Non-invasive real-time detection method and system for characteristic substances of yin-yang deficiency syndrome of human body

Abstract

The invention provides a non-invasive real-time detection method and system for characteristic substances of yin and yang deficiency syndromes of a human body, and relates to the technical field of substance detection.The specific scheme includes the steps that an exhaled gas sample containing life substances carrying physiological or pathological information is collected; analyzing the exhaled gas sample based on electrospray extraction ionization mass spectrometry, and marking exhaled gas characteristic molecule information to obtain a yin-yang data set; constructing an initial model by combining a random forest and a gradient boosting decision tree algorithm, training and testing the initial model based on the yin-yang data set to obtain a yin-yang identification model, and associating yin-yang attributes of traditional Chinese medicine diseases with exhaled gas characteristic molecule information; real-time exhaled gas of the subject is obtained, exhaled gas characteristic molecule information is detected, the yin-yang deficiency syndrome is recognized based on the yin-yang recognition model, and a yin-yang deficiency syndrome detection report is obtained and output. The invention aims to solve the problems that the traditional static analysis is difficult to directly distinguish yin and yang of a human body and the accuracy is relatively low by utilizing high-latitude mass spectrum data.

Description

Technical Field

[0001] The present invention relates to the technical field of material detection, and in particular to a non-invasive real-time detection method and system for human body yin-yang deficiency syndrome characteristic materials. Background Art

[0002] Yin-Yang deficiency syndrome is a manifestation of human pathology in TCM theory, which is reflected in abnormal human metabolism and material changes. At present, TCM physicians use their senses to collect information to diagnose Yin-Yang deficiency syndrome, but the determination of Yin-Yang deficiency syndrome has complex problems such as multi-label intersection and fuzziness, and the sensitivity and path of human senses are limited, information collection is subjective, and the amount of information is too small. Clinical diagnosis depends on the doctor's personal experience, which requires long-term learning and practice to ensure the efficacy. Due to the lack of objective instruments and equipment, the research on human basic materials has been slow, which has long restricted the development of TCM theory. With the introduction of modern science and technology and the urgent need for TCM to keep pace with the world, the subjective research method of TCM has gradually developed towards objectivity. In TCM theory, "lungs face all meridians" means that the chemical substances exhaled by the human body contain material information that can characterize the characteristics of Yin-Yang deficiency in the human body, and the exhaled substances come from various parts of the body. Therefore, by detecting metabolic exhaled excretions and analyzing the test results of each group with Yin-Yang deficiency, specific substances related to Yin-Yang deficiency in the organs can be accurately identified, and parameters such as the type and degree of abnormal manifestations of Yin-Yang deficiency in the human body can be analyzed and evaluated, so as to obtain more scientific and reliable results. Medical testing has developed from invasive to non-invasive real-time detection, and has great potential in the future in early diagnosis, dynamic tracking, telemedicine, etc. It is the inevitable direction of future medical equipment development. Among them, the in vivo mass spectrometry analysis method of exhaled gas can conduct real-time analysis of trace non-volatile metabolites and toxic and harmful substances in exhaled gas, providing new ideas and methods for non-invasive online diagnosis of diseases. So far, more than 3,500 components in exhaled breath have been analyzed. These components are emitted from different sources by the human respiratory system. These components not only contain important information related to the biochemical processes of the human body, but also contain material information in traditional Chinese medicine theory that can characterize the yin and yang characteristics of the human body. Through the research and development of real-time non-invasive detection instruments that can directly detect trace substances in human exhaled gas and are used to detect characteristic substances in real time, exhaled gas can be collected in real time under non-invasive conditions on the human body. Dynamic tracking of physiology and pathology, and elaboration of the dynamic laws of the growth of yin and yang in the human body and the material basis and scientific connotation of yin and yang false evidence are important ways to promote medical scientific research and are also the trend of development of modern medicine. However, when classical mass spectrometry technology is used to detect actual samples, the sample pretreatment process is lengthy and complicated, which may cause sample loss, contamination, degradation or cause biological stress, leading to low analysis efficiency and poor reliability, and ultimately unstable results. Electrospray extraction ionization mass spectrometry (EESI-MS) technology is a method that can quickly and directly analyze various forms of complex matrix samples (such as liquids, gases, aerosols, suspensions, solid surfaces, powders and colloids) without sample pretreatment. It has the advantages of high sensitivity, fast analysis speed and good stability. It has great potential for real-time analysis of polar, non-polar, volatile and non-volatile chemicals in various forms of samples.

[0003] However, it is difficult to directly distinguish between yin and yang in the human body using high-latitude mass spectrometry data, the accuracy is relatively low, and its use is greatly limited. Summary of the invention

[0004] The purpose of the present invention is to provide a non-invasive real-time detection method and system for characteristic substances of human yin and yang deficiency syndrome, which improves the problem that traditional static analysis using high-latitude mass spectrometry data is difficult to directly distinguish between human yin and yang and has relatively low accuracy.

[0005] The present invention provides a non-invasive real-time detection method and system for characteristic substances of human yin and yang deficiency syndrome, which adopts the following technical solutions: In the first aspect, a non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome specifically comprises: Collect exhaled breath samples containing vital substances that carry physiological or pathological information; Analyzing the exhaled gas sample based on electrospray extraction ionization mass spectrometry, marking characteristic molecular information of the exhaled gas to obtain a yin-yang data set; Combining random forest and gradient boosting decision tree algorithms to build an initial model, training and testing the initial model based on the Yin-Yang dataset to obtain a Yin-Yang identification model, and correlating the Yin-Yang attributes of TCM diseases with characteristic molecular information of exhaled breath; Acquire the real-time exhaled gas of the subjects and detect the characteristic molecular information of the exhaled gas. Construct a hyperplane based on the Yin-Yang identification model to classify the gas characteristic molecular information into two categories, obtain the Yin-Yang deficiency syndrome information of the subjects, and produce and output the Yin-Yang deficiency syndrome detection report.

[0006] Optionally, an exhaled breath sample containing life substances carrying physiological or pathological information is collected, wherein the exhaled breath sample includes amino acids and derivatives, fatty acids and lipids, steroids and hormones, organic compounds and acids, and organic compounds.

[0007] Optionally, the exhaled gas sample is analyzed based on electrospray extraction ionization mass spectrometry, and characteristic molecular information of the exhaled gas is marked to obtain a yin-yang data set, including: ionizing the exhaled gas sample based on electrospray extraction ionization mass spectrometry to obtain gas ions; Maintaining the vacuum environment, absorbing the gas ions and pressurizing them to perform mixed tandem mass analysis to obtain a mass-to-charge ratio data set and a mass analysis data set; The yin-yang dataset was obtained by fusing the mass-to-charge ratio dataset and the mass analysis dataset and unifying their formats, marking the timestamps and the characteristic molecular information of the exhaled gas.

[0008] Optionally, training and testing an initial model based on the yin-yang data set to obtain a yin-yang identification model includes: Initialize the parameters of the initial model; The yin-yang data set is randomly divided into a training set and a test set, the training set is input into the initial model to obtain a predicted value, the loss error between the predicted value and the true value is calculated, a loss function is constructed, the loss function is minimized based on the gradient descent method, the initial model parameters are iteratively updated until the loss function value tends to be stable or the maximum number of iterations is reached to terminate the training, and the initial model is evaluated and fine-tuned based on the test set to obtain a yin-yang identification model.

[0009] Optionally, a hyperplane is constructed based on the yin-yang identification model to perform binary classification on the gas characteristic molecule information, including: Construct a hyperplane based on the parameters and support vectors of the Yin-Yang identification model; Calculate the distance from the subject sample to the hyperplane, and determine the category to which the sample belongs based on the positive or negative value of the distance. If the sample distance is negative, it belongs to Yin deficiency syndrome, and if the sample distance is positive, it belongs to Yang deficiency syndrome.

[0010] The beneficial effect of the non-invasive real-time detection method of characteristic substances of human yin and yang deficiency syndrome provided by the present invention is: by collecting exhaled gas containing life substances carrying physiological / pathological information and marking it as an exhaled gas sample, analyzing the exhaled gas sample based on electrospray extraction ionization mass spectrometry, marking the characteristic molecular information of the exhaled gas, building a database and recognition model based on the analysis results through an artificial intelligence algorithm, associating the yin and yang attributes of traditional Chinese medicine diseases with the characteristic molecular information of exhaled gas, obtaining real-time exhaled gas of the subject, detecting the characteristic molecular information of exhaled gas, and identifying yin and yang deficiency syndrome based on the database and recognition model. The detection system combines EESI-MS technology and artificial intelligence measurement technology, and can realize the identification of specific substances of yin and yang deficiency syndrome and the objective quantitative expression of yin and yang deficiency syndrome results by real-time and non-invasive detection of exhaled substances, and then analyze and judge to obtain objective and accurate dialectical conclusions.

[0011] In the second aspect, a non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome specifically comprises: A sample collection module for collecting exhaled gas samples containing life substances carrying physiological or pathological information; A mass spectrometry analysis module analyzes the exhaled gas sample based on electrospray extraction ionization mass spectrometry, and marks characteristic molecular information of the exhaled gas to obtain a positive and negative data set; A model building module, for building an initial model by combining random forest and gradient boosting decision tree algorithms, training and testing the initial model based on the yin-yang data set to obtain a yin-yang identification model, and associating the yin-yang attributes of TCM diseases with characteristic molecular information of exhaled breath; The real-time detection module is used to obtain the real-time exhaled gas of the subject and detect the characteristic molecular information of the exhaled gas, identify the Yin-Yang deficiency syndrome based on the Yin-Yang identification model, and obtain and output the Yin-Yang deficiency syndrome detection report.

[0012] Optionally, the mass spectrometry module includes a differential vacuum system, an EESI ionizer, an ion introducer, an ion transporter, a hybrid tandem mass analyzer, and an ion detector; The differential vacuum system and the EESI ionizer for analog and digital conversion of the collected data use electrospray ionization technology to introduce the exhaled gas sample and convert it into ions. The ion introducer includes a mechanical pump with a pumping speed of 16L / s and operates at a pressure of 200Pa to maintain the vacuum environment and extract ions from the EESI ionizer. The ion transporter uses a molecular pump with a pumping speed of 80L / s and operates at a pressure of 0.2Pa to move the ions to the hybrid tandem mass analyzer in the differential vacuum system.

[0013] Optionally, the hybrid tandem mass analyzer comprises a quadrupole mass selector, a linear ion trap mass analyzer and a time-of-flight mass analyzer, the hybrid tandem mass analyzer separates and analyzes ions according to their mass-to-charge ratio, transports ions in the time-of-flight mass analyzer including a molecular pump, the pump speed is 300 L / s, and operates at a pressure of 0.0005 Pa, the time-of-flight mass analyzer is used to determine the mass of the ions based on the flight time of the ions through the drift region, the ions that have undergone mass analysis in the differential vacuum system are ultimately directed to an ion detector, the ion detector collects analog ions and converts the collected data into a digital format, records the number of ions detected, and marks a timestamp for each ion event.

[0014] Optionally, the yin-yang identification model in the model building module includes a random forest, a gradient boosting decision tree algorithm and a support vector machine algorithm. The random forest uses ensemble learning to improve classification accuracy, the gradient boosting decision tree adds decision trees in sequence to optimize prediction performance, and the support vector machine constructs a hyperplane in a high-dimensional space to perform yin-yang deficiency syndrome identification.

[0015] Optionally, the differential vacuum system adopts a four-stage differential vacuum structure, wherein: An ion transfer tube for placing a normal pressure ion introduction interface in the first-stage vacuum chamber; The second stage of vacuum is used to place the ion transporter; The third vacuum stage is used to place the quadrupole mass selector and ion trap mass analyzer; The fourth vacuum stage is used to house the time-of-flight mass analyzer.

[0016] The beneficial effects of the second aspect can refer to the description of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of the overall steps of a method for non-invasive real-time detection of characteristic substances of human yin and yang deficiency syndrome provided by the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be understood by people with general skills in the field to which the present invention belongs. "Including" and similar words used in this article mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0019] The embodiment of the present invention provides a non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome, comprising the following steps: S1. Collect exhaled gas samples containing life substances carrying physiological or pathological information; S2. analyzing the exhaled gas sample based on electrospray extraction ionization mass spectrometry, marking characteristic molecular information of the exhaled gas to obtain a yin-yang data set; S3, combining random forest and gradient boosting decision tree algorithms to build an initial model, training and testing the initial model based on the Yin-Yang dataset to obtain a Yin-Yang identification model, and correlating the Yin-Yang attributes of TCM diseases with characteristic molecular information of exhaled breath; S4. Acquire the real-time exhaled gas of the subject and detect the characteristic molecular information of the exhaled gas. Construct a hyperplane based on the Yin-Yang identification model to classify the gas characteristic molecular information into two categories, obtain the Yin-Yang deficiency syndrome information of the subject, and prepare and output the Yin-Yang deficiency syndrome detection report.

[0020] In some embodiments, when executing step S1 to collect an exhaled breath sample containing life substances carrying physiological or pathological information, the exhaled breath sample includes amino acids and derivatives, fatty acids and lipids, steroids and hormones, organic compounds and acids, and organic compounds.

[0021] In some embodiments, executing step S2 specifically includes: S2-1, ionizing the exhaled gas sample based on electrospray extraction ionization mass spectrometry to obtain gas ions; S2-2, absorbing the gas ions under vacuum and pressurizing them to perform mixed tandem mass analysis to obtain a mass-to-charge ratio data set and a mass analysis data set; S2-3. The mass-to-charge ratio dataset and the mass analysis dataset are integrated and formatted in a unified manner. The timestamp and characteristic molecular information of the exhaled gas are marked to obtain the yin-yang dataset.

[0022] Specifically, when performing step S2-1, an electrospray enabled solute-solute-solvent ionizer (EESI), a differential vacuum system, and an EESI ionizer for analog and digital conversion of collected data use electrospray ionization technology to introduce an exhaled gas sample and convert it into ions.

[0023] Specifically, when executing step S2-2, the mass spectrometry analysis of the breath samples of the Yin deficiency constitution group, the neutral constitution group, and the Yang deficiency constitution group uses the PLS-DA algorithm to screen important features according to the VIP value, and the eigenvalues ​​of the mass spectrum matrix (the first 100 VIP values) can be obtained by analyzing and calculating the simplified mass spectrum variables.

[0024] Specifically, when executing step S2-3, the number of detected ions is recorded, and a specific time (timestamp) is assigned to each ion event to facilitate further analysis and interpretation of the collected data.

[0025] In some embodiments, executing step S3, training and testing the initial model based on the yin-yang data set to obtain a yin-yang identification model, includes: Initialize the parameters of the initial model; The yin-yang data set is randomly divided into a training set and a test set, the training set is input into the initial model to obtain a predicted value, the loss error between the predicted value and the true value is calculated, a loss function is constructed, the loss function is minimized based on the gradient descent method, the initial model parameters are iteratively updated until the loss function value tends to be stable or the maximum number of iterations is reached to terminate the training, and the initial model is evaluated and fine-tuned based on the test set to obtain a yin-yang identification model.

[0026] Specifically, the initial model is constructed by combining random forest and gradient boosting decision tree algorithms. In the process of selecting model parameters, a grid search method is adopted, that is, parameters of various orders of magnitude are traversed to automatically select the parameters that best suit the model. The F1 value is used as a selection criterion in the selection process. The F1 value is an evaluation index that balances accuracy and recall, and is more accurate than simply using accuracy as an evaluation index. In order to prevent unexpected phenomena, a 10-fold cross-validation method is used, that is, the data set is randomly divided into 10 blocks, 9 of which are used as training sets, and the remaining 1 block is used as a test set, which is repeated 10 times. The average value of the 10 F1 values ​​is taken as the score of the model under this parameter.

[0027] Further, according to the influence curve of the number of decision trees on the model accuracy and F1 value in the random forest, it is found that the overall result gradually becomes stable when the number of decision trees is about 25, so the number of decision trees in the random forest is set to 25. In the ten-fold cross-validation process, the average accuracy is 82.00% and the average F1 value is 74.25%. After 60 trees, it gradually rises and falls. In the process of ten-fold cross-validation, the best result can reach 100% accuracy, but the performance form for some samples is poor, with an accuracy of only 60%, but the overall average accuracy can reach 80.00%.

[0028] In some embodiments, when executing step S4, it includes: S4-1. Acquire real-time exhaled gas of the subject and detect characteristic molecular information of the exhaled gas; S4-2. Identify Yin-Yang deficiency syndrome based on the Yin-Yang identification model, and obtain and output the Yin-Yang deficiency syndrome detection report.

[0029] Specifically, when executing step S4-1, the gas characteristic molecular information specifically includes various characteristic substances: amino acids include cysteine, n-amylglycine, aminocaprylic acid, n1-levulinic acid hydrochloride, γ-glutamyl-2-aminobutyric acid, tyramine glucosyl linolenate, docosapentaenoic acid, docosenoic acid, erucamide, glycerol palmitate, and adrenaline acid; steroids and hormones include deoxycorticosterone and adrenaline glucuronic acid; organic compounds and acids include methylthiourea, 4-nitrobenzoic acid, lauramide, phosphocitric acid, 2-octenoyl carnitine, linolenic acid, and 1,3-bis(4-nitrophenyl)urea; other organic compounds include creatine ethyl ester, phenanthrol, sphingosine, and tributyrin.

[0030] Furthermore, Table 1 shows the correlation between these characteristic substances and Yin-Yang deficiency syndrome, which can be used to judge Yin-Yang deficiency syndrome in the human body. Table 1 is as follows: Table 1. Mass-to-charge ratio and relative molecular mass data of characteristic substances

[0031] , Furthermore, the yin and yang properties of the disease are manifestations of human pathology, which can lead to metabolic abnormalities. The types and degrees of abnormal manifestations can be analyzed and judged through metabolic exhaled exhaled products. The exhaled gas in the human body carries a large amount of biological material information, which is a new "window" for people to understand the physiology and pathology of the body. The thousands of VOCs (volatile organic compounds) contained in the exhaled breath of the human body are the products of human metabolism and exposure to exogenous compounds. The endogenous VOCs produced by human metabolism are the media of respiratory metabolomics and can be used as biomarkers for human health status assessment and disease diagnosis. They have great potential value in clinical applications. For example, substances such as acetaldehyde, acetone, and formaldehyde in exhaled gas are associated with diseases such as diabetes, liver disease, and cancer. At the same time, the smell and taste in exhaled gas may also be related to certain diseases, such as bad breath and odor may be associated with digestive system diseases, oral diseases, etc.

[0032] Specifically, when executing step S4-2, an optimal hyperplane is constructed based on the parameters and support vectors of the Yin-Yang model, and the characteristic molecular information of the subject's exhaled gas is classified into two categories based on the optimal hyperplane to obtain the subject's Yin-Yang deficiency information, and a Yin-Yang deficiency detection report is prepared and output.

[0033] The optimal hyperplane is a straight line in two-dimensional space, a plane in three-dimensional space, and a hyperplane in higher-dimensional space. The optimal hyperplane separates sample points of different categories as much as possible and maximizes the distance to the nearest sample point (i.e., support vector).

[0034] Furthermore, the distance from the subject sample to the optimal hyperplane is calculated, and the category to which the sample belongs is determined based on the positive or negative value of the distance. If the sample distance is negative, it belongs to Yin deficiency syndrome, and if the sample distance is positive, it belongs to Yang deficiency syndrome.

[0035] In other embodiments, the molecular components of exhaled gas are detected by EESI-MS in situ mass spectrometry, and high-precision and high-sensitivity qualitative / quantitative detection of molecular components is achieved by Q-LIT-TOF multi-mass analyzer hybrid tandem technology; cloud database technology is used to build an open and shared human exhaled gas sample information database with services such as data archiving, knowledge search, data management, data calculation and data services; artificial intelligence algorithms such as incremental learning frameworks are used to identify the correlation between the yin and yang properties of traditional Chinese medicine diseases (taking the stable stage of chronic obstructive pulmonary disease and yin / yang deficiency syndrome of stable angina pectoris in coronary heart disease as examples) and molecular markers in exhaled gas, and the parameters of the yin and yang identification model are iteratively optimized.

[0036] Furthermore, the specific manifestations of Yin / Yang deficiency in the stable stage of COPD are: Yin deficiency is manifested by dry cough, little sputum, dry mouth and throat, hot hands and soles, and night sweats; Yang deficiency is manifested by fear of cold and cold limbs, weak cough, clear and thin sputum, shortness of breath and laziness to speak, fatigue and weakness; The specific manifestations of Yin / Yang deficiency in stable angina pectoris of coronary heart disease are: Yin deficiency is manifested by chest tightness and dull pain, which is aggravated by exertion, shortness of breath and dry mouth, palpitations and fatigue, dizziness and insomnia, spontaneous sweating and night sweats; Yang deficiency is manifested by palpitations and pain, chest tightness and shortness of breath, which is aggravated by movement, spontaneous sweating, fatigue and fear of cold, and cold limbs.

[0037] Furthermore, the exhaled gas of patients in the stable stage of COPD and patients with stable angina pectoris due to coronary heart disease was collected for yin / yang classification labeling, and an optimal hyperplane was constructed based on the parameters and support vectors of the yin-yang identification model. Based on the optimal hyperplane, the characteristic molecular information of the subjects' exhaled gas was binary classified to obtain the subjects' yin-yang deficiency detection report.

[0038] The embodiment of the present invention also provides a non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome, which specifically includes: A sample collection module for collecting exhaled gas samples containing life substances carrying physiological or pathological information; A mass spectrometry analysis module analyzes the exhaled gas sample based on electrospray extraction ionization mass spectrometry, and marks characteristic molecular information of the exhaled gas to obtain a positive and negative data set; A model building module, for building an initial model by combining random forest and gradient boosting decision tree algorithms, training and testing the initial model based on the yin-yang data set to obtain a yin-yang identification model, and associating the yin-yang attributes of TCM diseases with characteristic molecular information of exhaled breath; The real-time detection module is used to obtain the real-time exhaled gas of the subject and detect the characteristic molecular information of the exhaled gas, identify the Yin-Yang deficiency syndrome based on the Yin-Yang identification model, and obtain and output the Yin-Yang deficiency syndrome detection report.

[0039] In some embodiments, the mass spectrometry module includes a differential vacuum system, an EESI ionizer, an ion introducer, an ion transporter, a hybrid tandem mass analyzer, and an ion detector; The differential vacuum system and the EESI ionizer for analog and digital conversion of the collected data use electrospray ionization technology to introduce the exhaled gas sample and convert it into ions. The ion introducer includes a mechanical pump with a pumping speed of 16L / s and operates at a pressure of 200Pa to maintain the vacuum environment and extract ions from the EESI ionizer. The ion transporter uses a molecular pump with a pumping speed of 80L / s and operates at a pressure of 0.2Pa to move the ions to the hybrid tandem mass analyzer in the differential vacuum system.

[0040] Specifically, the hybrid tandem mass analyzer includes a quadrupole mass selector, a linear ion trap mass analyzer and a time-of-flight mass analyzer. The hybrid tandem mass analyzer separates and analyzes ions according to their mass-to-charge ratio, and transports ions in a time-of-flight mass analyzer including a molecular pump. The pump speed is 300 liters / second and operates at a pressure of 0.0005 pa. The time-of-flight mass analyzer is used to determine the mass of the ions according to the flight time of the ions through the drift region. The ions that have undergone mass analysis in the differential vacuum system are ultimately guided to an ion detector, which collects analog ions and converts the collected data into a digital format, records the number of ions detected, and marks a timestamp for each ion event.

[0041] In fact, the yin-yang identification model in the model building module includes random forest, gradient boosting decision tree algorithm and support vector machine algorithm. The random forest uses integrated learning to improve classification accuracy, the gradient boosting decision tree adds decision trees in sequence to optimize prediction performance, and the support vector machine constructs a hyperplane in a high-dimensional space to identify yin-yang deficiency syndromes.

[0042] Furthermore, the differential vacuum system adopts a four-stage differential vacuum structure, wherein: The first-stage vacuum chamber is used to place the ion transfer tube of the atmospheric pressure ion introduction interface; the second-stage vacuum is used to place the ion transmitter; the third-stage vacuum is used to place the quadrupole mass selector and ion trap mass analyzer; and the fourth-stage vacuum is used to place the time-of-flight mass analyzer.

[0043] It should be noted that the first-stage vacuum of the differential vacuum system is provided by a mechanical pump with a total pumping speed of better than 90 liters / hour, which is used to form a vacuum environment of <200 Pa at the rear end of the ion transfer tube and provide a fore-stage vacuum for the molecular pump. This project intends to use three mechanical pumps with a pumping speed of 30 liters / hour, or a scheme of using multiple mechanical pumps in parallel with one 60 liters / hour and one 30 liters / hour. The adoption of this scheme can greatly improve the fault tolerance of the vacuum system; when a mechanical pump fails, it will not directly cause the entire vacuum system to shut down, and at the same time, the faulty pump can be repaired without destroying the vacuum. The mechanical pump of this module will be the main source of noise, vibration, heat, and exhaust gas for the entire set of equipment. Therefore, this module intends to land the mechanical pump and arrange a reasonable space, and add components such as noise reduction, ventilation and heat dissipation, and exhaust gas collection and treatment to increase the applicability of the overall assembly.

[0044] Furthermore, the second, third and fourth stage vacuum are provided by three molecular pumps with different pumping speeds. The pumping speeds are planned to be 700 L / s, 300 L / s and 300 L / s, respectively, to obtain 0.2 Pa, 5× Pa and 5× This module intends to adopt a vacuum acquisition solution with a high pumping speed, which can increase the aperture of the vias between each level of vacuum, which is beneficial to reduce the loss of ions during ion crossing and improve the transmission efficiency and detection sensitivity.

[0045] Furthermore, the atmospheric pressure ion introduction interface is planned to adopt the ion transfer tube solution. The ion transfer tube is made of seamless 316L stainless steel with an inner diameter of 0.58mm, and the inside and outside of the tube are polished to improve the anti-pollution ability of the transfer tube. At the same time, the structure also needs to heat the ion transfer tube, using a heating rod and a temperature sensor to form a temperature control heating component, with a heating power better than 80W, and an adjustable working temperature between 150℃ and 350℃, to improve the ion transmission efficiency and reduce the residue of the test object.

[0046] Specifically, the ion detector includes a lateral excitation dual CEM detector of LIT and a time-to-digital conversion detector of TOF; The lateral excitation dual CEM detector of the LIT amplifies and detects the signal of the ion current through a low-bandwidth amplifier to obtain the ion abundance information of the corresponding mass number; The TOF time-to-digital conversion detector is used to amplify the signal of a single ion and convert it into a voltage pulse signal.

[0047] In fact, the LIT's side-excitation dual CEM detector amplifies and detects the signal of the ion current through a low-bandwidth amplifier, thereby obtaining the ion abundance information of the corresponding mass number. In this embodiment, a dynode is installed on each CEM detector. The dynode is made of stainless steel processed into a bowl shape and mirror-polished. It is used to attract all the ions thrown out from the LIT slit and at the same time play a certain current amplification role (releasing electrons with at least twice the charge). The electron flow emitted by the dynode is repelled into the CEM and amplified (about The two anodes collect the current signals released by the two CEMs respectively, and then insert them into the preamplifier (which has been out of vacuum) after being connected in parallel. The preamplifier to be used in this project has an amplification factor of up to V / A. The voltage signal output by the preamplifier is collected by a high-speed analog digital acquisition card (ADC) and can be uploaded to a computer and converted into spectrum data. According to this scheme, when the ion trap is scanned, 500 mass numbers are scanned in 0.5s, and the data waveform can achieve unit mass resolution, the peak width of each mass number ion is 1ms. Therefore, when 10 ions of a certain mass number are thrown out of the trap at the same time, they will hit the dynode to obtain at least 20 electrons, and then obtain CEM. Gain, get 2× electrons; converted to current, it is 3.2nA. ​​After amplification by the preamplifier, a 320mV pulse signal can be obtained. This level of signal can be easily detected by the ADC acquisition card.

[0048] Furthermore, the above calculation results show that under the detector design of this module, the detection capability of the ion trap mass analyzer itself is close to that of the single ion signal. Therefore, the sensitivity of the low-resolution LIT detection results of this module depends on the number of ions sent into the trap and the excitation efficiency of the ion trap itself. This provides the most basic data guidance for instrument performance debugging and the ultimate achievement of high-sensitivity detection goals. In the specific implementation process, the dynamic range of the sample concentration detected by this module can also be optimized and adjusted by adjusting the magnification of the pre-signal amplifier to meet the actual detection needs.

[0049] In fact, the time-to-digital conversion detector of TOF usually uses a double-layer microchannel plate (MCP) to form an ion current amplifier, and can gain the signal of a single ion and convert it into a voltage pulse signal. The pulse width of this signal is usually less than 10ns. Therefore, after the signal itself is amplified by a high-bandwidth amplifier, a voltage pulse better than 100mV can be obtained. This module intends to adopt the TDC data acquisition scheme, that is: first use a constant-fraction discriminator to process the above pulse signal, convert the pulse signal into a square wave signal that can be used for digital counting; and align the rising edge of the square wave with the extreme value of the pulse signal. This operation is to obtain higher time counting accuracy and spectral resolution. Then, the above square wave signal is sent to the time digital acquisition card dedicated to the time-of-flight mass spectrometer. Through the time-related counting algorithm, the flight time taken for each ion to reach the detector is recorded, and after multiple flight statistics, the abundance information of the corresponding mass number can be obtained.

[0050] It should be noted that the electrospray enabled solute-solute-solvent ionizer (EESI) is used to introduce the exhaled breath sample and convert it into ions using the electrospray ionization technique. The gas in the ionized state is transported in the ion introducer. The ion introducer includes a mechanical pump with a pumping speed of 16L / s, operating at a pressure of 200Pa, which is used to maintain the vacuum environment and extract ions from the EESI ionizer. Then, the ion transporter uses a molecular pump with a pumping speed of 80L / s, operating at a pressure of 0.2Pa, which helps the ions move to the hybrid tandem mass analyzer in the vacuum system. The hybrid tandem mass analyzer assembly contains a quadrupole mass selector, an ion trap mass analyzer, and a time-of-flight mass analyzer. It includes a molecular pump with a pumping speed of 300L / s, operating at a pressure of 0.005Pa. The hybrid tandem mass analyzer is responsible for separating and analyzing ions based on their mass-to-charge ratio. After that, the ions are transported in the time-of-flight mass analyzer including a molecular pump with a pumping speed of 300L / s and operating at a pressure of 0.0005Pa. A time-of-flight mass analyzer can be used to determine the mass of ions based on their flight time through the drift region. Ions that have undergone mass analysis in a differential vacuum system are ultimately directed to an ion detector. This ion detector allows for analog ion collection and conversion of the collected data into a digital format. It records the number of ions detected and assigns a specific time to each ion event to allow for further analysis and interpretation of the collected data.

[0051] Although the embodiments of the present invention are described in detail above, it is obvious to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as described in the claims. Moreover, the present invention described herein may have other embodiments and may be implemented or realized in a variety of ways.

Claims

1. A non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome, characterized in that: include: Collect exhaled breath samples containing vital substances that carry physiological or pathological information; Analyzing the exhaled gas sample based on electrospray extraction ionization mass spectrometry, marking characteristic molecular information of the exhaled gas to obtain a yin-yang data set; Combining random forest and gradient boosting decision tree algorithms to build an initial model, training and testing the initial model based on the Yin-Yang dataset to obtain a Yin-Yang identification model, and correlating the Yin-Yang attributes of TCM diseases with characteristic molecular information of exhaled breath; Acquire the real-time exhaled gas of the subjects and detect the characteristic molecular information of the exhaled gas. Construct a hyperplane based on the Yin-Yang identification model to classify the gas characteristic molecular information into two categories, obtain the Yin-Yang deficiency syndrome information of the subjects, and produce and output the Yin-Yang deficiency syndrome detection report.

2. The non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome according to claim 1, characterized in that: The exhaled gas sample containing life substances carrying physiological or pathological information is collected, wherein the exhaled gas sample includes amino acids and derivatives, fatty acids and lipids, steroids and hormones, organic compounds and acids, and organic compounds.

3. The non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome according to claim 1, characterized in that: The exhaled gas sample is analyzed based on electrospray extraction ionization mass spectrometry, and characteristic molecular information of the exhaled gas is marked to obtain a yin-yang data set, including: ionizing the exhaled gas sample based on electrospray extraction ionization mass spectrometry to obtain gas ions; Maintaining the vacuum environment, absorbing the gas ions and pressurizing them to perform mixed tandem mass analysis to obtain a mass-to-charge ratio data set and a mass analysis data set; The yin-yang dataset was obtained by fusing the mass-to-charge ratio dataset and the mass analysis dataset and unifying their formats, marking the timestamps and the characteristic molecular information of the exhaled gas.

4. The non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome according to claim 1, characterized in that: The yin-yang identification model is obtained by training and testing the initial model based on the yin-yang data set, including: Initialize the parameters of the initial model; The yin-yang data set is randomly divided into a training set and a test set, the training set is input into the initial model to obtain a predicted value, the loss error between the predicted value and the true value is calculated, a loss function is constructed, the loss function is minimized based on the gradient descent method, the initial model parameters are iteratively updated until the loss function value tends to be stable or the maximum number of iterations is reached to terminate the training, and the initial model is evaluated and fine-tuned based on the test set to obtain a yin-yang identification model.

5. A non-invasive real-time detection method for characteristic substances of human yin and yang deficiency syndrome according to claim 4, characterized in that: Based on the yin-yang identification model, a hyperplane is constructed to classify the gas characteristic molecular information into two categories, including: Construct a hyperplane based on the parameters and support vectors of the Yin-Yang identification model; Calculate the distance from the subject sample to the hyperplane, and determine the category to which the sample belongs based on the positive or negative value of the distance. If the sample distance is negative, it belongs to Yin deficiency syndrome, and if the sample distance is positive, it belongs to Yang deficiency syndrome.

6. A non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome, characterized in that: include: A sample collection module for collecting exhaled gas samples containing life substances carrying physiological or pathological information; A mass spectrometry analysis module analyzes the exhaled gas sample based on electrospray extraction ionization mass spectrometry, and marks characteristic molecular information of the exhaled gas to obtain a positive and negative data set; A model building module, for building an initial model by combining random forest and gradient boosting decision tree algorithms, training and testing the initial model based on the yin-yang data set to obtain a yin-yang identification model, and associating the yin-yang attributes of TCM diseases with characteristic molecular information of exhaled breath; The real-time detection module is used to obtain the real-time exhaled gas of the subject and detect the characteristic molecular information of the exhaled gas, identify the Yin-Yang deficiency syndrome based on the Yin-Yang identification model, and obtain and output the Yin-Yang deficiency syndrome detection report.

7. The non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome according to claim 6, characterized in that: The mass spectrometry module includes a differential vacuum system, an EESI ionizer, an ion introducer, an ion transmitter, a hybrid tandem mass analyzer, and an ion detector; The differential vacuum system and the EESI ionizer for analog and digital conversion of the collected data use electrospray ionization technology to introduce the exhaled gas sample and convert it into ions. The ion introducer includes a mechanical pump with a pumping speed of 16L / s and operates at a pressure of 200Pa to maintain the vacuum environment and extract ions from the EESI ionizer. The ion transporter uses a molecular pump with a pumping speed of 80L / s and operates at a pressure of 0.2Pa to move the ions to the hybrid tandem mass analyzer in the differential vacuum system.

8. The non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome according to claim 7 is characterized in that: The hybrid tandem mass analyzer comprises a quadrupole mass selector, a linear ion trap mass analyzer and a time-of-flight mass analyzer. The hybrid tandem mass analyzer separates and analyzes ions according to their mass-to-charge ratio. The ions are transported in the time-of-flight mass analyzer including a molecular pump. The pump speed is 300 liters / second and the analysis is performed at a pressure of 0.0005 Pa. The time-of-flight mass analyzer is used to determine the mass of the ions according to the flight time of the ions through the drift region. The ions that have undergone mass analysis in the differential vacuum system are ultimately directed to an ion detector, which collects analog ions and converts the collected data into a digital format, records the number of ions detected, and marks a timestamp for each ion event.

9. The non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome according to claim 8, characterized in that: The yin-yang identification model in the model building module includes a random forest, a gradient boosting decision tree algorithm and a support vector machine algorithm. The random forest uses ensemble learning to improve classification accuracy, the gradient boosting decision tree adds decision trees in sequence to optimize prediction performance, and the support vector machine constructs a hyperplane in a high-dimensional space to perform yin-yang deficiency syndrome identification.

10. The non-invasive real-time detection system for characteristic substances of human yin and yang deficiency syndrome according to claim 9, characterized in that: The differential vacuum system adopts a four-stage differential vacuum structure, wherein: An ion transfer tube for placing a normal pressure ion introduction interface in the first-stage vacuum chamber; The second stage of vacuum is used to place the ion transporter; The third vacuum stage is used to place the quadrupole mass selector and ion trap mass analyzer; The fourth vacuum stage is used to house the time-of-flight mass analyzer.

Images