CN111696679A

CN111696679A - Human health state analysis scheme based on quantum Petri net

Info

Publication number: CN111696679A
Application number: CN202010677406.9A
Authority: CN
Inventors: 张盛; 章越新
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-07-14
Filing date: 2020-07-14
Publication date: 2020-09-22

Abstract

The invention provides a human health state analysis scheme based on a quantum Petri network, which integrates machine learning, quantum informatics and traditional Chinese medicine theories to realize intelligent analysis and prediction of human health states. The method comprises the steps of establishing association between human viscera and health data by using a traditional Chinese medicine theory, establishing a health state disturbance equation by using quantum bits as a data processing carrier, and finally obtaining a health state analysis result by using Petri network calculation, so that possible illness conditions of a human body in a period of time in the future are predicted.

Description

Human health state analysis scheme based on quantum Petri net

Technical Field

The invention relates to a human health state analysis scheme based on a quantum Petri network, and belongs to the field of human health state intelligent analysis and prediction.

Background

In the field of medical research, disease prediction models are often used to predict future development of a disease. Specifically, the risk that a certain group of people or individuals may suffer from a certain disease in a future period of time is predicted by researching the medical mechanism of the certain disease and selecting a proper mathematical tool for modeling analysis.

Traditional disease Prediction is mainly achieved by Clinical Prediction tools/models (CPRs). For example, Benjamin et al proposed a demographic-based Framingham method in the literature [ Impact of atrial fibrillation on the riskof death: the Framingham Heart Study ] to Study the effect of atrial fibrillation on mortality. However, the method needs to track a large number of experimental objects, the follow-up time is dozens of years, the cost is high, the efficiency is low, and the experimental objects have uncertainty and are difficult to be widely applied.

In response to the above problems, Biehl et al have proposed a method of determining the level of intensive care using an advanced-person risk assessment (ERA) in the literature [ Prediction of clinical assessment of patient out of patient usage estimation: a probability-based study ]. The method establishes an ERA scoring standard from daily management data recorded in an electronic medical record of a subject, and effectively improves the prediction efficiency. However, the method mainly aims at evaluating and predicting high-risk serious diseases, and neglects analysis of common diseases.

On the other hand, the prediction accuracy based on the CPRs model is difficult to verify, and the recognition degree of the same row is low. Grady et al in the literature [ what is a good clinical prediction rule so hard to find? Defects of the CPRs model are indicated: the CPR model is good in modeling, but after the actual application scene is changed, the prediction effect of the model is greatly reduced because the physical condition and the disease severity of a patient are mostly different from those of a modeling sample.

At the same time, the study by Siontis et al showed that only 10 of the 118 CPRs were studied in different populations 4 times and more, while only one of the 10 tools gave consistent predictions in different databases.

Because the traditional disease prediction schemes all depend on the medical mechanism of the disease, the specialization is strong and the application range is limited. In order to solve the above problems, in recent years, disease prediction schemes based on artificial intelligence techniques such as machine learning and deep learning have been proposed in succession, so that disease prediction is more intelligent, accurate and efficient. For example, Qingyu Zhao et al propose a general regression model based on a variational self-coding framework, which is mainly applied to the brain aging prediction problem of structured nuclear magnetic images. But the model can not predict the future morbidity risk of other similar patients according to the diagnosis records of the patients.

Miranda et al, in the literature [ induced prediction of the bone surface at the first viral pathological MRI patient ] propose a whole brain classification method using a Support Vector Machine (SVM) to achieve future course prediction at the individual level from MRI data obtained at the first psychotic episode of a patient. However, the method must predict the disease according to the existing diseased condition of the patient, and cannot analyze and prevent the disease before the disease.

In view of the above-mentioned drawbacks, Nguyen et al propose a new end-to-end Deep learning system in the document [ Deep: A volumetric net for media records ], which can extract features from medical records, automatically predict future risks, and realize analysis of various diseases. However, according to the scheme, the disease prediction is realized only by extracting the characteristics of the historical medical records of the user, and the multi-cause analysis of the disease cannot be realized.

In addition, the disease prediction scheme based on the artificial intelligence search engine is widely applied to practical scenes, for example, various medical search engines such as 360-degree medical search, dog-searching medical search, Baidu thumb doctor, clove park search and the like can search according to entity words containing clinical symptoms in search information input by a user, and statistically weight extracted entity words to provide disease prediction information. However, the weight of the search engine has strong subjective knowledge, and the result of the search engine depends on subjective cognition of medical staff on diseases, so that the deviation between the final prediction result and the actual clinical symptoms of the user is large.

Although the existing disease prediction scheme based on the artificial intelligence technology solves the limitation problem of the traditional scheme, the traditional scheme still has the defect of being not negligible. Firstly, the data analysis of diseases is emphasized, so that the key medical causes of the diseases are easy to ignore, and the prediction precision is difficult to achieve expectation. Secondly, the computing power requirement on the equipment is relatively high, which causes the corresponding increase of the cost, thereby being difficult to be widely applied to a low computing power platform. Thirdly, the method focuses on historical disease record analysis, so that the real-time performance is not high.

Disclosure of Invention

In order to overcome the defects of the existing scheme and meet the development trend of the future disease prediction application field, the invention provides a human health state analysis scheme based on a quantum Petri network, and intelligent analysis and prediction of the human health state are realized by combining machine learning, quantum informatics and traditional Chinese medicine theories.

In order to achieve the above purpose, the scheme of the invention specifically comprises the following steps:

s 1: acquiring health information related to human viscera and carrying out preprocessing operation on the health information to obtain preprocessing data corresponding to the health information;

s 2: the human viscera are equivalent to qubits, and a health state quantum system is established according to the qubits;

s 3: performing regression processing on the preprocessed data to obtain a perturbation operator corresponding to the quantum bit;

s 4: acting the perturbation operator on the qubit to obtain a quantum system perturbation equation;

s 5: constructing and operating a human health state Petri net according to a quantum system disturbance equation to obtain a probability amplitude corresponding to the eigenstate of a health state quantum system;

s 6: and obtaining the analysis result of the health state of the human body according to the probability amplitude corresponding to the eigen state, namely possible diseased species and diseased probability.

According to the health information related to the human viscera obtained in the step s1, preprocessing operation is performed on the health information to obtain preprocessing data corresponding to the health information, and generally speaking, most of application requirements can be met by selecting the five viscera and six viscera as the human viscera.

The health information refers to various factors that can reflect or have strong influence on the health status of the viscera, including but not limited to environmental information and physiological information.

According to the s2, the human viscera are equivalent to qubits, and a health state quantum system is established according to the qubits, wherein two basic vectors of the qubits correspond to two health states of the human viscera, namely 'normal' and 'abnormal', and are respectively represented by Dirac symbols |0 > and |1 >; the state-of-health quantum system is a composite quantum system spanned by qubits.

Performing regression processing on the preprocessed data according to the s3 to obtain a perturbation operator corresponding to the qubit, wherein the regression processing mode includes: machine learning, least square method, interpolation, fitting and the like. Machine learning algorithms are generally preferred because of their greater degree of intelligence.

And according to the s4, applying perturbation operators to the qubits to obtain a quantum system perturbation equation, wherein the perturbation operators have single mapping and only act on corresponding qubits.

Constructing and operating a human health state Petri net according to the quantum system disturbance equation in the s5 to obtain a probability amplitude corresponding to the eigen state of the health state quantum system, wherein the human health state Petri net is described as follows:

the device comprises a perturbation module, a quantum bit module and an eigen-state module;

the disturbance module is used for representing the change condition of the quantum system under the environmental disturbance;

the qubit module is used for describing qubits and state change conditions caused by environmental disturbance;

the eigen-state module is used for describing all eigen states and state change conditions thereof which form the health state quantum system;

each perturbation operator in the perturbation module acts on the corresponding qubit in the qubit module respectively, so that the state of the qubit is changed;

the state of the qubit in the qubit module determines the probability amplitude corresponding to the eigen-states in the eigen-state module.

Obtaining a human health state analysis result, namely a possible disease type and a disease probability according to the probability amplitude corresponding to the eigen state in s6, wherein:

calculating probability values corresponding to the eigenstates according to the probability amplitude corresponding to the eigenstates, sequencing the probability values in an ascending mode to obtain the eigenstates with the maximum probability values, wherein all abnormal bits in the eigenstates are used for describing the diseased species;

and calculating to obtain probability values corresponding to all the 'abnormal' eigenstates, summing the probability values, and taking the summed result as a diseased probability value.

Drawings

Fig. 1 is a schematic diagram of a human health state analysis scheme based on a quantum Petri net according to embodiment a.

Fig. 2 is a schematic diagram of a Petri net in human health status according to an embodiment a.

Detailed Description

The following is a specific embodiment of the present invention, and the technical solution of the present invention is further described with reference to the accompanying drawings. It should be noted that the present embodiment has fully illustrated the principle of the solution of the present invention, but the present invention is not limited to the present embodiment.

Example A

The embodiment provides a human health state analysis scheme based on a quantum Petri network, as shown in FIG. 1, which includes the steps of:

s 11: acquiring health information related to human viscera and carrying out preprocessing operation on the health information to obtain preprocessing data corresponding to the health information;

s 12: the human viscera are equivalent to qubits, and a health state quantum system is established according to the qubits;

s 13: performing regression processing on the preprocessed data to obtain a perturbation operator corresponding to the quantum bit;

s 14: acting the perturbation operator on the qubit to obtain a quantum system perturbation equation;

s 15: constructing and operating a human health state Petri net according to a quantum system disturbance equation to obtain a probability amplitude corresponding to the eigenstate of a health state quantum system;

s 16: and obtaining the analysis result of the health state of the human body according to the probability amplitude corresponding to the eigen state, namely possible diseased species and diseased probability.

Implementation of s11 in this example. The health information related to the zang-fu organs of the human body is defined as follows:

the heart rate information is health information associated with the heart, and is due to: the health status of the heart can be directly reflected by the heart rate;

the reason why the sleep information is health information related to the liver is that: the theory of traditional Chinese medicine holds that the liver belongs to wood in five elements, stores blood mainly, has the functions of storing and regulating blood, and the health state of the liver is directly reflected in sleep information;

the reason why the exercise information is the health information associated with the spleen is that: the theory of traditional Chinese medicine considers that the spleen belongs to earth in five elements, mainly transports and transforms food essence into essence, qi, blood and body fluid, and distributes the essence, qi, blood and body fluid to the whole body, and the health state of the spleen is closely related to exercise information;

the reason why the temperature information is health information related to the lung is that: the traditional Chinese medicine theory considers that the lung belongs to metal in five elements, is mainly descending and has the function of defending against invasion of exogenous pathogens, and the temperature information reflects the invasion intensity of the exogenous pathogens to a certain extent, so that the temperature information is selected as a basis for predicting the health condition of the lung;

the walking information is health information related to the kidney, and is caused by: the theory of traditional Chinese medicine considers that the kidney belongs to water in five elements, is the little yin in yin, stores essence and runs through Yongquan, and the walking information is closely related to the state of Yongquan point.

It should be noted that the selection criteria of the health information is not unique, but must conform to medical theory.

It should be understood that the sampling interval of the health information must be earlier than the health state analysis waiting interval.

In this embodiment, the acquisition of the preprocessed data includes the following two steps:

step 1: sampling data to obtain the sampling data d of the health information_X(t_i,j) Wherein t is_i,jIs the ith sampling time point in the jth sampling period; the sampling interval is a small observation period of the health information, the unit is second, minute, hour, day, month, year and the like, and the sampling interval can be determined according to the actual application condition; the sampling time interval is the overview of the health informationMeasuring period, the value of which is generally larger than sampling interval; let the unit of the sampling interval be day, the unit of the sampling period be month, and the sampling result be expressed as follows:

d_H(t_i,j) Is t_i,jHeart rate value at the moment in units of times/minute;

d_ST(t_i,j) Is t_i,jThe value of the time of falling asleep is accurate to the hour, and the unit is the hour;

d_SL(t_i,j) Is t_i,jThe value of the sleep duration at a moment is rounded off in the unit of hour;

d_EF(t_i,j) Is t_i,jThe motion frequency order value of the moment;

d_EI(t_i,j) Is t_i,jThe exercise intensity value at the moment is in units of minutes/time;

d_T(t_i,j) Is t_i,jThe average temperature value at the moment is expressed in units of;

d_WN(t_i,j) Is t_i,jStep number of the moment;

d_WF(t_i,j) Is t_i,jThe step frequency value of the moment is in the unit of meter/second;

step 2: for d above_X(t_i,j) Preprocessing to obtain the jth sampling period t_jIntermediate preprocessing data D of_X(t_j) To D, pair_X(t_j) Carrying out normalization processing to obtain final preprocessing data D_{N_X}(t_j)。

In this embodiment, the heart rate information, the sleep information, the exercise information, the temperature information, and the walking information correspond to D_X(t_j) Are respectively marked as D_H(t_j)、D_S(t_j)、D_E(t_j)、D_T(t_j)、D_W(t_j) Corresponding to D_{N_X}(t_j) Are respectively marked as D_{N_H}(t_j)、D_{N_S}(t_j)、D_{N_E}(t_j)、D_{N_T}(t_j)、D_{N_W}(t_j)。

In the step 2, the algorithm of the preprocessing is determined according to the actual application scenario, but should conform to the medical theory. In this embodiment, in order to facilitate the acquisition of health information and subsequent experimental verification, the following preprocessing algorithm is specifically selected:

wherein n represents t_jThe number of internal sampling time points, in this example, n is the total number of days per month;

wherein t is_{max_j}Is D_X(t_j) And taking the corresponding sampling period when the maximum value is obtained.

Implementation of s12 in this example. Order to

Qubits corresponding to the human zang-fu, which are generally expressed as:

wherein, | α_X|²+|β_X|²＝1，|0＞_XRepresents XThe corresponding zang-fu organs are in "normal" state, correspondingly, |1 >_XAn "abnormal" state is represented. In this example, take

Is the initial probability amplitude. Specifically, the quanta of the human five zang organs are expressed as follows:

heart:

liver:

spleen:

lung:

kidney:

make the state of health quantum system | phi_BIs an arbitrary vector in Hilbert space, i.e.

Wherein the content of the first and second substances,

representing a two-dimensional hilbert space. This gives:

implementation of s13 in this example. By using

To represent

Corresponding perturbation operator in the form of matrix

Wherein F_X(t) and G_X(t) are each independently

The matrix elements of (1).

Act on

Can be expressed as

Wherein:

|ψ_X(t)＞＝F_X(t)|0＞_X+G_X(t)|1＞_Xand | F_X(t)|²+|G_X(t)|²＝1。

In this embodiment, in order to reduce the calculation cost for the subsequent experimental verification, the least square method is adopted to perform the D pair_{N_X}(t_j) Performing regression to obtain

Corresponding F_X(t), regression was achieved by python programming.

Implementation of s14 in this example. It is known that

Has single mapping property, so the quantum system disturbance equation Eq (| phi) can be obtained_B>) general representation:

in other words, there are:

wherein P is_m(t) is time | a_mThe corresponding probability amplitude.

Implementation of s15 in this example. Establishing a Petri net of the human health state, and marking as P (| phi)_B>) the Petri Net was implemented using visual object Net + + software, as detailed in fig. 2.

P(|Φ_B>) the modules, libraries, and transitions are defined as follows:

m [ d ] is a disturbance module; m [ s ] is a qubit module; m [ e ] is an eigenmode module;

m [ d1], M [ d2], M [ d3], M [ d4] and M [ d5] are disturbance units in M [ d ], respectively;

m [ s1], M [ s2], M [ s3], M [ s4] and M [ s5] are quantum bit units in M [ s ], respectively;

P_His a perturbation operator library, P_XFIs a library of perturbation elements, P_XIs a qubit depot, P_X0Is a "normal" state library, P_X1Is an "abnormal" state repository, P_BIs a health State Quantum System library, P_aIs an eigenstate library;

T_XHfor disturbance transitions, T_X0For "normal" state transitions, T_X1For "abnormal" state transitions, T_aFor transition of eigenstates, T_tenFor tensor transitions, T_recRecording transitions for data;

M[d]the middle disturbance unit has a composition of T_XH、P_XF；

M[s]The medium quantum bit unit is composed of P_X、T_X0、T_X1、P_X0、P_X1；

M[e]The medium eigenstate unit is composed of P_a、P_B、T_ten、T_a、T_rec；

And the library places and the transitions carry out state migration through directed arcs.

P(|Φ_B>) librariesThe Token values indicate:

P_Hand P_XCorresponding to the number of predicted time points, P, in a future period of time_XFAnd P_X0Having a Token value of F_X(t)，P_X1Having a Token value of G_X(t)，P_aToken value of_m(t)|²，P_BHas a Token value of 32;

initial time, P_XF、P_X0、P_X1、P_aAnd P_BThe Token values of (A) are all 0, P_HAnd P_XThe value of Token of (2) is preset.

Next, the constructed Petri Net P (| Φ) is run_B>) to obtain the health status of the human body in a future period of time.

Definition of t_kFor the kth predicted time point within a future period of time, in particular:

p 1: at t_kTime, M [ d]P in (1)_H Outputting 1 Token, passing through M [ d ] respectively]So that P is_XFOccurrence of a status response, P_XFIs changed to F_X(t_k)；

p 2: when P is present_XFAfter the status response occurs, M [ s ]]P in (1)_XRespectively output 1 Token so that P_X0And P_X1Occurrence of a status response, P_X0And P_X1Respectively become | F_X(t_k)|²And | G_X(t_k)|²；

p3：P_X0And P_X1Will output the current Token value, so that P_BOccurrence of a status response, P_BThe Token value of (a) becomes 32;

p4：P_Bwill output the current Token value, so that P_aOccurrence of a status response, P_aIs changed to | P_m(t_k)|²；

p5：P_aWill output the current Token value, all | P_m(t_k)|²Will be recorded, then P (| Φ)_B>) transfer into t_k+1And at the moment, repeating the steps till the end.

It should be understood that fig. 2 is only a diagram of the human health state Petri net structure provided in the present embodiment, and does not limit any other descriptions following the human health state Petri net described in the solution of the present invention.

Implementation of s16 in this example. To | P_m(t_k)|²Sorting in an ascending manner to obtain the maximum value max { | P_m(t_k)|²And its corresponding eigen state max { | a_m> -) and according to max { | a_mThe "abnormal" bit (value is "1") in > "judges the diseased species. At the same time for t_kAll | a's at time are "abnormal_mProbability summation is carried out to obtain the value of the sick probability which is marked as P_X(t_k). In other words, there is P_X(t_k)＝∑_m|P_m(t_k)|²。

And verifying the reliability of the scheme. In this embodiment, a user is randomly selected as an experimenter, a health state analysis result of the experimenter, that is, a possible disease type and a disease probability, is obtained according to the steps, and consistency analysis is performed on the result and a disease record of the experimenter.

Let experimenter number a1, its basic information is: in men, the disease is 26 years old, gastroenteritis is diagnosed in 12 months in 2018, chronic bronchitis is diagnosed in 8 months in 2019, gout is diagnosed in 4 months in 2020, and Harbin city, Heilongjiang province, is common.

In this embodiment, the sampling start time of the health information is 2018, 1 month and 1 day.

Respectively defining a sampling interval 1 from 1/2018 to 11/30/2018, a sampling interval 2 from 1/2018 to 7/31/2019, and a sampling interval 3 from 1/2018 to 2020/3/31. Experimenter D1D_{N_X}(t_j) The results are detailed in tables 1 to 5.

TABLE 1 experimenter A1D_{N_H}(t_j)

t_j	D_{N_H}(t_j)	t_j	D_{N_H}(t_j)	t_j	D_{N_H}(t_j)	t_j	D_{N_H}(t_j)
								2018.01	0.94	2018.08	0.91	2019.03	0.91	2019.10	0.90
2018.02	1.00	2018.09	0.91	2019.04	0.94	2019.11	0.95
								2018.03	0.95	2018.10	0.92	2019.05	0.92	2019.12	0.92
2018.04	0.92	2018.11	0.92	2019.06	0.92	2020.01	0.91
								2018.05	0.91	2018.12	0.95	2019.07	0.97	2020.02	0.90
2018.06	0.92	2019.01	0.96	2019.08	0.91	2020.03	0.94
								2018.07	0.94	2019.02	0.92	2019.09	0.92

TABLE 2 experimenter A1D_{N_S}(t_j)

t_j	D_{N_s}(t_j)	t_j	D_{N_S}(t_j)	t_j	D_{N_S}(t_j)	t_j	D_{N_S}(t_j)
								2018.01	0.33	2018.08	0.42	2019.03	0.50	2019.10	0.75
2018.02	0.83	2018.09	0.75	2019.04	1.00	2019.11	0.42
								2018.03	0.25	2018.10	0.33	2019.05	0.33	2019.12	1.00
2018.04	0.75	2018.11	0.33	2019.06	0.83	2020.01	0.83
								2018.05	0.33	2018.12	0.75	2019.07	0.42	2020.02	0.92
2018.06	0.42	2019.01	0.83	2019.08	0.42	2020.03	0.83
								2018.07	0.83	2019.02	0.83	2019.09	0.33

TABLE 3D of experimenter A1_{N_E}(t_j)

t_j	D_{N_E}(t_j)	t_j	D_{N_E}(t_j)	t_j	D_{N_E}(t_j)	t_j	D_{N_E}(t_j)
								2018.01	0.14	2018.08	1.00	2019.03	0.22	2019.10	0.97
2018.02	0.33	2018.09	0.83	2019.04	0.58	2019.11	0.75
								2018.03	0.56	2018.10	0.89	2019.05	0.97	2019.12	0.39
2018.04	0.50	2018.11	0.69	2019.06	0.89	2020.01	0.22
								2018.05	0.89	2018.12	0.67	2019.07	0.83	2020.02	0.28
2018.06	0.69	2019.01	0.42	2019.08	0.89	2020.03	0.33
								2018.07	0.83	2019.02	0.33	2019.09	1.00

TABLE 4 experimenter A1D_{N_T}(t_j)

t_j	D_{N_T}(t_j)	t_j	D_{N_T}(t_j)	t_j	D_{N_T}(t_j)	t_j	D_{N_T}(t_j)
								2018.01	0.05	2018.08	0.27	2019.03	0.73	2019.10	0.69
2018.02	0.31	2018.09	0.42	2019.04	0.58	2019.11	1.00
								2018.03	0.92	2018.10	0.62	2019.05	0.58	2019.12	0.77
2018.04	0.88	2018.11	0.81	2019.06	0.35	2020.01	0.08
								2018.05	0.27	2018.12	0.81	2019.07	0.31	2020.02	0.31
2018.06	0.81	2019.01	0.04	2019.08	0.23	2020.03	0.85
								2018.07	0.23	2019.02	0.35	2019.09	0.31

TABLE 5D of experimenter A1_{N_W}(t_j)

t_j	D_{N_W}(t_j)	t_j	D_{N_W}(t_j)	t_j	D_{N_W}(t_j)	t_j	D_{N_W}(t_j)
								2018.01	0.55	2018.08	0.93	2019.03	0.53	2019.10	0.65
2018.02	0.43	2018.09	0.65	2019.04	0.60	2019.11	0.70
								2018.03	0.52	2018.10	0.74	2019.05	0.68	2019.12	0.54
2018.04	0.67	2018.11	0.82	2019.06	0.75	2020.01	0.52
								2018.05	0.71	2018.12	0.66	2019.07	0.77	2020.02	0.40
2018.06	1.00	2019.01	0.45	2019.08	1.00	2020.03	0.42
								2018.07	0.89	2019.02	0.51	2019.09	0.56

According to the data in tables 1 to 5, D of the three sampling intervals is respectively calculated_{N_X}(t_j) Performing regression to obtain corresponding F_X(t), specifically as follows:

f of sampling interval 1_X(t)：

F_1H(t)＝0.02sin(2π×0.181t+0.2)+0.93；

F_1S(t)＝0.29sin(2π×0.364t-2.67)+0.52；

F_1E(t)＝-1.02sin(2π×0.028t+3.42)-0.13；

F_1T(t)＝0.31sin(2π×0.136t-1.13)+0.49；

F_1W(t)＝-0.20sin(2π×0.068t+1.64)+0.70；

F of sampling interval 2_X(t)：

F_2H(t)＝0.02sin(2π×0.167t+0.54)+0.93；

F_2S(t)＝0.25sin(2π×0.421t-1.90)+0.59；

F_2E(t)＝0.30sin(2π×0.079t-1.90)+0.67；

F_2T(t)＝-0.25sin(2π×0.158t+1.53)+0.50；

F_2W(t)＝-0.19sin(2π×0.079t+1.53)+0.69；

F of sampling interval 3_X(t)：

F_3H(t)＝0.02sin(2π×0.167t+0.33)+0.93；

F_3S(t)＝-0.14sin(2π×0.093t+2.22)+0.60；

F_3E(t)＝0.33sin(2π×0.074t-1.88)+0.66；

F_3T(t)＝0.28sin(2π×0.157t-1.72)+0.52；

F_3W(t)＝-0.19sin(2π×0.078t+1.39)+0.67。

Since the regression data is observed at monthly intervals, it is necessary to predict the health status of the user every day in the future for F_X(t) carrying out a scale transformation to obtain F_X(t/n)。

Constructing and operating a Petri net P (| phi)_B>) to obtain the health status of experimenter a1 in 12 months in 2018, 8 months in 2019 and 4 months in 2020, and the detailed results are shown in tables 6 to 8.

Table 6 health status of experimenter a1 at 12 months in 2018

t_k	max{\|a_m＞}	\|1＞_X
			k＝6	\|a₁₅＞	\|1＞₂、\|1＞₃、\|1＞₄
k＝12	\|a₁₅＞	\|1＞₂、\|1＞₃、\|1＞₄
			k＝18	\|a₁₅＞	\|1＞₂、\|1＞₃、\|1＞₄
k＝24	\|a₁₅＞	\|1＞₂、\|1＞₃、\|1＞₄
			k＝30	\|a₁₅＞	\|1＞₂、\|1＞₃、\|1＞₄

Table 7 health status of experimenter a1 at 8 months in 2019

t_k	max{\|a_m＞}	\|1＞_X
			k＝6	\|a₃＞	\|1＞₄
k＝12	\|a₃＞	\|1＞₄
			k＝18	\|a₃＞	\|1＞₄
k＝24	\|a₃＞	\|1＞₄
			k＝30	\|a₃＞	\|1＞₄

Table 8 health status of experimenter a1 at month 4 of 2020

t_k	max{\|a_m＞}	\|1＞_X
			k＝6	\|a₁₀＞	\|1＞₂、\|1＞₅
k＝12	\|a₁₀＞	\|1＞₂、\|1＞₅
			k＝18	\|a₁₀＞	\|1＞₂、\|1＞₅
k＝24	\|a₁₀＞	\|1＞₂、\|1＞₅
			k＝30	\|a₁₀＞	\|1＞₂、\|1＞₅

Further, all | a's of the experimenter A1 which are in "abnormal" during the above three periods are obtained_mAnd P corresponding thereto_X(t_k) The results are detailed in tables 9 to 11.

TABLE 9P of experimenter A1 at 12 months of 2018_X(t_k)

t_k	P₁(t_k)	P₂(t_k)	P₃(t_k)	P₄(t_k)	P₅(t_k)
						k＝6	11.55％	93.99％	81.82％	91.75％	24.78％
k＝12	10.98％	94.69％	83.91％	93.45％	23.50％
						k＝18	10.50％	93.38％	85.91％	94.73％	22.36％
k＝24	10.13％	89.64％	87.81％	95.66％	21.37％
						k＝30	9.88％	82.78％	89.60％	96.28％	20.55％

TABLE 10 experimenter A1P 8 months in 2019_X(t_k)

t_k	P₁(t_k)	P₂(t_k)	P₃(t_k)	P₄(t_k)	P₅(t_k)
						k＝6	17.16％	39.26％	67.00％	43.77％	53.32％
k＝12	17.19％	52.17％	64.21％	44.09％	51.15％
						k＝18	17.10％	65.85％	61.24％	45.41％	48.94％
k＝24	16.92％	77.13％	58.05％	47.67％	46.70％
						k＝30	16.61％	84.51％	54.84％	51.74％	44.46％

TABLE 11 experimenter A1P at month 4 of 2020_X(t_k)

t_k	P₁(t_k)	P₂(t_k)	P₃(t_k)	P₄(t_k)	P₅(t_k)
						k＝6	16.94％	70.39％	17.82％	36.07％	51.23％
k＝12	16.65％	67.93％	21.01％	37.06％	53.36％
						k＝18	16.27％	66.40％	24.36％	39.21％	55.46％
k＝24	15.80％	64.81％	27.86％	42.38％	57.50％
						k＝30	15.27％	63.18％	31.46％	46.42％	61.48％

From the results of tables 6 and 9, it can be seen that: the organs of the experimenter a1 in the "abnormal" state in 12 months in 2018 are the liver, the spleen and the lung, and the probability values of the organs in the "abnormal" state can be respectively as high as 94.69%, 89.60% and 96.28%, so that the experimenter a1 can be predicted to be possibly suffered from diseases of the liver, the spleen or the lung (respiratory) in 12 months in 2018. The experimenter A1 is known to have been diagnosed with gastroenteritis in 2018 in 12 months, and according to the theory of traditional Chinese medicine, the gastrointestinal diseases are spleen diseases.

From the results of tables 7 and 10, it can be seen that: the organ of the experimenter a1 in the "abnormal" state in 8 months in 2019 is the lung, and the probability value of the organ in the "abnormal" state can be as high as 51.74%, so that the experimenter a1 can be predicted to be possibly suffered from lung (respiratory) diseases in 8 months in 2019. Experimenters A1 are known to have been diagnosed with chronic bronchitis in 8 months of 2019, and according to the theory of traditional Chinese medicine, the chronic bronchitis is a respiratory disease.

From the results of tables 8 and 11, it can be seen that: the organs of the experimenter A1 in the "abnormal" state in 4 months of 2020 are the liver and the kidney, and the probability values of the organs in the "abnormal" state can be respectively as high as 70.39% and 61.48%, so that the experimenter A1 can be predicted to possibly have liver or kidney diseases in 4 months of 2020. The experimenter A1 is known to be diagnosed with gout in the year 2020, 4 months, and according to the traditional Chinese medicine theory, gout is liver and kidney diseases.

In summary, the analysis result of the health status of experimenter a1 in this embodiment substantially matches the actual disease history.

The invention discloses a human body health state analysis scheme based on a quantum Petri net, which utilizes a traditional Chinese medicine theory to establish medical association between human viscera and health information, and utilizes a quantum information theory to establish association between the human viscera and quantum bits and quantum disturbance, thereby predicting possible illness conditions of a human body in a period of time in the future.

Claims

1. A human health state analysis scheme based on a quantum Petri network is characterized by comprising the following steps:

step 1: acquiring health information related to human viscera and carrying out preprocessing operation on the health information to obtain preprocessing data corresponding to the health information;

step 2: the human viscera are equivalent to qubits, and a health state quantum system is established according to the qubits;

and step 3: performing regression processing on the preprocessed data to obtain a perturbation operator corresponding to the quantum bit;

and 4, step 4: acting the perturbation operator on the qubit to obtain a quantum system perturbation equation;

and 5: constructing and operating a human health state Petri net according to a quantum system disturbance equation to obtain a probability amplitude corresponding to the eigenstate of a health state quantum system;

step 6: and obtaining the analysis result of the health state of the human body according to the probability amplitude corresponding to the eigen state, namely possible diseased species and diseased probability.

2. The human health state analysis scheme based on the quantum Petri net as claimed in claim 1, wherein the health information related to human viscera is acquired in the step 1 and is preprocessed to obtain preprocessed data corresponding to the health information, and the method comprises the following steps: the health information refers to various factors that can reflect or have strong influence on the health status of the viscera, including but not limited to environmental information and physiological information.

3. The human health state analysis scheme based on quantum Petri net as claimed in claim 1, wherein in the step 2, human viscera are equivalent to qubits, and a health state quantum system is established according to the qubits, wherein: the two basis vectors of qubits correspond to two health states of the human viscera, namely "normal" and "abnormal"; a health state quantum system is a complex quantum system spanned by qubits.

4. The human health state analysis scheme based on the quantum Petri net according to claim 1, wherein the preprocessed data is subjected to regression processing in the step 3 to obtain perturbation operators corresponding to quantum bits, and the scheme is characterized in that: regression processing methods include, but are not limited to, machine learning, least squares, interpolation, fitting, and the like.

5. The human health state analysis scheme based on the quantum Petri net according to claim 1, wherein in the step 4, perturbation operators are applied to qubits to obtain a quantum system perturbation equation, and the quantum system perturbation equation is characterized in that: the perturbation operator has single mapping and only acts on corresponding quantum bits.

6. The human health state analysis scheme based on the quantum Petri net according to claim 1, wherein the human health state Petri net is constructed and operated according to a quantum system disturbance equation in the step 5 to obtain a probability amplitude corresponding to the eigen state of a quantum system in the health state, and the scheme is characterized in that: the human health state Petri net automatically operates in time, and consists of three modules, namely a disturbance module, a quantum bit module and an eigen state module, wherein:

the state of the qubit in the qubit module determines the probability amplitude corresponding to the eigenstates in the eigenstate module.

7. The human health status analysis scheme based on the quantum Petri net according to claim 1, wherein the human health status analysis result, namely the possible disease type and disease probability, is obtained according to the probability amplitude corresponding to the eigen state in the step 6, and is characterized in that: and calculating probability values corresponding to the eigenstates according to the probability amplitude corresponding to the eigenstates, sequencing the probability values in an ascending mode to obtain the eigenstates with the maximum probability values, wherein all abnormal bits in the eigenstates are used for describing the diseased species.

8. The human health status analysis scheme based on the quantum Petri net according to claim 1, wherein the human health status analysis result, namely the possible disease type and disease probability, is obtained according to the probability amplitude corresponding to the eigen state in the step 6, and is characterized in that: and calculating to obtain probability values corresponding to all the 'abnormal' eigenstates, summing the probability values, and taking the sum result as a diseased probability value.