curve and the expected response time), and recovery trend covariance (the dynamic consistency between the curve recovery segment and the resting baseline). Taking segment 6 under L2 load as an example, the peak lag is 3.2s, the acceleration offset is 0.8s, and the covariance decreases by 38%, reflecting significant mechanical response sluggishness and recovery difficulty in this region. In the elastic coefficient inversion analysis, the system constructs a two-dimensional response-recovery feature map, using the strain response slope at each load level (e.g., 0.45 for segment 6 at L2) as the X-axis and the recovery slope (-0.15 at L2) as the Y-axis to plot the evolution trajectory. It is found that from L1 to L3, there is a clear trend of shifting to the lower left quadrant, forming a break at L3, i.e., entering the uncompensated interval, indicating that the myocardium has lost its effective recovery ability under load. Through this two-dimensional mapping, the system ultimately generated a stress response elastic coefficient index of 0.61 for this segment, which is lower than the individual safety threshold (0.75), and it was determined to be a functional degradation risk zone.

[0039] In summary, by identifying the initial stress point from electromechanical delay, establishing intercellular conduction chains, implementing phased incremental load, extracting load response characteristics, and performing elastic inversion, we successfully identified potential cardiotoxicity-related local functional deterioration areas, providing quantitative support and data basis for subsequent treatment strategy adjustments before the 4th cycle (such as delaying the continued use of doxorubicin and increasing β-blocker intervention).

[0040] Example 2: Combined with appendix Figure 5 Patient Ms. Wang received a multi-cycle chemotherapy regimen (AC-T regimen, a total of 6 cycles), and quantitative assessments based on myocardial stress were performed before and after each cycle. 48 hours before the first chemotherapy session (referred to as...) ), 48 hours before and after the third cycle (denoted as ), and the 48 hours before and after the 6th cycle (the last one) (referred to as Acquire electromechanical coupling signals; data includes: Electromechanical delay data (EMD) (unit: ms): The delay between the QRS initiation and the peak local strain rate of the myocardium in the acquired pulsating signal, for example, in The time is [34, 35, 33] ms. The time was [38, 42, 45] ms. The time reached [52,55,58]ms; Intercellular hysteresis chain length (ICDL) (unit: μm / ms): The intercellular signal propagation path is plotted outward from the initial stress region, and the propagation velocity change at each node is recorded to calculate the hysteresis chain derivative. , for example, 0.45, 0.39, 0.28, respectively; Load response strain perturbation (unit %): The amplitude of the strain response recorded by the micro-load excitation method is 3.2%, 2.6%, and 1.4% at three time points, respectively; Decay factor setting To ensure that the data within the last 48 hours has a higher weight.

[0041] The following main function is introduced to map the stress trend: Where is the range of electromechanical delay distribution at each time, for example, the 3rd cycle is , and the last cycle is .

[0042] Substitute: Take as an example, take the representative average value: ; ; ; The formula kernel value is calculated as: Set the integral from 48 hours ago ( ) to the current time hours, and bring in the exponential decay kernel function, which is: Substitute the variable , which is: Since , then: Trend change explanation: If , , , respectively, we get: ; ; ; Then the first derivative is: From : The rate of decline is From : The rate of decline accelerates to Combining the function slope trend continues to decline and the derivative is negative, confirm Ms. Wang in the third cycle after the emergence of myocardial dysfunction compensatory decline, should be considered before the fourth cycle to reduce drug dosage, or early introduction of heart protection strategy, such as the use of cardioprotective agent Dexrazoxane.

[0043] Before and after the third cycle of chemotherapy, patient Ms. Wang completed the electromechanical delay (EMD), intercellular delay chain (ICDL) and load response (LRD) data collection, entered the data fusion and evaluation stage. First, to ensure that the three types of heterogeneous signals have consistency at the level of myocardial anatomical region, the system uses MRI modeling to obtain Ms. Wang's left ventricular standard 17 segment atlas, and implements spatial mapping in TTE ultrasound and electrical signal diagram to ensure that each electromechanical delay data point can match the conduction path node of the intercellular delay chain in the same segment, thereby constructing the spatially registered joint feature space. For example, the EMD delay high point existing in segment 6 (middle segment of the anterior interventricular wall) corresponds to the initial response cell, which is completely coincident with the initial node of the ICDL chain in geometry, eliminating the interference of anatomical drift on data comparison.

[0044] Into the cross-modal time comparison analysis, the system unifies the sampling frequency of different data streams, and standardizes each beat cycle. The response derivative rate of each modality signal is calculated by time axis gradient comparison algorithm. Taking Ms. Wang's third cycle data as an example, the EMD signal of segment 6 appears a maximum derivative mutation within 98 ms after the QRS trigger of electrocardiogram, while the conduction velocity in the same region of the ICDL signal jumps in rate at the 110th ms, indicating that there is an 11 ms response shift between the two modalities. This time derivative jump is identified by the system as a potential abnormal response acceleration, which is a key criterion for judging early stress signs. In further comparison, it is found that segment 6 has a broken rising edge in the EMD signal waveform in multiple consecutive heartbeats, forming an abnormal step, i.e. the stress waveform appears a morphological fault. This fault is manifested as a termination or reverse propagation of the conduction path in ICDL, and the system marks this region as a stress non-continuous area, inferring that a myocardial microstructural barrier or functional activation disconnection is formed at this place.

[0045] More notably, during the collection of Ms. Wang's fifth cycle, the system identified that part of the cell group in segment 8 (left ventricular middle and posterior interventricular wall) appeared strain response signals earlier than segment 6, although it was spatially far from the initial stress area, but it triggered ahead of time, about 25 ms earlier than the stress response time. After spatial comparison, this phenomenon was excluded as an artifact, and it was inferred that there was a distal ectopic activation or cross-segment short circuit propagation. This kind of abnormal unit that responds first away from the starting point is classified as a potential heterogenous stress node, which has a hinting value for subsequent rhythm disorder.

[0046] In the longitudinal trend analysis phase, the system pairs the stress response elasticity coefficient (SREC) of Ms. Wang in all assessment periods on the time axis, and takes the first derivative change as the sensitive criterion for functional change. For example, from T1 to T3, SREC decreases from 0.92 to 0.66, the derivative is ; from T3 to , it decreases to 0.49, the derivative slows down to , although the trend has not reversed, the decline rate has decreased. The system introduces a backtracking segmentation strategy, reviews the historical periods of Ms. Wang by treatment stage (induction period, consolidation period, maintenance period), and draws multiple trend lines, finally confirms T3 as the trend turning point of dose response, and accordingly suggests considering entering the functional protection strategy window in the subsequent course.

[0047] In terms of context calibration, in the data collection 48 hours before and after each key time point, the respiratory status parameters (respiratory rate, tidal volume) and sympathetic activation scores (HRV standard deviation, LF / HF ratio, skin electrical activity peak frequency) are included to construct the physiological background index. Ms. Wang's score increased to 87 before T3, indicating a state of emotional tension; this score is then used for dynamic weighted comparison results to eliminate the interference of non-pathological fluctuations on stress recognition. In addition, all myocardial signal collections of the system are triggered by ECG R waves, so that three-dimensional myocardial image collection is completed under the same cardiac phase, ensuring that different modal data have phase-aligned physical basis within the cardiac cycle.

[0048] Finally, the system constructs a chain length curve for segment 6 of Ms. Wang in combination with the time derivative of the average length change of the lag chain in multiple consecutive periods, and finds that the average growth rate from T1 to T3 is , and after , it suddenly drops to , the system identifies this mutation point as the stress inflection point, and suggests starting supportive heart treatment combined with blocking agents.

[0049] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for quantitative assessment of myocardial stress in breast cancer chemotherapy, characterized by The application relates to a method for evaluating the stress level of myocardial tissue after breast cancer chemotherapy. The method comprises the following steps: acquiring an electromechanical response signal of myocardial tissue of a chemotherapy patient in a beating cycle, extracting an electromechanical coupling delay parameter of local tissue, and identifying a starting area of a chemotherapy-induced myocardial stress reaction; Collecting a response time difference between cell membranes in the identified stress starting area and adjacent myocardial tissue, and calculating a stress conduction lag difference index to represent the asymmetric stress conduction characteristics induced by chemotherapy; Implementing controllable micro-load intervention on the patient, and monitoring the strain response curve of the stress starting area and the surrounding tissue after the load is applied; Based on the response lag time, strain peak value change rate and dynamic trend of recovery to the baseline in the load response process, the stress response elastic coefficient of the myocardial tissue is obtained by inversion, which is used for quantifying the stress level of the myocardial tissue after breast cancer chemotherapy.

2. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 1, characterized in that The acquisition of the electromechanical response signal comprises simultaneously recording local depolarization signals and synchronous mechanical strain rate signals of the myocardial tissue to form a beating level data pair for synchronous comparative analysis; and the electromechanical coupling delay parameter is based on the adaptive adjustment of a reference point according to the signal waveform form in the beating cycle to adapt to the alignment error caused by the ECG or strain form variation after chemotherapy.

3. The method of quantitative assessment of cardiomyocyte stress in breast cancer chemotherapy according to claim 2, characterized in that The identification of the stress starting area comprises that the electromechanical coupling delay deviates from a basic threshold in three continuous heartbeats, and the area has boundary stability in space; a one-dimensional lag chain is established based on the radial expansion of the starting area center to the intercellular conduction path of the adjacent area, and is used for marking the conduction directionality; and the stress conduction lag difference index comprises a bidirectional conduction delay difference ratio between local adjacent myocardial units, so as to reveal the potential asymmetric diffusion trend.

4. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 3, characterized in that The micro-load intervention is a phased incremental load mode, and the disturbance is applied in multiple incremental levels to obtain a stress threshold transfer point of the strain response; and the response acquisition of the load intervention comprises a strain rate variation difference in a time window in the starting area and the peripheral area, which is used for evaluating the consistency of the load response.

5. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 4, characterized in that The strain response curve after the load response is comprehensively analyzed by calculating the peak lag, acceleration starting point offset and recovery trend covariance; The inversion of the stress response elastic coefficient comprises two-dimensional mapping of the response slope and the recovery speed curve under the condition of the incremental load, and identification of a non-compensation interval.

6. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 5, characterized in that The electromechanical delay data, the intercellular lag chain data and the load response data simultaneously enter a joint feature space for multidimensional dynamic comparison; and the stress response elastic coefficient generated by each evaluation is longitudinally time-paired with historical records for judging the myocardial function change trend; The time nodes for collecting the electromechanical response signal are selected within 48 hours before and after the first, third and last chemotherapy in a chemotherapy cycle, so as to capture the stress response starting inflection point.

7. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 6, characterized in that The construction of the joint feature space adopts a spatial alignment strategy of myocardial units in each data source, so that the anatomical areas between the electromechanical delay data and the intercellular lag chain data are consistent; and the comparison comprises a time gradient alignment process across the data sources, and detects myocardial abnormal response acceleration as an early stress signal based on the time derivative change rate.

8. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 7, characterized in that If a response waveform form fault mutation appears in the comparison result, the region is marked as a stress non-continuous region, which is used to infer the myocardial inter-conduction barrier; in the comparison process, the stress unit far from the starting point in the data points is identified to respond before the adjacent region.

9. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 8, characterized in that In the longitudinal time pairing analysis, the first derivative of the stress response elastic coefficient change amplitude of each historical cycle of the patient is used as a functional change sensitivity criterion; wherein the longitudinal time pairing analysis introduces a backtracking segmentation strategy to divide the historical data according to the chemotherapy stage to identify the dose turning point.

10. The method of quantitative assessment of myocardial stress in chemotherapy of breast cancer according to claim 9, characterized in that The 48-hour data segment before and after the time node includes the respiratory state and the sympathetic activation score as context reference signals, which are used to calibrate the myocardial stress baseline state; the time node data acquisition includes three-dimensional myocardial region directional scanning marked by ECG trigger, so that each cycle is collected at the same cardiac phase; The judgment of the stress response starting inflection point takes the mutation of the average length change rate of the lag chain as the main discrimination criterion.

Images