CN109741827B - Chest and abdomen surface area respiratory signal period prediction method combining double period judgment - Google Patents

Abstract

The method for predicting the breathing signal period of the thoracic and abdominal surface area combined with the dual-period judgment belongs to the technical fields of oncology medicine, precision instruments, engineering technology, mathematics, etc. Periodic breathing signal, and finally the ideal multi-period breathing signal is periodically extracted. The entire extraction process is carried out between two imaginary cycles with the same amount of data, and the periodic extraction can be completed within two cycles; this method can accurately extract the ideal multi-cycle breathing signal. The period of the periodic breathing signal lays a simulation experiment foundation for accurately predicting the breathing movement.

Description

Prediction method of respiratory signal period in thoracic and abdominal surface area combined with double-period judgment

technical field

The method for predicting the period of the respiratory signal in the surface area of the chest and abdomen combined with double-period judgment belongs to the technical fields of tumor medicine, precision instruments, engineering technology and mathematics.

Background technique

Radiation therapy (hereinafter referred to as radiotherapy) is one of the most important means of treating cancer. About 70% of cancer patients will use radiotherapy in the course of cancer treatment, and about 40% of cancers can be cured by radiotherapy. The effect of radiation therapy is related to the precision and dose of radiation delivered to the tumor area. Malignant tumors of the digestive system are located in the thoracic and abdominal cavity. Affected by respiratory movement, the position and volume of the tumor will change over time, resulting in a decrease in the accuracy of radiation irradiation, thereby affecting the effect of radiotherapy.

In order to solve the problem that respiratory movement reduces the effect of radiotherapy, many scholars have carried out research work on technologies or methods such as target area expansion, breath-holding, respiratory gating, passive compression, four-dimensional radiotherapy, and real-time tracking. These works have effectively reduced the effect of respiratory movement on radiotherapy effects. However, since these works are all lag compensation work carried out after the occurrence of respiratory movement, they cannot fundamentally overcome the problem that respiratory movement reduces the effect of radiotherapy.

In order to solve the above problems, some scholars have begun to study the advanced compensation method, which first needs to predict the breathing movement. At present, respiratory motion prediction methods can be divided into the following two categories: first, model prediction methods, which establish a respiratory motion model by monitoring previous respiratory motion data, and then predict future respiratory motion based on the model; second, model-free prediction method, which is trained by observed breathing motion patterns to form a heuristic learning algorithm that learns to predict future breathing motions.

Although the breathing movement repeats the inhalation and exhalation movements continuously, the breathing movement does not simply repeat the inhalation and exhalation as a cycle. and amplitude will change, and the model prediction method cannot take care of this change, so the prediction accuracy will be restricted. Therefore, more scholars have begun to try in the model-free prediction method, for example, using Gaussian process regression to predict respiratory movement.

Gaussian process regression is a universal prediction algorithm that predicts future data by learning known data, and the prediction results are given in the form of mean and variance. Since the Gaussian process regression algorithm itself is not proposed purely for respiratory motion prediction, some constraints of respiratory motion will not be considered. Taking into account the constraints of breathing motion narrows the prediction range and makes the predictions more accurate.

For respiratory movement, its frequency will be limited within a certain range, so the prediction result can be constrained by obtaining the period or frequency of the respiratory signal, and the prediction accuracy can be improved. It can be seen that obtaining the cycle of breathing motion is beneficial to improve the prediction accuracy.

However, since the respiratory movement is not a simple repetition of a single period, it is difficult to judge the period by the distance of two identical eigenvalues. Then, how to obtain the cycle of respiratory motion has become a key technical issue to improve the prediction accuracy.

Contents of the invention

In order to predict the period of the respiratory signal, the present invention starts from the ideal respiratory signal and proposes a method for extracting the period of the respiratory signal. Work.

The purpose of the present invention is achieved like this:

The method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment consists of the following steps:

Step a, constructing an ideal single-period respiratory signal;

Step b, performing period extension on the ideal single-period respiratory signal obtained in step a, to obtain an ideal multi-period respiratory signal;

Step c, carry out periodic extraction to the ideal multi-period breathing signal obtained in step b, including:

Step c1, set the period data variable to 1, and enter step c2;

Step c2. Calculate the sum of all data in the first period and the sum of all data in the second period according to the period data variables, and proceed to step c3;

Step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if:

Yes, go to step c4;

No, period data variable +1, return to step c2;

Step c4, the period data variable is the period of the multi-period respiratory signal.

The above method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment, the construction of an ideal single-period respiratory signal as described in step a, includes the following steps:

Step a1, according to the normal breathing process of human beings, the breathing movement is divided into three stages: inhalation process, exhalation process and quasi-pause process;

Step a2, respectively determine the duration T ₁ of the inhalation process, the duration T ₂ of the exhalation process and the duration T ₃ of the pause-like process;

Step a3, simulate the inhalation process with a sine function with a period of 4T and _a phase of [0, π/2] to obtain a simulated inspiratory signal; use a sine function with a period of 4T and a phase of [π/ ₂ , π] to simulate exhalation Breathing process, to obtain a simulated expiratory signal; use a sine function with a period of 4T and a phase of [π, 3π/ ₂ ] to simulate a class pause process, to obtain a simulated class pause signal;

Step a4. Adjust the relative amplitude of the simulated inspiratory signal to [a, b], adjust the relative amplitude of the simulated exhalation signal to [c, b], and adjust the relative amplitude of the simulated pause signal to Adjust to between [a,c];

Step a5, connecting the obtained simulated inspiratory signal, simulated expiratory signal and simulated pause signal end to end to obtain an ideal single-period respiratory signal.

The method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment consists of the following steps:

Step a, constructing an ideal single-period respiratory signal;

Step a1, according to the normal breathing process of human beings, the breathing movement is divided into three stages: inhalation process, exhalation process and quasi-pause process;

Step a2, respectively determine the duration T ₁ of the inhalation process, the duration T ₂ of the exhalation process and the duration T ₃ of the pause-like process;

Step a3, simulate the inhalation process with a sine function with a period of 4T and _a phase of [0, π/2] to obtain a simulated inspiratory signal; use a sine function with a period of 4T and a phase of [π/ ₂ , π] to simulate exhalation Breathing process, to obtain a simulated expiratory signal; use a sine function with a period of 4T and a phase of [π, 3π/ ₂ ] to simulate a class pause process, to obtain a simulated class pause signal;

Step a4. Adjust the relative amplitude of the simulated inspiratory signal to [a, b], adjust the relative amplitude of the simulated exhalation signal to [c, b], and adjust the relative amplitude of the simulated pause signal to Adjust to between [a,c];

Step a5, connecting the obtained simulated inspiratory signal, simulated expiratory signal and simulated pause signal end to end to obtain an ideal single-period respiratory signal;

Step b, performing period extension on the ideal single-period respiratory signal obtained in step a, to obtain an ideal multi-period respiratory signal;

Step c, carry out periodic extraction to the ideal multi-period breathing signal obtained in step b, including:

Step c1, set the period data variable to 1, and enter step c2;

Step c2. Calculate the sum of all data in the first period and the sum of all data in the second period according to the period data variables, and proceed to step c3;

Step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if:

Yes, go to step c4;

No, period data variable +1, return to step c2;

Step c4, the period data variable is the period of the multi-period respiratory signal.

A method for constructing an ideal single-period respiratory signal, comprising the following steps:

Step a1, according to the normal breathing process of human beings, the breathing movement is divided into three stages: inhalation process, exhalation process and quasi-pause process;

Step a2, respectively determine the duration T ₁ of the inhalation process, the duration T ₂ of the exhalation process and the duration T ₃ of the pause-like process;

Step a3, simulate the inhalation process with a sine function with a period of 4T and _a phase of [0, π/2] to obtain a simulated inspiratory signal; use a sine function with a period of 4T and a phase of [π/ ₂ , π] to simulate exhalation Breathing process, to obtain a simulated expiratory signal; use a sine function with a period of 4T and a phase of [π, 3π/ ₂ ] to simulate a class pause process, to obtain a simulated class pause signal;

Step a4. Adjust the relative amplitude of the simulated inspiratory signal to [a, b], adjust the relative amplitude of the simulated exhalation signal to [c, b], and adjust the relative amplitude of the simulated pause signal to Adjust to between [a,c];

Step a5, connecting the obtained simulated inspiratory signal, simulated expiratory signal and simulated pause signal end to end to obtain an ideal single-period respiratory signal.

A method for extracting an ideal multi-period breathing signal period, comprising the following steps:

Step c1, set the period data variable to 1, and enter step c2;

Step c2. Calculate the sum of all data in the first period and the sum of all data in the second period according to the period data variables, and proceed to step c3;

Step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if:

Yes, go to step c4;

No, period data variable +1, return to step c2;

Step c4, the period data variable is the period of the multi-period respiratory signal.

Beneficial effect:

First, although the present invention belongs to the technical field of tumor medicine, the present invention extracts the signal feature quantity from a technical point of view, and in view of the technical purpose and function of the present invention alone, it is impossible to diagnose and diagnose diseases. Therefore, it does not belong to the diagnosis and treatment methods of diseases mentioned in Article 25 of the Patent Law, and there is no object problem in the present invention.

Second, the disclosure of the present invention is sufficient and sufficient. As long as the source code described in the specification is run in MATLAB software, the running result can be obtained, and this application can be realized by anyone skilled in the art or not.

Third, the present invention also provides a method for constructing an ideal single-period respiratory signal. This method only needs to use three sections of sine functions to fit the respiratory signal within one period. The method is simple, the function is simple, and the fitting result Corresponds to real breathing movements.

Fourth, the method of the present invention is carried out during two imaginary weeks with the same amount of data, and compared with the invention patent "Method for Predicting the Period of Respiratory Signals in the Human Chest and Abdomen Surface Area Based on Variance Calculation" declared by the research team on the same day, as long as one The cycle extraction can be completed within the real cycle, so the amount of calculation is smaller. It can also be seen from the simulation results that the running time is increased from 0.40s to 0.14s.

Fifth, the method of the present invention is carried out during two imaginary weeks with the same amount of data. The invention patent "Method for Predicting the Period of Respiratory Signals in the Chest and Abdominal Surface Area Combining Three-Period Judgment" declared by the research team on the same day takes into account the breathing movement and exhalation It is asymmetrical with the inspiratory process, so data mutual verification can be realized within two imaginary cycles. While avoiding false cycles and making prediction results more accurate, the running time can be further improved. It can also be seen from the simulation results that running Time improved from 0.32s to 0.14s.

Description of drawings

Figure 1 is an ideal multi-cycle respiratory signal diagram.

Figure 2 is the program running interface.

specific embodiment

The specific embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings.

Specific embodiment one

This embodiment is an embodiment of a method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment.

The method for predicting the period of the respiratory signal in the thoracoabdominal surface area combined with double-period judgment in this embodiment consists of the following steps:

Step a, constructing an ideal single-period respiratory signal;

Step b, performing period extension on the ideal single-period respiratory signal obtained in step a, to obtain an ideal multi-period respiratory signal;

Step c, carry out periodic extraction to the ideal multi-period breathing signal obtained in step b, including:

Step c1, set the period data variable to 1, and enter step c2;

Step c2. Calculate the sum of all data in the first period and the sum of all data in the second period according to the period data variables, and proceed to step c3;

Step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if:

Yes, go to step c4;

No, period data variable +1, return to step c2;

Step c4, the period data variable is the period of the multi-period respiratory signal.

Specific embodiment two

This embodiment is an embodiment of a method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment.

The method for predicting the breathing signal period of the thoracoabdominal surface area combined with double-period judgment in this embodiment, on the basis of the specific embodiment 1, further limits the construction of an ideal single-period respiratory signal described in step a, including the following steps:

Step a1, according to the normal breathing process of human beings, the breathing movement is divided into three stages: inhalation process, exhalation process and quasi-pause process;

Step a2, respectively determine the duration T ₁ of the inhalation process, the duration T ₂ of the exhalation process and the duration T ₃ of the pause-like process;

Step a3, simulate the inhalation process with a sine function with a period of 4T and _a phase of [0, π/2] to obtain a simulated inspiratory signal; use a sine function with a period of 4T and a phase of [π/ ₂ , π] to simulate exhalation Breathing process, to obtain a simulated expiratory signal; use a sine function with a period of 4T and a phase of [π, 3π/ ₂ ] to simulate a class pause process, to obtain a simulated class pause signal;

Step a4. Adjust the relative amplitude of the simulated inspiratory signal to [a, b], adjust the relative amplitude of the simulated exhalation signal to [c, b], and adjust the relative amplitude of the simulated pause signal to Adjust to between [a,c];

Step a5, connecting the obtained simulated inspiratory signal, simulated expiratory signal and simulated pause signal end to end to obtain an ideal single-period respiratory signal.

Specific embodiment three

This embodiment is an embodiment of a method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment.

The method for predicting the period of the respiratory signal in the thoracoabdominal surface area combined with double-period judgment in this embodiment consists of the following steps:

Step a, constructing an ideal single-period respiratory signal;

Step a1, according to the normal breathing process of human beings, the breathing movement is divided into three stages: inhalation process, exhalation process and quasi-pause process;

Step a2, respectively determine the duration T ₁ of the inhalation process, the duration T ₂ of the exhalation process and the duration T ₃ of the pause-like process;

Step a3, simulate the inhalation process with a sine function with a period of 4T and _a phase of [0, π/2] to obtain a simulated inspiratory signal; use a sine function with a period of 4T and a phase of [π/ ₂ , π] to simulate exhalation Breathing process, to obtain a simulated expiratory signal; use a sine function with a period of 4T and a phase of [π, 3π/ ₂ ] to simulate a class pause process, to obtain a simulated class pause signal;

Step a4. Adjust the relative amplitude of the simulated inspiratory signal to [a, b], adjust the relative amplitude of the simulated exhalation signal to [c, b], and adjust the relative amplitude of the simulated pause signal to Adjust to between [a,c];

Step a5, connecting the obtained simulated inspiratory signal, simulated expiratory signal and simulated pause signal end to end to obtain an ideal single-period respiratory signal;

Step b, performing period extension on the ideal single-period respiratory signal obtained in step a, to obtain an ideal multi-period respiratory signal;

Step c, carry out periodic extraction to the ideal multi-period breathing signal obtained in step b, including:

Step c1, set the period data variable to 1, and enter step c2;

Step c2. Calculate the sum of all data in the first period and the sum of all data in the second period according to the period data variables, and proceed to step c3;

Step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if:

Yes, go to step c4;

No, period data variable +1, return to step c2;

Step c4, the period data variable is the period of the multi-period respiratory signal.

Specific embodiment four

This embodiment is an embodiment of a method for constructing an ideal single-period respiratory signal.

A method for constructing an ideal single-period respiratory signal in this embodiment includes the following steps:

Step a1, according to the normal breathing process of human beings, the breathing movement is divided into three stages: inhalation process, exhalation process and quasi-pause process;

Step a2, respectively determine the duration T ₁ of the inhalation process, the duration T ₂ of the exhalation process and the duration T ₃ of the pause-like process;

Step a3, simulate the inhalation process with a sine function with a period of 4T and _a phase of [0, π/2] to obtain a simulated inspiratory signal; use a sine function with a period of 4T and a phase of [π/ ₂ , π] to simulate exhalation Breathing process, to obtain a simulated expiratory signal; use a sine function with a period of 4T and a phase of [π, 3π/ ₂ ] to simulate a class pause process, to obtain a simulated class pause signal;

Step a4. Adjust the relative amplitude of the simulated inspiratory signal to [a, b], adjust the relative amplitude of the simulated exhalation signal to [c, b], and adjust the relative amplitude of the simulated pause signal to Adjust to [a,c];

Step a5, connecting the obtained simulated inspiratory signal, simulated expiratory signal and simulated pause signal end to end to obtain an ideal single-period respiratory signal.

Specific embodiment five

This embodiment is an embodiment of a method for extracting periods of an ideal multi-period respiratory signal.

A kind of ideal multi-period respiration signal period extraction method of the present embodiment, comprises the following steps:

Step c1, set the period data variable to 1, and enter step c2;

Step c2. Calculate the sum of all data in the first period and the sum of all data in the second period according to the period data variables, and proceed to step c3;

Step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if:

Yes, go to step c4;

No, period data variable +1, return to step c2;

Step c4, the period data variable is the period of the multi-period respiratory signal.

Specific embodiment seven

This embodiment is an embodiment of a method for predicting the period of the respiratory signal in the chest and abdomen surface area combined with double-period judgment.

In order to verify the method of the present invention, the method was run on MATLAB R2014a software. The applied computer configuration is as follows:

computer configuration table

operating system Windows 7 Ultimate X86 processor Intel(R) Core(TM) i5-8250OU@1.60GHz 1.80GHz Install memory 8.00GB

The written MATLAB program is as follows:

The program running results are shown in Figure 1 and Figure 2 respectively. Among them, Fig. 1 is a signal diagram of an ideal multi-cycle breathing signal, and Fig. 2 is a program running interface.

The result of program operation shows that the method of the present invention can accurately extract the signal period from the respiratory motion signal.

Finally, thanks to the general project of the National Natural Science Foundation of China "Dynamic three-dimensional measurement of human chest and abdomen surface in radiotherapy and analysis of regional respiratory motion and prediction of space-time integration" (Project No. Funding support for this patent from Research on Prediction of Respiratory Movement on the Abdominal Surface (Project No.: 201810214267).

Claims (2)

1. The method for predicting the respiratory signal period of the chest and abdomen surface area by combining double-period judgment is characterized by comprising the following steps of:

step a, constructing an ideal single-cycle respiratory signal, comprising the following steps:

a1, according to the normal breathing process of human, dividing the breathing movement into three stages of an inspiration process, an expiration process and a pause-like process;

step a2, respectively determining the duration T1 of an inspiration process, the duration T2 of an expiration process and the duration T3 of a pause-like process;

a3, simulating an air suction process by using a sine function with a period of 4T1 and a phase of [0, pi/2 ] to obtain a simulated air suction signal; simulating the expiration process by using a sine function with the period of 4T2 and the phase of [ pi/2, pi ] to obtain a simulated expiration signal; simulating a class pause process by using a sine function with a period of 4T3 and a phase of [ pi, 3 pi/2 ] to obtain a simulated class pause signal;

step a4, adjusting the relative amplitude of the analog inspiration signal to be between [ a and b ], adjusting the relative amplitude of the analog expiration signal to be between [ c and b ], and adjusting the relative amplitude of the analog pause signal to be between [ a and c ];

step a5, connecting the obtained simulated inspiration signal, simulated expiration signal and simulated pause signal end to obtain an ideal single-cycle respiration signal;

b, carrying out periodic continuation on the ideal single-period respiratory signal obtained in the step a to obtain an ideal multi-period respiratory signal;

step c, carrying out periodic extraction on the ideal multi-periodic respiratory signal obtained in the step b, wherein the periodic extraction comprises the following steps:

step c1, setting a period data variable as 1, and entering step c2;

step c2, calculating the sum of all data in the first period and the sum of all data in the second period according to the period data variable, and entering the step c3;

step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if so:

if yes, go to step c4;

if not, the periodic data variable is +1, and the step c2 is returned;

and c4, the period data variable is the period of the multicycle respiratory signal.

2. A method for extracting ideal multicycle respiratory signal cycle is characterized by comprising the following steps:

step c1, setting a period data variable as 1, and entering step c2;

step c2, calculating the sum of all data in the first period and the sum of all data in the second period according to the period data variable, and entering the step c3;

step c3, judging whether the sum of all data in the first period is equal to the sum of all data in the second period, if so:

if yes, go to step c4;

if not, the periodic data variable is +1, and the step c2 is returned;

and c4, the period data variable is the period of the multicycle respiratory signal.

Images