Pulmonary edema risk prediction method and system based on electromagnetic field biological detection

Abstract

The invention discloses a pulmonary edema risk prediction method and system based on electromagnetic field biological detection, and the method comprises the steps: obtaining a disturbance coefficient in a preset rotation detection mode through six electrode slices disposed on the chest of a to-be-detected user, inputting the disturbance coefficient into a pre-trained pulmonary water volume prediction model, and carrying out the prediction of the pulmonary water volume. Predicting to obtain a predicted value of the lung water volume of the to-be-detected user; matching a corresponding preset chest circumference segment based on chest circumference data of the to-be-detected user, and comparing the predicted value of the lung water volume with a preset threshold value corresponding to the preset chest circumference segment to obtain a prediction result; or calculating the pulmonary impedance spectrum of the to-be-detected user based on the detected disturbance coefficient; matching a corresponding preset chest circumference segment based on the chest circumference data of the to-be-detected user, and calculating Similarity between the pulmonary impedance spectrum of the to-be-detected user and a standard pulmonary impedance spectrum corresponding to the preset chest circumference segment based on a cosine similarity theory; and then prediction is carried out based on Similarity.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and particularly to a method and system for predicting the risk of pulmonary edema based on electromagnetic field biological detection. Background Art

[0002] Heart failure (HF) is a severe manifestation or advanced stage of various heart diseases. HF has a high prevalence, readmission rate, mortality rate, and treatment cost globally, especially in China, imposing a heavy burden on the medical system. According to the latest data, there are approximately 64.3 million HF patients worldwide. Cardiogenic pulmonary edema is a common and severe clinical manifestation of HF, and its prevalence has been increasing year by year in recent years. Its in-hospital mortality rate is as high as 10% - 20%. Cardiogenic pulmonary edema will exacerbate the condition of HF, and due to the accumulation of fluid in the alveoli, it seriously affects the patient's respiratory function, leading to acute dyspnea and hypoxemia, rapidly progressing to acute respiratory failure, and even sudden cardiac death. Relevant research indicates that the mortality rate of HF patients with pulmonary edema is more than twice that of HF patients without pulmonary edema (48% VS 21%).

[0003] Hypervolemic overload is an important pathophysiological process in the development of HF. Detecting symptoms of pulmonary congestion is an important means for the volume management of HF. Therefore, when facing cardiogenic pulmonary edema, it is necessary to relieve and improve symptoms as soon as possible, stabilize the hemodynamic state, maintain the function of important organs, avoid recurrence, and improve the prognosis. Early identification and diagnosis, real-time dynamic monitoring, and intelligent early warning and assessment of the pulmonary edema situation are the key links in the diagnosis and treatment of HF or cardiogenic pulmonary edema. However, there is still a lack of accurate, dynamic, and economical monitoring technologies. Currently, the diagnosis and monitoring of cardiogenic pulmonary edema mainly rely on the determination of plasma brain natriuretic peptide levels, chest imaging examinations such as X-ray, CT, MRI, bedside cardiopulmonary ultrasound, pulmonary capillary wedge pressure (PCWP), non-invasive hemodynamic monitoring devices such as pulse index continuous cardiac output measurement (PICCO), etc. These methods have limitations, such as time-consuming and laborious, invasive, risk of radiation exposure, uneven quality of doctor's film reading, lack of real-time and dynamic monitoring capabilities, poor specificity or sensitivity, complex operation, no parameters specifically for pulmonary water, no big data analysis and artificial intelligence, resulting in the inability to take effective treatment measures in a timely manner and aggravating the development of HF.

[0004] In the prior art, it has been proposed to monitor the pulmonary fluid by measuring the resistance across the lung. The more fluid present in the lung, the lower the impedance, and an implantable medical device, such as a cardiac pacemaker, has been proposed. Generally, the resistance measurement is performed between a right ventricular electrode connected to the implantable medical device via a wire and another electrode located at the implantable medical device itself. Such impedance measurements sample the thoracic tissue, including the lung.

[0005] For example, the US patent application of US10303305 discloses a method for measuring lung impedance using an implantable medical device. The resistance measurement is carried out between an epicardial electrode placed on the left ventricular wall of the heart and connected to the implantable medical device, and another electrode located at the implantable medical device. By applying a relatively large stimulating current to the built-in electrode and measuring the generated voltage with the built-in electrode, and then calculating the ratio of the voltage and the current, the impedance is measured.

[0006] For another example, the Chinese invention patent application of CN101163443A discloses a pathological evaluation method for impedance measurement by using a converging bioelectric wire field. It implants a first electrode and a second electrode in vivo, then injects a current to generate a first electric wire field and a second electric wire field, and then calculates the impedance value based on the potential difference and the injected current, and further evaluates the pathological condition based on the impedance value. However, the above-mentioned prior arts all implant electrodes in the lungs or the lesions to be measured, that is, they adopt an invasive detection method.

[0007] In view of this, the present invention provides a non-invasive method for predicting the risk of pulmonary edema. Summary of the Invention

[0008] The purpose of the present invention is to provide a method and system for predicting the risk of pulmonary edema based on electromagnetic field biological detection, which partially solves or alleviates the above deficiencies in the prior art and can achieve non-invasive prediction of the risk of pulmonary edema.

[0009] To solve the above-mentioned technical problems, the present invention specifically adopts the following technical solutions: In the first aspect of the present invention, it is to provide a method for predicting the risk of pulmonary edema based on electromagnetic field biological detection, which includes: Six electrode patches are arranged at six specified positions on the chest of the user to be detected, and the perturbation coefficient in the preset rotation detection mode is obtained; the six preset positions are respectively: the first electrode is located at the 3rd and 4th intercostal spaces on the right midclavicular line in front of the chest, the second and third electrodes are arranged side by side between the 6th and 7th ribs on the right axillary midline on the side of the chest, the fourth and fifth electrodes are arranged side by side between the 6th and 7th ribs on the left axillary midline on the side of the chest, and the sixth electrode is located at the 3rd and 4th intercostal spaces on the left midclavicular line in front of the chest; the perturbation coefficient is the mean value of the perturbation coefficients on the propagation paths from all electrodes as the transmitting end to each electrode as the receiving end; The disturbance coefficient detected in the rotation detection mode is input into a pre-trained lung water volume prediction model to predict the lung water volume prediction value of the user to be detected; and based on the pre-acquired chest circumference data of the user to be detected, the chest circumference data is matched to the corresponding preset chest circumference segment, and the lung water volume prediction value is compared with the preset threshold value corresponding to the preset chest circumference segment. If the lung water volume prediction value is less than the first preset threshold value, the prediction result is no risk; if the lung water volume prediction value is greater than the second preset threshold value, the prediction result is high risk; if the lung water volume prediction value is ≥ the first preset threshold value and < the third preset threshold value, the prediction result is low risk; if the lung water volume prediction value is ≥ the third preset threshold value and < the second preset threshold value, the prediction result is medium risk; Alternatively, based on the disturbance coefficient detected in the rotation detection mode, the lung impedance spectrum of the user to be detected is calculated. ; and based on the pre-acquired chest circumference data of the user to be detected, matching it to the corresponding preset chest circumference segment, calculating the lung impedance spectrum Z of the user to be detected based on the cosine similarity theory b The standard lung impedance spectrum corresponding to the preset chest circumference segment calculated in advance The cosine similarity Similarity between them; then based on the cosine similarity Similarity, the prediction result of the occurrence of pulmonary edema in the user to be detected is predicted; if the cosine similarity Similarity is 1, the prediction result is no risk; if the cosine similarity Similarity is -1, the prediction result is high risk; if the cosine similarity Similarity ≥ 0, and < 1, the prediction result is low risk; if the cosine similarity Similarity < 0, and > -1, the prediction result is medium risk; in, , for The i-th frequency component of for The i-th frequency component of , n is the total number of frequency components.

[0010] In some embodiments, the preset rotation detection mode refers to completing detection in multiple detection modes within a preset cycle time, and applying an excitation signal in the frequency band of 10KHz-100KHz in each detection mode in turn; wherein the multiple detection modes specifically include: The first detection mode: any two electrodes among the six electrodes are used as transmitting ends, any two electrodes among the remaining four electrodes are used as receiving ends, and the other two electrodes are used as grounding ends; and / or, Second detection mode: any two electrodes among the six electrodes are used as transmitting ends and receiving ends at the same time, and the remaining four electrodes are used as ground ends; and / or, Third detection mode: Any two of the six electrodes are simultaneously used as the transmitting end and the receiving end, any two of the remaining four electrodes are grounded, and the other two electrodes are floating.

[0011] In some embodiments, the step of calculating the standard lung impedance spectrum corresponding to the preset chest circumference segment specifically includes the steps of: Setting at least three chest circumference segments based on a preset maximum chest circumference threshold and a minimum chest circumference threshold; Screening the corresponding healthy subject groups based on each chest circumference segment to obtain at least three healthy subject groups; Obtaining the set of healthy perturbation coefficients of all healthy subjects in the preset rotation detection mode, and calculating the lung tissue impedance spectrum of each healthy subject based on the respective healthy perturbation coefficients of each healthy subject; Using the least squares method to fit the lung tissue impedance spectra of all healthy subjects in each group respectively to obtain the standard lung impedance spectrum of each group .

[0012] In some embodiments, the step of training the lung water volume prediction model specifically includes: Constructing a training data set: Setting at least three chest circumference segments based on a preset maximum chest circumference threshold and a minimum chest circumference threshold; Obtaining the training set of healthy perturbation coefficients of healthy subjects with different chest circumferences in each chest circumference segment in the preset rotation detection mode; Obtaining the training set of abnormal perturbation coefficients of pulmonary edema subjects with different chest circumferences and different lung water volumes in each chest circumference segment in the preset rotation detection mode; Model training: Using a machine learning algorithm to learn the training set of healthy perturbation coefficients and the data set of abnormal perturbation coefficients to obtain the lung water volume prediction model.

[0013] In some embodiments, when obtaining the training set of abnormal perturbation coefficients of pulmonary edema subjects with different chest circumferences and different lung water volumes in each chest circumference segment in the preset rotation detection mode, for the same pulmonary edema subject, respectively obtaining the first abnormal perturbation coefficient and the second abnormal perturbation coefficient of the pulmonary edema subject in the sitting position and the lying position; And taking the mean value of the first abnormal perturbation coefficient in the sitting position and the second abnormal perturbation coefficient corresponding to the lying position under the same excitation signal as the abnormal perturbation coefficient of the pulmonary edema subject to obtain the training set of abnormal perturbation coefficients corresponding to each chest circumference segment.

[0014] In a second aspect of the present invention, there is provided a pulmonary edema risk prediction system based on electromagnetic field biological detection, which includes: Six electrode patches are used to construct an excitation signal propagation path between six specified positions on the chest of the user to be detected, and obtain the perturbation coefficient of each propagation path in a preset rotation detection mode; the preset six positions are respectively: the first electrode is located at the 3rd and 4th intercostal spaces on the right midclavicular line in front of the user's chest, the second and third electrodes are arranged side by side between the 6th and 7th ribs on the right midaxillary line on the side of the user's chest, the fourth and fifth electrodes are arranged side by side between the 6th and 7th ribs on the left midaxillary line on the side of the user's chest, and the sixth electrode is located at the 3rd and 4th intercostal spaces on the left midclavicular line in front of the user's chest; A data processing module is used to calculate the mean value of the perturbation coefficients on all propagation paths in the preset rotation detection mode as the perturbation coefficient of the user to be detected; specifically, a perturbation coefficient will be generated for each propagation path from the electrode as the transmitter to each electrode as the receiver; An excitation generator is used to send electrical stimulation signals to any two of the six electrode patches respectively based on the preset rotation detection mode; A first prediction module is used to input the perturbation coefficient detected in the rotation detection mode into a pre-trained pulmonary water volume prediction model to predict the predicted value of the pulmonary water volume of the user to be detected; and match the corresponding preset chest circumference segment based on the chest circumference data of the user to be detected, calculate and compare the predicted value of the pulmonary water volume with the preset threshold corresponding to the preset chest circumference segment. If the predicted value of the pulmonary water volume is less than the first preset threshold, the prediction result is no risk; if the predicted value of the pulmonary water volume is greater than the second preset threshold, the prediction result is high risk; if the predicted value of the pulmonary water volume ≥ the first preset threshold and < the third preset threshold, the prediction result is low risk; if the predicted value of the pulmonary water volume ≥ the third preset threshold and < the second preset threshold, the prediction result is medium risk; or, A second prediction module is used to calculate the pulmonary impedance spectrum of the user to be detected based on the perturbation coefficient detected in the rotation detection mode ; and match the corresponding preset chest circumference segment based on the chest circumference data of the user to be detected, and then calculate the pulmonary impedance spectrum of the user to be detected based on the cosine similarity theory and the standard pulmonary impedance spectrum of the preset chest circumference segment calculated in advance to obtain the cosine similarity Similarity therebetween. Finally, based on the cosine similarity Similarity, predict the prediction result of the user to be detected having pulmonary edema; if the cosine similarity Similarity is 1, the prediction result is no risk; if the cosine similarity Similarity is -1, the prediction result is high risk; if the cosine similarity Similarity ≥ 0 and < 1, the prediction result is low risk; if the cosine similarity Similarity < 0 and > -1, the prediction result is medium risk; Among them, , is the i-th frequency component of is the i-th frequency component of , and n is the total number of frequency components. In some embodiments, the above-mentioned preset rotation detection mode means that detections under multiple detection modes are completed within a preset cycle time, and excitation signals in the frequency band of 10 KHz - 100 KHz are sequentially applied under each detection mode; wherein, the multiple detection modes specifically include: First detection mode: Any two of the six electrodes are used as the transmitting end, any two of the remaining four electrodes are used as the receiving end, and the other two electrodes are used as the grounding end; and / or, Second detection mode: Any two of the six electrodes are simultaneously used as the transmitting end and the receiving end, and the remaining four electrodes are used as the grounding end; and / or, Third detection mode: Any two of the six electrodes are simultaneously used as the transmitting end and the receiving end, any two of the remaining four electrodes are grounded, and the other two electrodes are floating.

[0015] In some embodiments, the second prediction module specifically includes: A first calculation unit, configured to calculate the lung tissue impedance spectrum of each healthy subject in each group based on the preset healthy perturbation coefficient sets of each group of healthy subjects detected under the rotation detection mode; A second calculation unit, configured to respectively fit the lung tissue impedance spectra of each healthy subject in each group by using the least squares method to obtain the standard lung impedance spectra of each group .

[0016] In some embodiments, the pulmonary edema risk prediction system based on electromagnetic field biological detection further includes: A model training module, configured to respectively obtain the abnormal perturbation coefficient sets of multiple pulmonary edema subjects in the sitting position and the lying position and under the preset rotation detection mode to construct a training sample library, and use a machine learning algorithm to learn the abnormal perturbation coefficient sets in the training sample library to obtain the lung water volume prediction model.

[0017] In some embodiments, the pulmonary edema risk prediction system based on electromagnetic field biological detection further includes a data preprocessing module, configured to perform data preprocessing on the electrical signals obtained through six electrode patches, and the data preprocessing includes data cleaning, denoising, and normalization processing.

[0018] Beneficial effects: The method and system of the present invention can achieve non-invasive (or non-intrusive), real-time, and continuous (e.g., 24 hours a day) monitoring and prediction of pulmonary water volume. Compared with traditional auxiliary screening methods such as ultrasound, CT, and X-ray, it has lower costs and is more convenient. In particular, CT and X-ray cannot achieve real-time monitoring. Usually, they can only be monitored once a day, that is, continuous monitoring cannot be achieved. Moreover, the method and system of the present invention avoid the risk of patients receiving ionizing radiation and are particularly suitable for long-term systemic circulation volume management of heart failure patients. In addition, by using a non-invasive detection method, monitoring can be carried out at the bedside at any time using adhesive electrode patches, avoiding the infection risk caused by invasive or intrusive detection, as well as the possibility of triggering a systemic inflammatory response, reducing the pain and medical risks of patients.

[0019] The method and system of the present invention can timely capture the dynamic changes of pulmonary edema, providing more timely and accurate data support for doctors to make diagnoses and treatments, thereby adjusting treatment plans and avoiding delays in the condition due to untimely examinations.

[0020] The method and system of the present invention adopt a six-electrode rotating measurement structure ( Figure 2 the electrode arrangement structure shown in), breaking the limitations of a single pair of electrodes of ReDS Pro. The advantage of this is to achieve multi-dimensional signal acquisition and effectively reduce signal offset caused by the anisotropy of thoracic tissues.

[0021] The method and system of the present invention support supine pulmonary water monitoring. Compared with the sitting position, the lung compliance and lung resistance of heart failure patients will change, and the lung volume will also change, resulting in differences in the collected data. Therefore, it is very suitable for real-time and continuous monitoring of supine pulmonary edema in heart failure patients. Moreover, through the design of the above-mentioned rotating measurement structure, the deviation between the pulmonary water volume measured in the supine position and the pulmonary water volume in the forward sitting position is reduced to a certain extent, making them almost the same, which enables real-time and continuous monitoring even for ICU intubated patients (who cannot maintain a sitting position), expanding the applicable population.

[0022] The system of the present invention has great advantages in cost control. The terminal selling price is compressed to 1 / 3 of that of competing products. The terminal selling price of ReDS Pro is about 300,000 yuan per unit, enabling even grass-roots hospitals with relatively low budgets to configure this system. Description of the Drawings

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to actual scale. Obviously, the drawings described below are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0024] Figure 1 It is a flowchart of an embodiment of a method for predicting pulmonary edema of the present invention; Figure 2 It is a schematic diagram of six electrodes located at six designated positions on the chest during detection; Figure 3 It is a flowchart of another embodiment of a method for predicting pulmonary edema of the present invention; Figure 4 It is the comparison result of the perturbation coefficient R obtained by using the method of the present invention to detect the experimental group (i.e., the group of pulmonary edema subjects) and the control group (i.e., the group of healthy subjects) respectively; Figure 5 It is the result obtained by statistically analyzing the perturbation coefficients obtained by monitoring 66 healthy subjects by using the method of the present invention and according to the monitoring results; Figure 6 It is the result obtained by analyzing the perturbation coefficients obtained by repeatedly monitoring some of the 66 healthy subjects twice; Figure 7 It is the result obtained by statistically analyzing the perturbation coefficients obtained by monitoring 16 pulmonary edema subjects by using the method of the present invention and according to the monitoring results; Figure 8 It is to compare the mean values of the perturbation coefficients of 66 healthy subjects and 16 pulmonary edema subjects.

[0025] Summary of reference numeral identification: The first electrode 0, the second electrode 1, the third electrode 3, the fourth electrode 4, the fifth electrode 3, the sixth electrode 5. Detailed implementation manners

[0026] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0027] In this document, suffixes such as "module", "component", or "unit" used to denote elements are only for the convenience of explaining the present invention and have no specific meaning in themselves. Therefore, "module", "component", or "unit" can be used interchangeably.

[0028] In this document, the orientation or positional relationship indicated by terms such as "upper", "lower", "inner", "outer", "front", "rear", "one end", "the other end", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation on the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and should not be construed as indicating or implying relative importance.

[0029] In this document, unless otherwise clearly defined and limited, terms such as "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, a direct connection, or an indirect connection through an intermediate medium, and can also be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0030] As used in this document, "and / or" includes any and all combinations of one or more of the listed related items.

[0031] As used in this document, "a plurality of" means two or more, that is, it includes two, three, four, five, etc.

[0032] As used in this specification, the term "about" typically represents + / - 5% of the value, more typically + / - 4% of the value, more typically + / - 3% of the value, more typically + / - 2% of the value, even more typically + / - 1% of the value, and even more typically + / - 0.5% of the value.

[0033] In this specification, certain embodiments may be disclosed in a format within a certain range. It should be understood that this kind of description "within a certain range" is only for convenience and brevity, and should not be construed as a rigid limitation on the disclosed range. Therefore, the description of the range should be considered as having specifically disclosed all possible sub-ranges and the individual numerical values within this range. For example, the description of the range 1 - 6 should be regarded as having specifically disclosed sub-ranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6, etc., and the individual numbers within this range, such as 1, 2, 3, 4, 5, and 6. The above rules apply regardless of the breadth of the range.

[0034] The prior art CN 115295139 A discloses an edema monitoring system based on bioelectrical impedance, which includes: a real-time control system, an upper computer operation system, and a cloud storage system; the real-time control system is communicatively connected to the upper computer operation system, and the upper computer operation system is communicatively connected to the cloud storage system; among them, the real-time control system is worn by the patient and is used to measure bioelectrical impedance and send it to the upper computer operation system; the upper computer operation system exists in the patient's mobile terminal and is used to extract bioelectrical impedance characteristics, display the bioelectrical impedance and its characteristics to the patient, and at the same time send it to the cloud storage system; the cloud storage system exists in the hospital information system and is used to store the bioelectrical impedance and its characteristics, and display it to the doctor. After the doctor makes an evaluation of the edema situation, the evaluation result is returned for the patient to view. This monitoring system uses two excitation electrodes and two receiving electrodes to calculate the bioimpedance only based on two voltages. However, for different body parts, their internal tissues and structures are different. Using the same method for all of them lacks pertinence, and the reliability of its detection results needs to be improved.

[0035] Embodiment 1: Refer to Figure 1 , which is a flowchart of a method for predicting (or evaluating) pulmonary water volume based on electromagnetic field biological detection according to the present invention. Specifically, the method includes the steps: S101 Obtain the perturbation coefficient in the preset rotation detection mode through six electrode patches set at six preset positions on the chest of the user to be detected.

[0036] Refer to Figure 2 , in some embodiments, the six preset positions are respectively: the first electrode is located at the 3rd and 4th intercostal spaces on the right midclavicular line in front of the chest of the user to be detected, the second and third electrodes are arranged side by side at the 6th and 7th ribs on the right axillary midline on the side of the chest of the user to be detected, the fourth and fifth electrodes are arranged side by side at the 6th and 7th ribs on the left axillary midline on the side of the chest of the user to be detected, and the sixth electrode is located at the 3rd and 4th intercostal spaces on the left midclavicular line in front of the chest of the user to be detected. Correspondingly, the above-mentioned perturbation coefficient is the mean value of the perturbation coefficients on the propagation paths from all electrodes as the transmitting end to each electrode as the receiving end.

[0037] Compared with the difference between the perturbation coefficients of the chest (i.e., the first electrode and the sixth electrode) and the perturbation coefficients of the side of the chest (i.e., the second to fifth electrodes), the difference between the perturbation coefficients of the chest and back (i.e., the mirror positions on the user's back corresponding to the 3rd and 4th intercostal spaces on the right midclavicular line in front of the chest and the mirror positions on the left midclavicular line in front of the chest) and the perturbation coefficients of the side of the chest is relatively large. Therefore, in order to reduce the error, a perturbation coefficient set at a specified position on the chest and a perturbation coefficient set at a specified position on the side of the chest are obtained.

[0038] For example, a BORN-BE non-invasive dynamic brain edema monitor is used to collect the perturbation coefficient data set.

[0039] In some embodiments, the above-mentioned preset rotation detection mode means that within a preset cycle time (e.g., 3 min - 5 min), detections in multiple detection modes are completed, and an excitation signal in the frequency band of 10 KHz - 100 KHz is sequentially applied under each detection mode; wherein, the multiple detection modes specifically include: The first detection mode: Any two of the six electrodes are used as the transmitting ends, any two of the remaining four electrodes are used as the receiving ends, and the other two electrodes are used as the grounding ends (correspondingly, four disturbance coefficients will be obtained); and / or, The second detection mode: Any two of the six electrodes are simultaneously used as the transmitting end and the receiving end, and the remaining four electrodes are used as the grounding ends (correspondingly, four disturbance coefficients will be obtained); and / or, The third detection mode: Any two of the six electrodes are simultaneously used as the transmitting end and the receiving end, any two of the remaining four electrodes are grounded, and the other two electrodes are suspended (correspondingly, four disturbance coefficients will be obtained). Preferably, the finally output disturbance coefficient is the average value of the above 12 disturbance coefficients.

[0040] Specifically, after the excitation signal generated by the excitation signal generator in the BORN-BE non-invasive cerebral edema dynamic monitor flows through the chest via the transmitting end electrodes, it flows into the receiving device in the BORN-BE non-invasive cerebral edema dynamic monitor from the receiving end electrodes, and corresponding disturbance coefficients are obtained through data analysis and processing. Specifically, the preset frequency band is 10 kHz to 100 kHz. Assuming the preset frequency adjustment value is 10 kHz, the excitation signal generator sequentially generates a reference electromagnetic wave signal and an adjusted electromagnetic wave signal at preset time intervals. Assuming the frequency of the reference electromagnetic wave signal is 10 kHz (corresponding to excitation signal I), the frequency of the first adjusted electromagnetic wave signal is 10 + 10 = 20 kHz (corresponding to excitation signal II), the frequency of the second adjusted electromagnetic wave signal is 20 + 10 = 30 kHz (corresponding to excitation signal III), and so on. That is, within the preset cycle time, detections under 10 excitation signals need to be completed. Preferably, under each excitation signal, the disturbance coefficients in the above three detection modes are obtained, thereby obtaining a set of disturbance coefficients.

[0041] Furthermore, in order to improve the reliability of the data, data within multiple cycle times are obtained to construct this data set.

[0042] Low-frequency signals are mainly sensitive to extracellular hydrates, while high-frequency signals can penetrate the cell membrane and enter the cell, that is, high-frequency signals are sensitive to both inside and outside the cell. The impedance of tissues at the same location in the same state is lower at high frequencies. Therefore, when pulmonary edema occurs (for example, pleural effusion, pericardial effusion), the electrical conductivity of the lungs increases and the bioelectrical impedance decreases. At this time, the decrease in the resistance of high-frequency signals is more obvious, so high-frequency signals are more sensitive to pulmonary edema. Therefore, preferably, in this embodiment, an excitation signal in the frequency band of 10KHz - 100KHz is used.

[0043] S102 inputs the perturbation coefficient detected in the rotation detection mode into a pre-trained pulmonary water volume prediction model to predict the pulmonary water volume prediction value of the user to be detected.

[0044] Since the perturbation coefficient is positively correlated with the chest circumference, it is necessary to segment the chest circumference in advance and obtain the pulmonary water volume and perturbation coefficient of healthy subjects and subjects with pulmonary edema with different chest circumferences in each segment for model training. Accordingly, the steps of training the pulmonary water volume prediction model specifically include: Construct a training data set: Set at least three chest circumference segments based on a preset maximum chest circumference threshold and a minimum chest circumference threshold; Obtain a healthy perturbation coefficient training set of healthy subjects with different chest circumferences (whose pulmonary water volume is a normal physiological experience value) in the preset rotation detection mode in each chest circumference segment; obtain an abnormal perturbation coefficient training set of subjects with pulmonary edema with different chest circumferences and different pulmonary water volumes in the preset rotation detection mode in each chest circumference segment; Model training: Use a machine learning algorithm to learn the healthy perturbation coefficient training set and the abnormal perturbation coefficient data set to obtain the pulmonary water volume prediction model.

[0045] In some embodiments, to reduce errors, multiple segments are divided. Specifically, based on a preset maximum chest circumference threshold, a minimum chest circumference threshold, a preset chest circumference interval threshold, and the number of segments N, segmentation is performed to obtain a normal perturbation coefficient set and an abnormal perturbation coefficient set (that is, the perturbation coefficient training set for each segment). For example, the minimum chest circumference is 798mm, the maximum chest circumference is 986mm, the preset chest circumference interval threshold is 23mm, and the preset number of segments is 8, then the segments obtained are: segments less than 798mm, segments from 798mm to 821mm, ···, segments from 936mm to 959mm, segments greater than 982mm.

[0046] Further, due to the different genders having different threshold ranges for chest circumference, different maximum and minimum chest circumference thresholds are preset for different genders respectively, and then segmented. For example, for males, those less than 874 mm are in one segment, those from 874 mm to 980 mm are in one segment, and those greater than 980 mm are in one segment; while for females, those less than 798 mm are in one segment, those from 798 mm to 986 mm are in one segment, and those greater than 986 mm are in one segment. Correspondingly, in the subsequent process, a training set of healthy perturbation coefficients of healthy subjects with different chest circumferences in each chest circumference segment for different genders under the preset rotation detection mode is obtained. Similarly, a training set of abnormal perturbation coefficients of pulmonary edema subjects with different chest circumferences and different pulmonary water volumes in each chest circumference segment for different genders under the preset rotation detection mode is obtained. By obtaining the perturbation coefficient training set by gender and chest circumference segmentation, the error introduced when segmenting the chest circumference without distinguishing gender is eliminated.

[0047] For the user to be detected, almost all can achieve the lying position, but for some users, the sitting position cannot be achieved. However, there are slight differences in the perturbation coefficients detected in different body positions. Therefore, preferably, a set of perturbation coefficients of healthy subjects in the lying position and the perturbation coefficients of non-edema subjects are obtained. Correspondingly, the perturbation coefficients of the user to be detected in the lying position are also obtained during the subsequent prediction process.

[0048] Furthermore, in order to eliminate the influence of body position, the abnormal perturbation coefficients of each pulmonary edema subject in different body positions under the same excitation signal can also be obtained respectively. For example, the first abnormal perturbation coefficient in the sitting position and the second abnormal perturbation coefficient in the lying position under the same detection mode, and then the mean value of the two perturbation coefficients is calculated and used as the abnormal perturbation coefficient of the pulmonary edema subject. Similarly, the mean value of the perturbation coefficients of healthy subjects in different body positions is used as their healthy perturbation coefficient; and during the subsequent prediction process, the mean value of the abnormal perturbation coefficients obtained in different body positions calculated based on the same principle is used as the abnormal perturbation coefficient of the user to be detected and input into the prediction model for prediction.

[0049] In some other embodiments, before obtaining the perturbation coefficient data set based on the electrical signals obtained from six electrode pads, data preprocessing is performed, and the data preprocessing includes data cleaning, denoising, and normalization processing.

[0050] Among them, the steps for data cleaning include: The first electrode and the sixth electrode are respectively arranged at the above two mirror image positions on the backs of healthy subjects and subjects with pulmonary edema, and a set of healthy perturbation coefficients and a set of abnormal perturbation coefficients in a preset rotation detection mode are obtained; specifically, after obtaining the set of perturbation coefficients at the above specified positions on the chest and the side of the chest, the first electrode and the sixth electrode are directly switched to the mirror image positions, and the same detection mode is used for detection again, and then the difference in perturbation coefficients between the chest and the back and the side of the chest is calculated (for example, the difference between the perturbation coefficient when the first electrode at the mirror image position on the chest and back is used as the receiving end and the perturbation coefficient when the second electrode / third electrode is used as the receiving end). If the difference is less than or equal to the difference in perturbation coefficients between the chest and the side of the chest (for example, the difference between the perturbation coefficient when the first electrode located on the chest is used as the receiving end and the perturbation coefficient when the second electrode / third electrode is used as the receiving end), the corresponding training data of the chest and the side of the chest are excluded.

[0051] Embodiment 2: Based on the above method, the present invention also provides a method for predicting (or assessing) the risk of pulmonary edema based on electromagnetic biological detection. Refer to Figure 1 , specifically, in addition to the steps in the above embodiment, the method further includes the steps: S103 Based on the chest circumference data of the user to be detected, the corresponding preset chest circumference segment is matched, and the predicted value of the lung water volume is compared with the preset threshold corresponding to the preset chest circumference segment. If the predicted value of the lung water volume is less than the first preset threshold, the prediction result is no risk; if the predicted value of the lung water volume is greater than the second preset threshold, the prediction result is high risk; if the predicted value of the lung water volume ≥ the first preset threshold and < the third preset threshold, the prediction result is low risk; if the predicted value of the lung water volume ≥ the third preset threshold and < the second preset threshold, the prediction result is medium risk.

[0052] As described above, at least three chest circumference segments are preset, and different preset thresholds are set for each chest circumference segment. Correspondingly, when making a risk determination, the corresponding chest circumference segment and the preset threshold corresponding to the chest circumference segment are first matched for the user to be detected, and then the risk assessment is carried out.

[0053] Specifically, for each segmentation, since the abnormal perturbation coefficient training set of pulmonary edema subjects with different pulmonary water volumes and the healthy perturbation coefficient training set of healthy subjects are obtained in advance, and the oxygenation index is the gold standard for judging the severity of the case, therefore, the oxygenation index of each pulmonary edema subject can be obtained in advance (obtaining the oxygenation index is an existing technology and will not be elaborated here), and then user classification is performed based on the oxygenation index to obtain low-risk users, medium-risk users, high-risk users, and risk-free users respectively, and the corresponding perturbation coefficient training sets are marked. Then, for low-risk users, medium-risk users, and high-risk users, the corresponding average pulmonary water volume is calculated as the preset threshold for each segmentation and each type of user.

[0054] Embodiment 3: Refer to Figure 3 , which is a flowchart of another embodiment of a method for predicting the risk of pulmonary edema based on electromagnetic field biological detection according to the invention. Specifically, the method includes the above step S101, except that in this embodiment, the risk prediction is not based on the pulmonary water volume, but on the pulmonary impedance spectrum Z b of the user to be detected and the standard pulmonary impedance spectrum Z a of the corresponding healthy subject for risk prediction. Specifically, after performing step S101, the following steps are executed: S104 Calculate the pulmonary impedance spectrum of the user to be detected based on the perturbation coefficient detected under the rotation detection mode , where are the frequency components of the vector.

[0055] S105 Match the corresponding preset chest circumference segmentation of the user to be detected based on the chest circumference data of the user to be detected, and the standard pulmonary impedance spectrum corresponding to the preset chest circumference segmentation.

[0056] In some embodiments, the step of calculating the standard pulmonary impedance spectrum of the corresponding chest circumference segmentation specifically includes the following steps: Set at least three chest circumference segments based on the preset maximum chest circumference threshold and minimum chest circumference threshold; Screen the corresponding healthy subject groups based on each chest circumference segment to obtain at least three healthy subject groups; Obtain the healthy perturbation coefficient set of all healthy subjects under the preset rotation detection mode, and calculate the pulmonary tissue impedance spectrum of each healthy subject based on the healthy perturbation coefficient detected by each healthy subject under the rotation detection mode; Use the least squares method to fit the pulmonary tissue impedance spectra of all healthy subjects in each group respectively to obtain the standard pulmonary impedance spectrum corresponding to each group, that is, each chest circumference segment ; Among them, , is each frequency component of the vector , and n is the total number of frequency components.

[0057] In this embodiment, by segmenting the chest circumference and making the proportion of the perturbation coefficient set in each segment the same, the chest circumference distribution of healthy subjects is made more reasonable, thereby minimizing the influence caused by a large chest circumference difference.

[0058] Referring to the above embodiment, when segmenting the chest circumference without distinguishing gender, in order to reduce errors, multiple segments are divided. Specifically, based on a preset maximum chest circumference threshold, a minimum chest circumference threshold, a preset chest circumference interval threshold, and the number of segments N, the segmentation is performed to obtain a normal perturbation coefficient set and an abnormal perturbation coefficient set (i.e., the perturbation coefficient training set) for each segment. For example, the minimum chest circumference is 798 mm, the maximum chest circumference is 986 mm, the preset chest circumference interval threshold is 23 mm, and the preset number of segments is 8, then the segments obtained are: the segment less than 798 mm, the segment from 798 mm to 821 mm, ···, the segment from 936 mm to 959 mm, the segment greater than 982 mm.

[0059] In other embodiments, different maximum chest circumference thresholds and minimum chest circumference thresholds are preset for different genders and segmented. For example, for men, the segment less than 874 mm is one segment, the segment from 874 mm to 980 mm is one segment, and the segment greater than 980 mm is one segment; for women, the segment less than 798 mm is one segment, the segment from 798 mm to 986 mm is one segment, and the segment greater than 986 mm is one segment. Correspondingly, in the subsequent process, a healthy perturbation coefficient training set of healthy subjects in different chest circumference segments under different genders in a preset rotation detection mode is obtained. Similarly, an abnormal perturbation coefficient training set of pulmonary edema subjects with different pulmonary water volumes in different chest circumference segments under different genders in a preset rotation detection mode is obtained. By obtaining the training set by gender and segment, the error introduced when segmenting without distinguishing gender is eliminated.

[0060] S106 Calculate the cosine similarity Similarity between the lung impedance spectrum of the user to be detected and the standard lung impedance spectrum of the corresponding chest circumference segment .

[0061] In some embodiments, , is the i-th frequency component of the lung impedance spectrum , is the i-th frequency component of the standard lung impedance spectrum ; n is the total number of frequency components; among them, is to make the vector Projection onto a vector The inner product of these two vectors The angle between two vectors

[0062] S107 Based on the predicted cosine similarity Similarity, obtain the prediction result of pulmonary edema occurring in the user to be detected; if the cosine similarity Similarity is 1, the prediction result is no risk; if the cosine similarity Similarity is -1, the prediction result is high risk; if the cosine similarity Similarity ≥ 0 and < 1, the prediction result is low risk; if the cosine similarity Similarity < 0 and > -1, the prediction result is medium risk.

[0063] Using this method, the perturbation coefficient sets of nearly 200 healthy people and patients with pulmonary edema were obtained and analyzed, and the results showed Figure 4 as follows: The perturbation coefficients of the pulmonary edema group were significantly lower than those of the control group (i.e., the healthy group). Therefore, the risk of pulmonary edema can be predicted based on the perturbation coefficient. Compared with the traditional screening method, using this method for the evaluation of patients' pulmonary edema has good stability, high primary screening accuracy, and monitoring accuracy.

[0064] See Figure 5 , and the perturbation coefficients of 66 healthy subjects (including 39 males and 27 females) were statistically analyzed, and all were found to satisfy the normal distribution (p = 0.69 > 0.05). The maximum perturbation coefficient Max of these 66 healthy subjects was 330, and the minimum perturbation coefficient was 105.

[0065] To verify the reliability of this method, 8 healthy subjects were repeatedly monitored using the method of the present invention, and 1 of them was invalid data. See Figure 6 , among which, there was no significant difference in the two measurement data of the 7 effective subjects (T = 0.374, P = 0.715 > 0.05). Among them, P and T are common parameters in statistics.

[0066] See Figure 7 , and the perturbation coefficients of 16 pulmonary edema subjects (12 males and 4 females) among the effective cases were statistically analyzed and found not to satisfy the normal distribution (p = 0.04 < 0.05).

[0067] See Figure 8 It can be seen that the perturbation coefficients of the healthy group satisfy the normal distribution, while those of the case group do not. The rank sum test was selected to compare the differences in perturbation coefficients between the healthy group and the case group, and the results showed that the differences in the perturbation coefficient values between the two groups were statistically significant (p < 0.001).

[0068] Embodiment 4: Based on the above method, the present invention also provides an edema risk prediction system based on electromagnetic field biological detection, which includes: Six electrode patches, which are used to construct an excitation signal propagation path between six specified positions on the chest of the user to be detected, and obtain the perturbation coefficient of each propagation path in a preset rotation detection mode; see Figure 2 , the six preset positions are: the first electrode is located at the 3rd and 4th intercostal spaces on the right midclavicular line, the second and third electrodes are arranged side by side between the 6th and 7th ribs on the right midaxillary line, the fourth and fifth electrodes are arranged side by side between the 6th and 7th ribs on the left midaxillary line, and the sixth electrode is located at the 3rd and 4th intercostal spaces on the left midclavicular line; A data processing module, which is used to calculate the mean value of the perturbation coefficients on all propagation paths in the preset rotation detection mode as the perturbation coefficient of the user to be detected; An excitation generator, which is used to generate electrical stimulation signals (i.e., excitation signals) to the six electrode patches respectively based on the preset rotation detection mode; A first prediction module, which is used to input the perturbation coefficient detected in the rotation detection mode into a pre-trained pulmonary water volume prediction model to predict the predicted value of the pulmonary water volume of the user to be detected; and predict the prediction result of the user to be detected having pulmonary edema based on the predicted value of the pulmonary water volume; if the predicted value of the pulmonary water volume is less than the first preset threshold, the prediction result is no risk; if the predicted value of the pulmonary water volume is greater than the second preset threshold, the prediction result is high risk; if the predicted value of the pulmonary water volume ≥ the first preset threshold and < the third preset threshold, the prediction result is low risk; if the predicted value of the pulmonary water volume ≥ the third preset threshold and < the second preset threshold, the prediction result is medium risk; or A second prediction module, which is used to calculate the pulmonary impedance spectrum of the user to be detected based on the perturbation coefficient detected in the rotation detection mode ; and calculate the pulmonary impedance spectrum of the user to be detected based on the cosine similarity theory The cosine similarity Similarity between the pulmonary impedance spectrum of the user to be detected and the pre-calculated pulmonary impedance spectrum of healthy subjects Finally, predict the prediction result of the user to be detected having pulmonary edema based on the cosine similarity Similarity; if the cosine similarity Similarity is 1, the prediction result is no risk; if the cosine similarity Similarity is -1, the prediction result is high risk; if the cosine similarity Similarity ≥ 0 and < 1, the prediction result is low risk; if the cosine similarity Similarity < 0 and > -1, the prediction result is medium risk.

[0069] In some embodiments, the above-mentioned preset rotation detection mode means that detections under multiple detection modes are completed within a preset cycle time, and excitation signals in the frequency band of 10K - 100K are sequentially applied under each detection mode; wherein, the multiple detection modes specifically include: The first detection mode: any two of the six electrodes are used as the transmitting ends, any two of the remaining four electrodes are used as the receiving ends, and the other two electrodes are used as the grounding ends; and / or, The second detection mode: any two of the six electrodes are simultaneously used as both the transmitting end and the receiving end, and the remaining four electrodes are used as the grounding ends; and / or, The third detection mode: any two of the six electrodes are simultaneously used as both the transmitting end and the receiving end, any two of the remaining four electrodes are grounded, and the other two electrodes are floating.

[0070] Compared with the low-frequency band: below 10KHz and the high-frequency band: above 50KHz, the mean value of the perturbation coefficient obtained in the frequency band of 10KHz - 100KHz is more stable. Therefore, in order to reduce errors, in this embodiment, excitation signals in the frequency band of 10KHz - 100KHz are mainly used for stimulation.

[0071] In some embodiments, at least three chest circumference segments are preset based on a preset maximum chest circumference threshold and a minimum chest circumference threshold; then, corresponding groups of healthy subjects are screened based on each chest circumference segment to obtain at least three groups of healthy subjects, and finally, the lung tissue impedance spectra of each group are calculated respectively. Specifically, the above-mentioned second prediction module specifically includes: a first calculation unit, configured to calculate the lung tissue impedance spectrum of each healthy subject in each group based on the set of healthy perturbation coefficients of each group of healthy subjects detected under the rotation detection mode; a second calculation unit, configured to use the least squares method to fit the lung tissue impedance spectra of each healthy subject in each group respectively to obtain the standard lung impedance spectrum of each group .

[0072] In some embodiments, when the system includes a first prediction module, the system further includes: A dataset construction module, configured to obtain a training set of healthy perturbation coefficients of healthy subjects within different chest circumference segments under a preset rotation detection mode; obtain a training set of abnormal perturbation coefficients of pulmonary edema subjects with different chest circumference segments and different lung water volumes under a preset rotation detection mode; wherein, the chest circumference segments are at least three chest circumference segments preset based on a preset maximum chest circumference threshold and a minimum chest circumference threshold; Model training: Use a machine learning algorithm to learn from the training set of healthy perturbation coefficients and the dataset of abnormal perturbation coefficients to obtain the lung water volume prediction model.

[0073] In some embodiments, when obtaining the abnormal perturbation coefficient training set of pulmonary edema subjects with different chest circumference segments and different pulmonary water volumes in a preset rotation detection mode, for the same pulmonary edema subject, the first abnormal perturbation coefficient of the pulmonary edema subject in the sitting position and the second abnormal perturbation coefficient are obtained respectively, and the mean value of the first abnormal perturbation coefficient in the sitting position and the second abnormal perturbation coefficient corresponding to the lying position under the same excitation signal is used as the abnormal perturbation coefficient of the pulmonary edema subject, so as to obtain the abnormal perturbation coefficient set of each group.

[0074] In some embodiments, the system further includes a data preprocessing module for preprocessing the electrical signals obtained through six electrode patches, and the data preprocessing includes data cleaning, denoising and normalization processing.

[0075] It should be noted that in this article, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0076] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc), and includes several instructions for causing a computer terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0077] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit and scope protected by the present invention and the claims. All of these are within the protection scope of the present invention.

Claims

1. A method for predicting the risk of pulmonary edema based on electromagnetic field biological detection, characterized in that, Including: Six electrode patches are set at six specified positions on the chest of the user to be detected, and the perturbation coefficient in a preset rotation detection mode is obtained; the six preset positions are respectively: the first electrode is located at the 3rd and 4th intercostal spaces on the right midclavicular line in front of the chest, the second and third electrodes are arranged side by side between the 6th and 7th ribs on the right midaxillary line on the side of the chest, the fourth and fifth electrodes are arranged side by side between the 6th and 7th ribs on the left midaxillary line on the side of the chest, and the sixth electrode is located at the 3rd and 4th intercostal spaces on the left midclavicular line in front of the chest; the perturbation coefficient is the average value of the perturbation coefficients on the propagation path from all electrodes as the transmitting end to each electrode as the receiving end; The perturbation coefficient detected in the rotation detection mode is input into a pre-trained lung water volume prediction model to predict the lung water volume prediction value of the user to be detected; and based on the chest circumference data of the user to be detected obtained in advance, the corresponding preset chest circumference segment is matched, and the lung water volume prediction value is compared with the preset threshold corresponding to the preset chest circumference segment. If the lung water volume prediction value is less than the first preset threshold, the prediction result is no risk; if the lung water volume prediction value is greater than the second preset threshold, the prediction result is high risk; if the lung water volume prediction value ≥ the first preset threshold and < the third preset threshold, the prediction result is low risk; If the lung water volume prediction value ≥ the third preset threshold and < the second preset threshold, the prediction result is medium risk; Alternatively, calculate the lung impedance spectrum of the user to be detected based on the perturbation coefficient detected in the rotation detection mode. ; and match the corresponding preset chest circumference segment based on the chest circumference data of the user to be detected obtained in advance, and calculate the lung impedance spectrum of the user to be detected based on the cosine similarity theory. The standard lung impedance spectrum corresponding to the preset chest circumference segment calculated in advance. The cosine similarity Similarity between them; then predict the prediction result of pulmonary edema of the user to be detected based on the cosine similarity Similarity; if the cosine similarity Similarity is 1, the prediction result is no risk; if the cosine similarity Similarity is -1, the prediction result is high risk; if the cosine similarity Similarity ≥ 0 and < 1, the prediction result is low risk; if the cosine similarity Similarity < 0 and > -1, the prediction result is medium risk. Among them, , is the i-th frequency component of is the i-th frequency component of, and n is the total number of frequency components.

2. The method for predicting the risk of pulmonary edema based on electromagnetic field biological detection according to claim 1, wherein, The preset rotation detection mode means that the detection in multiple detection modes is completed within a preset cycle time, and an excitation signal in the frequency band of 10KHz - 100KHz is applied in sequence in each detection mode; wherein, the multiple detection modes specifically include: The first detection mode: any two of the six electrodes are used as the transmitting end, any two of the remaining four electrodes are used as the receiving end, and the other two electrodes are used as the grounding end; and / or, The second detection mode: any two of the six electrodes are used as both the transmitting end and the receiving end at the same time, and the remaining four electrodes are used as the grounding end; and / or, The third detection mode: any two of the six electrodes are used as both the transmitting end and the receiving end at the same time, any two of the remaining four electrodes are grounded, and the other two electrodes are suspended.

3. The method for predicting the risk of pulmonary edema based on electromagnetic field biological detection according to claim 2, wherein, The steps of calculating the standard lung impedance spectrum Za corresponding to the preset chest circumference segment specifically include the steps of: Setting at least three chest circumference segments based on the preset maximum chest circumference threshold and minimum chest circumference threshold; Screening the corresponding healthy subject groups based on each chest circumference segment to obtain at least three healthy subject groups; Obtaining the set of healthy perturbation coefficients of all healthy subjects in the preset rotation detection mode, and calculating the lung tissue impedance spectrum of each healthy subject based on the respective healthy perturbation coefficients of each healthy subject; Using the least squares method to fit the lung tissue impedance spectra of all healthy subjects in each group respectively, to obtain the standard lung impedance spectra for each group .

4. A method for predicting the risk of pulmonary edema based on electromagnetic field biological detection according to claim 1, characterized in that, The steps of training the lung water volume prediction model specifically include: Constructing a training dataset: Set at least three chest circumference segments based on a preset maximum chest circumference threshold and a minimum chest circumference threshold; Obtain a training set of healthy perturbation coefficients of healthy subjects with different chest circumferences within each chest circumference segment under a preset rotation detection mode; Obtain a training set of abnormal perturbation coefficients of pulmonary edema subjects with different chest circumferences and different pulmonary water volumes within each chest circumference segment under a preset rotation detection mode. Model training: Use a machine learning algorithm to learn from the training set of healthy perturbation coefficients and the dataset of abnormal perturbation coefficients to obtain the pulmonary water volume prediction model.

5. A method for predicting the risk of pulmonary edema based on electromagnetic field biological detection according to claim 4, characterized in that, When obtaining the training set of abnormal perturbation coefficients of pulmonary edema subjects with different chest circumferences and different pulmonary water volumes within each chest circumference segment under a preset rotation detection mode, for the same pulmonary edema subject, respectively obtain the first abnormal perturbation coefficient and the second abnormal perturbation coefficient of the pulmonary edema subject in the sitting position and the lying position; And take the mean of the first abnormal perturbation coefficient in the sitting position and the second abnormal perturbation coefficient corresponding to the lying position under the same excitation signal as the abnormal perturbation coefficient of the pulmonary edema subject, and obtain the training set of abnormal perturbation coefficients corresponding to each chest circumference segment.

6. A pulmonary edema risk prediction system based on electromagnetic field biological detection, characterized in that, Including: Six electrode patches, used to construct an excitation signal propagation path between six specified positions on the chest of the user to be detected, and obtain the perturbation coefficient of each propagation path under a preset rotation detection mode; The six preset positions are: The first electrode is located at the 3rd and 4th intercostal spaces on the right midclavicular line in front of the user's chest, the second and third electrodes are arranged side by side between the 6th and 7th ribs on the right axillary midline on the side of the user's chest, the fourth and fifth electrodes are arranged side by side between the 6th and 7th ribs on the left axillary midline on the side of the user's chest, and the sixth electrode is located at the 3rd and 4th intercostal spaces on the left midclavicular line in front of the user's chest. A data processing module, used to calculate the mean value of the perturbation coefficients on all propagation paths under a preset rotation detection mode as the perturbation coefficient of the user to be detected. Specifically, a perturbation coefficient will be generated for each propagation path from the electrode as the transmitter to each electrode as the receiver. An excitation generator, used to send electrical stimulation signals to any two of the six electrode patches respectively based on a preset rotation detection mode. A first prediction module, used to input the perturbation coefficient detected under the rotation detection mode into a pre-trained pulmonary water volume prediction model to predict the predicted value of the pulmonary water volume of the user to be detected; And match the corresponding preset chest circumference segment based on the chest circumference data of the user to be detected, and calculate and compare the predicted value of the pulmonary water volume with the preset threshold corresponding to the preset chest circumference segment. If the predicted value of the pulmonary water volume is less than the first preset threshold, the prediction result is no risk; If the predicted value of the pulmonary water volume is greater than the second preset threshold, the prediction result is high risk; If the predicted value of the pulmonary water volume ≥ the first preset threshold and < the third preset threshold, the prediction result is low risk. If the predicted value of the pulmonary water volume ≥ the third preset threshold and < the second preset threshold, the prediction result is medium risk. Or, A second prediction module, configured to calculate the lung impedance spectrum of the user to be detected based on the perturbation coefficient detected in the rotation detection mode ; match the corresponding preset chest circumference segment based on the chest circumference data of the user to be detected, and then calculate the lung impedance spectrum of the user to be detected based on the cosine similarity theory and the standard lung impedance spectrum of the preset chest circumference segment calculated in advance to obtain the cosine similarity Similarity therebetween. Finally, predict the prediction result of pulmonary edema occurrence of the user to be detected based on the cosine similarity Similarity; if the cosine similarity Similarity is 1, the prediction result is no risk; if the cosine similarity Similarity is -1, the prediction result is high risk; if the cosine similarity Similarity ≥ 0 and < 1, the prediction result is low risk; if the cosine similarity Similarity < 0 and > -1, the prediction result is medium risk; Among them, , is the i-th frequency component of is the i-th frequency component of, and n is the total number of frequency components.

7. The risk prediction system for pulmonary edema based on electromagnetic field biological detection according to claim 6, wherein The above-mentioned preset rotation detection mode refers to completing detections under multiple detection modes within a preset cycle time, and successively applying excitation signals in the frequency band of 10KHz - 100KHz under each detection mode; wherein, the multiple detection modes specifically include: The first detection mode: using any two of the six electrodes as the transmitting end, any two of the remaining four electrodes as the receiving end, and the other two electrodes as the grounding end; and / or, The second detection mode: using any two of the six electrodes as both the transmitting end and the receiving end simultaneously, and the remaining four electrodes as the grounding end; and / or, The third detection mode: using any two of the six electrodes as both the transmitting end and the receiving end simultaneously, grounding any two of the remaining four electrodes, and leaving the other two electrodes floating.

8. The risk prediction system for pulmonary edema based on electromagnetic field biological detection according to claim 6, wherein The second prediction module specifically includes: a first calculation unit, configured to calculate the lung tissue impedance spectrum of each healthy subject in each group based on the preset healthy perturbation coefficient sets of each group of healthy subjects detected in the rotation detection mode; and a second calculation unit, configured to respectively fit the lung tissue impedance spectra of each healthy subject in each group by using the least squares method to obtain the standard lung impedance spectrum of each group .

9. The risk prediction system for pulmonary edema based on electromagnetic field biological detection according to claim 6, characterized in that, It further includes: A model training module, which is used to respectively obtain the abnormal perturbation coefficient sets of multiple pulmonary edema subjects in the sitting position and the lying position under the preset rotation detection mode, so as to construct a training sample library, and use a machine learning algorithm to learn the abnormal perturbation coefficient sets in the training sample library to obtain the pulmonary water volume prediction model.

10. The pulmonary edema risk prediction system based on electromagnetic field biological detection according to claim 6, characterized in that, It further includes a data preprocessing module, which is used to perform data preprocessing on the electrical signals obtained through six electrode patches, and the data preprocessing includes data cleaning, denoising, and normalization processing.

Images