Method and device for determining respiratory stage, magnetic resonance imaging method and system
By extracting the characteristic values of the respiratory signal and training the model, accurately identifying the respiratory stage of the respiratory signal, the problem of respiratory motion artifacts in magnetic resonance imaging is solved and the image quality is improved.
Patent Information
- Application Number
- CN201910772771.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2019-08-21
- Publication Date
- 2024-12-20
- Estimated Expiration
- 2040-06-14
AI Technical Summary
During magnetic resonance imaging, the respiratory movement of the subject causes a decrease in image quality, and it is difficult for the prior art to accurately determine the respiratory stage of the respiratory signal, resulting in the presence of motion artifacts.
By extracting distance eigenvalues, point eigenvalues and area eigenvalues from the breathing signal, the breathing signal model is trained to accurately determine the respiratory stage of the breathing signal. Specific methods include calculating the extreme points and time points of adjacent waveforms, and using these eigenvalues to calculate the distance, number of points and area ratios to identify the breathing stage.
Accurate phase recognition of the respiratory signal is achieved, motion artifacts in magnetic resonance imaging are avoided, and image quality is improved.
Smart Images
Figure CN112401867B_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of medical magnetic resonance imaging, and in particular to methods and apparatus for determining respiratory phases, magnetic resonance imaging (MRI) methods and systems, and storage media. Background Technology
[0002] Magnetic resonance imaging (MRI) is a type of tomographic imaging that uses the magnetic resonance phenomenon to obtain electromagnetic signals from a subject, such as the human body, and reconstruct information. Specifically, an MRI system applies a radio frequency pulse of a specific frequency to the human body in a static magnetic field, exciting hydrogen protons within the body and causing them to resonate. After the pulse stops, the protons generate magnetic resonance (MR) signals during relaxation. Through processing such as receiving the MR signals, spatial encoding, and image reconstruction, information about the human body is reconstructed.
[0003] During magnetic resonance imaging (MRI), the respiratory motion of the subject degrades the image quality of areas affected by respiratory motion, such as the abdomen. To eliminate motion artifacts in MRI images caused by respiratory motion, a respiratory sensor is typically used to detect the subject's respiratory status and trigger MRI data acquisition at an appropriate time.
[0004] However, different coil units in a magnetic resonance imaging system receive inconsistent respiratory motion signals, which results in each receiving channel obtaining different respiratory motion information, and this information will vary for different subjects and different locations. Summary of the Invention
[0005] According to one aspect of the present disclosure, a method for determining the respiratory phase of a respiratory signal is provided. The method includes: extracting distance feature values, point count feature values, and area feature values from the respiratory signal, wherein the distance feature values, the point count feature values, and the area feature values are used to indicate waveform changes between two adjacent waveforms in the respiratory signal; training a respiratory signal model based on the distance feature values, the point count feature values, and the area feature values to determine the respiratory phase of the respiratory signal using the respiratory signal model.
[0006] The above method solves the problem of accurately determining the respiratory stage of respiratory signals in related technologies, and has the beneficial effect of accurately determining the respiratory stage.
[0007] In one exemplary embodiment of this disclosure, the distance feature value is the ratio of the width of the second half of the first waveform in the two adjacent waveforms to the width of the first half of the second waveform in the two adjacent waveforms; and / or the point count feature value is the ratio of the width of the rising segment of the first waveform to the peak point to the width of the falling segment of the first waveform to the trough point; and / or the area feature value is the ratio of the area formed by the two peak points of the two adjacent waveforms and the trough point between the two waveforms to the area formed by the second waveform.
[0008] By extracting waveform change features that characterize two adjacent waveforms of the respiratory signal, such as distance feature values, number of points feature values, and area feature values, the respiratory phase can be identified by comparing the different curve shapes of inhalation and exhalation.
[0009] In an exemplary embodiment of this disclosure, before extracting distance feature values, point count feature values, and area feature values from the respiratory signal, the method further includes: obtaining the peak points of the first waveform and the second waveform, respectively, as the first maximum point and the second maximum point; obtaining the trough point between the two waveforms, as the first minimum point, and obtaining the trough point of the second waveform after the second maximum point, as the second minimum point; determining the time points corresponding to the first maximum point, the first minimum point, the second maximum point, and the second minimum point, respectively, as the first maximum time point, the first minimum time point, the second maximum time point, and the second minimum time point.
[0010] Using the above method, we can obtain the relevant extreme points of two adjacent waveforms of a respiratory signal and their corresponding time points. Thus, we can use the obtained extreme points and corresponding time points to easily and quickly calculate the respiratory characteristic values of the respiratory signal related to waveform changes.
[0011] In one exemplary embodiment of this disclosure, extracting distance feature values from a respiratory signal includes: calculating the ratio of the difference between the first minimum time point and the first maximum time point to the difference between the second maximum time point and the first minimum time point; and using the calculated ratio as the extracted distance feature value.
[0012] By calculating the ratio of the bottom width of the second half of the first waveform to the bottom width of the first half of the second waveform in two adjacent waveforms, the changes between the second half of the first waveform and the first half of the second waveform can be compared, thus making it easier to extract the distance feature value of the respiratory signal.
[0013] In one exemplary embodiment of this disclosure, extracting point count feature values from a respiratory signal includes: determining, based on a predetermined amplitude ratio parameter and the first maximum and first minimum points, the start time points of an upward segment of the first waveform up to the first maximum and a downward segment of the first waveform up to the first minimum, respectively, as a first time point and a second time point for determining the end-expiratory start time point of the respiratory phase; calculating the ratio of the difference between the first maximum time point and the first time point to the difference between the first minimum time point and the second time point; and using the calculated ratio as the extracted point count feature value.
[0014] By calculating the ratio between the time corresponding to the rising waveform to the peak point and the time corresponding to the falling waveform to the trough point in the first waveform, the changes in the waveform near the peak and the waveform near the trough of the first waveform can be compared, thus making it easier to extract the point feature value of the respiratory signal.
[0015] In an exemplary embodiment of this disclosure, the first time point and the second time point are determined according to the following two formulas: mag(t_1)=mag(t_max1)*p+mag(t_min1)*(1-p); mag(t_2)=mag(t_max1)*(1-p)+mag(t_min1)*p; where t_1 is the first time point, t_max1 is the first maximum time point, p is the amplitude ratio parameter, t_min1 is the first minimum time point, and mag is the amplitude function.
[0016] The above formula can be used to accurately calculate the point characteristic value of the respiratory signal.
[0017] In one exemplary embodiment of this disclosure, the amplitude scaling parameter p is equal to 80%.
[0018] The amplitude ratio parameter of 80% is an empirical value determined in practice taking into account other parameters of the magnetic resonance imaging system. By setting the amplitude ratio parameter to 80%, the point characteristic value of the respiratory signal can be calculated more accurately.
[0019] In one exemplary embodiment of this disclosure, extracting area feature values from a respiratory signal includes: calculating the ratio of the area of a first polygon determined by the first maximum point, the second maximum point, and the first minimum point to the area of a second polygon determined by the first minimum point, the second maximum point, and the second minimum point; and using the calculated ratio as the extracted area feature value.
[0020] By calculating the ratio of the area formed by the two peaks of two adjacent waveforms and the trough between the two waveforms, and the ratio of the area formed by the two troughs of the two waveforms and the peak of the second waveform, the changes between the first and second waveforms can be compared, thus making it easier to extract the area feature value of the respiratory signal.
[0021] According to another aspect of the present disclosure, an apparatus for determining the respiratory stage of a respiratory signal is provided, comprising: an extraction module configured to extract distance feature values, point count feature values, and area feature values from the respiratory signal, wherein the distance feature values, the point count feature values, and the area feature values are used to indicate waveform changes between two adjacent waveforms in the respiratory signal; and a determination module configured to train a respiratory signal model based on the distance feature values, the point count feature values, and the area feature values to determine the respiratory stage of the respiratory signal using the respiratory signal model.
[0022] The aforementioned device solves the problem in related technologies where the respiratory phase of a respiratory signal cannot be accurately determined, thus providing the beneficial effect of accurately determining the respiratory phase.
[0023] In one exemplary embodiment of this disclosure, the extraction module includes: a maximum point acquisition unit configured to acquire the peak point of the first waveform and the peak point of the second waveform, respectively, as the first maximum point and the second maximum point; a minimum point acquisition unit configured to acquire the trough point between the two waveforms, as the first minimum point, and acquire the trough point of the second waveform after the second maximum point, as the second minimum point; and a time point determination unit configured to determine the time points corresponding to the first maximum point, the first minimum point, the second maximum point, and the second minimum point, respectively, as the first maximum time point, the first minimum time point, the second maximum time point, and the second minimum time point.
[0024] The above structure allows us to obtain the relevant extreme points of two adjacent waveforms of a respiratory signal and their corresponding time points. This allows us to quickly and easily calculate the respiratory characteristic values of the respiratory signal that are related to waveform changes using the obtained extreme points and corresponding time points.
[0025] In one exemplary embodiment of this disclosure, the extraction module includes: a distance calculation unit configured to calculate the ratio of the difference between the first minimum time point and the first maximum time point to the difference between the second maximum time point and the first minimum time point; and a distance extraction unit configured to use the calculated ratio as the extracted distance feature value.
[0026] By calculating the ratio of the bottom width of the second half of the first waveform to the bottom width of the first half of the second waveform in two adjacent waveforms, the changes between the second half of the first waveform and the first half of the second waveform can be compared, thus making it easier to extract the distance feature value of the respiratory signal.
[0027] In one exemplary embodiment of this disclosure, the extraction module includes: a feature point determination unit, configured to determine, based on a predetermined amplitude ratio parameter and the first maximum point and the first minimum point, the start time point of an upward segment of the first waveform up to the first maximum point and the start time point of a downward segment of the first waveform up to the first minimum point, respectively serving as a first time point and a second time point for determining the end-expiration start time point of the respiratory phase; a feature point calculation unit, configured to calculate the ratio of the difference between the first maximum point and the first time point to the difference between the first minimum point and the second time point; and a feature point extraction unit, configured to use the calculated ratio as the extracted point feature value.
[0028] By calculating the ratio between the time corresponding to the rising waveform to the peak point and the time corresponding to the falling waveform to the trough point in the first waveform, the changes in the waveform near the peak and the waveform near the trough of the first waveform can be compared, thus making it easier to extract the point feature value of the respiratory signal.
[0029] In one exemplary embodiment of this disclosure, the extraction module includes: an area calculation unit configured to calculate the ratio of the area of a first polygon determined by the first maximum point, the second maximum point, and the first minimum point to the area of a second polygon determined by the first minimum point, the second maximum point, and the second minimum point; and an area feature extraction unit configured to use the calculated ratio as the extracted area feature value.
[0030] By calculating the ratio of the area formed by the two peaks of two adjacent waveforms and the trough between the two waveforms, and the ratio of the area formed by the two troughs of the two waveforms and the peak of the second waveform, the changes between the first and second waveforms can be compared, thus making it easier to extract the area feature value of the respiratory signal.
[0031] According to another aspect of this disclosure, a magnetic resonance imaging method is provided, comprising: determining the respiratory phase of a respiratory signal according to the above-described method for determining the respiratory phase of a respiratory signal provided in this disclosure; and performing magnetic resonance imaging on a subject at a site of examination based on the determined respiratory phase.
[0032] In the above method, by determining the end-expiratory phase of the respiratory signal, the magnetic resonance imaging system can be triggered to scan at the beginning time of the determined end-expiratory phase, thereby avoiding motion artifacts in magnetic resonance imaging.
[0033] According to another aspect of this disclosure, a magnetic resonance imaging system is provided, comprising: the means for determining the respiratory phase of a respiratory signal as described above, configured to determine the respiratory phase of a respiratory signal; and an imaging device configured to perform magnetic resonance imaging on a subject at a site of examination based on the determined respiratory phase.
[0034] In this magnetic resonance imaging system, by determining the end-expiratory phase of the respiratory signal, the magnetic resonance imaging system can be triggered to scan at the determined start time of the end-expiratory phase, thereby avoiding motion artifacts in magnetic resonance imaging.
[0035] According to another aspect of this disclosure, a non-transitory readable storage medium is provided having a program stored thereon, which, when executed, causes a processor to perform the method provided in this disclosure for determining the respiratory phase of a respiratory signal. Attached Figure Description
[0036] The accompanying drawings, which are included to provide a further understanding of this disclosure and form part of this disclosure, illustrate exemplary embodiments of the present disclosure and are used to explain the disclosure, but do not constitute an undue limitation of the disclosure. In the drawings:
[0037] Figure 1 It is a time-series diagram of respiratory signals based on relevant technologies;
[0038] Figure 2 This is a flowchart of a method for determining the respiratory phase of a respiratory signal according to an embodiment of the present disclosure;
[0039] Figure 3 This is a flowchart of obtaining respiratory feature values according to an embodiment of the present disclosure;
[0040] Figure 4 This is a schematic diagram showing two adjacent waveforms of a respiratory signal according to an embodiment of the present disclosure;
[0041] Figure 5 This is a schematic diagram illustrating the distance characteristics of a respiratory signal according to an embodiment of the present disclosure;
[0042] Figure 6 This is a schematic diagram illustrating the point-based characteristics of a respiratory signal according to an embodiment of the present disclosure;
[0043] Figure 7 This is a schematic diagram illustrating the area characteristics of a respiratory signal according to an embodiment of the present disclosure;
[0044] Figure 8 This is a schematic diagram of the structure of a device for determining the respiratory phase of a respiratory signal according to an embodiment of the present disclosure;
[0045] Figure 9 This is a schematic diagram illustrating the classification effect of respiratory feature values using embodiments of the present disclosure; and
[0046] Figure 10 This is a schematic diagram of an example of a computing device 100, which is part of the hardware portion of an apparatus for implementing a method for determining the respiratory phase of a respiratory signal according to an embodiment of the present disclosure. Detailed Implementation
[0047] To enable those skilled in the art to better understand this disclosure, the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.
[0048] Based on the natural understanding of respiratory movements, the general respiratory cycle consists of three phases: inhalation, exhalation, and end-exhalation. Figure 1 It is a time-series diagram of respiratory signals based on relevant technologies, where the horizontal axis represents time and the vertical axis represents the amplitude of respiration. For example... Figure 1 As shown, EX represents the exhalation phase, IN represents the inhalation phase, and EOE represents the end of exhalation. At the EOE stage, just before the end of exhalation, respiratory motion is minimal. For magnetic resonance imaging, if image data is acquired at the end of exhalation, motion artifacts are negligible. Therefore, the key issue is how to determine the end of exhalation in the respiratory signal.
[0049] An exemplary embodiment of this disclosure provides a method for determining the respiratory phase of respiratory movements. Figure 2 This is a flowchart of a method for determining the respiratory phase of a respiratory signal according to an embodiment of this disclosure. Figure 2 As shown, the method includes the following steps:
[0050] Step S22: Extract respiratory feature values from the respiratory signal.
[0051] Respiratory features such as distance, number of points, and area are extracted from the respiratory signal. These respiratory features are mainly used to indicate the waveform changes between two adjacent waveforms in the respiratory signal. In this way, by comparing the waveform changes of two adjacent waveforms and utilizing the differences in the curve shapes of inhalation and exhalation, the end-expiratory phase of a waveform can be determined.
[0052] Step S24: Based on the extracted respiratory feature values, train a respiratory signal model and use the respiratory signal model to determine the respiratory stage of the respiratory signal.
[0053] After extracting these respiratory features from respiratory signals, such as Pilot Tone respiratory signals, they are used to train a Support Vector Machine (SVM) model. Once the SVM model is trained, it can be used to determine the respiratory stage of the respiratory signal.
[0054] Figure 3 This is a flowchart of obtaining respiratory feature values according to an embodiment of this disclosure, such as... Figure 3 As shown, the method includes the following steps:
[0055] Step S32: Determine the extreme points and corresponding time points of two adjacent waveforms of the respiratory signal.
[0056] Figure 4 This is a schematic diagram showing two adjacent waveforms of a respiratory signal according to an embodiment of the present disclosure, wherein the horizontal axis represents time and the vertical axis represents the amplitude of the respiratory signal.
[0057] First, the peak points of the first and second waveforms are obtained and designated as the first maximum point max1 and the second maximum point max2, respectively. Next, the trough point between the two waveforms is obtained and designated as the first minimum point min1, and the trough point immediately following the second maximum point max2 is obtained and designated as the second minimum point min2.
[0058] Next, determine the time points corresponding to the first maximum point max1, the first minimum point min1, the second maximum point max2, and the second minimum point min2, respectively, namely, the time points of the first maximum point t_max1, the first minimum point t_min1, the second maximum point t_max2, and the second minimum point t_min2.
[0059] Step S34: Determine the distance feature value.
[0060] After determining the extreme points and their corresponding time points, the distance eigenvalues are calculated using the corresponding time points.
[0061] Figure 5 This is a schematic diagram illustrating the distance characteristics of a respiratory signal according to an embodiment of the present disclosure, such as... Figure 5 As shown, D1 represents the distance difference between the first minimum time point t_min1 and the first maximum time point t_max1, and D2 represents the distance difference between the second maximum time point t_max2 and the first minimum time point t_min1. The distance characteristic value can be calculated using the following formula:
[0062] f1=(t_min1–t_max1) / (t_max2–t_min1).
[0063] The above formula can also be expressed as f1 = D1 / D2, where f1 represents the distance feature value.
[0064] Step S36: Determine the feature value of the number of points.
[0065] After determining the extreme points and their corresponding time points, the characteristic values of the points are calculated using the corresponding time points.
[0066] Figure 6 This is a schematic diagram illustrating the point-based characteristics of a respiratory signal according to an embodiment of the present disclosure, such as... Figure 6 As shown, N1 and N2 represent the possible end-expiratory phases on the first waveform, and h represents the amplitude of the respiratory signal at the first maximum point.
[0067] Based on the predetermined amplitude ratio parameter p, the first maximum point max1, and the first minimum point min1, determine the start time point t_1 (not shown) of the rising waveform N1 on the first waveform up to the first maximum point max1, and determine the start time point t_2 (not shown) of the falling waveform N2 on the first waveform up to the first minimum point min_1. t_1 is called the first time, and t_2 is called the second time.
[0068] In one exemplary embodiment of this disclosure, the first time t_1 and the second time t_2 are calculated according to the following two formulas:
[0069] mag(t_1)=mag(t_max1)*p+mag(t_min1)*(1-p);
[0070] mag(t_2)=mag(t_max1)*(1-p)+mag(t_min1)*p.
[0071] Where mag(t) is the amplitude of the respiratory signal as a function of time t, and p is a predetermined amplitude scaling parameter.
[0072] For example, assuming the predetermined amplitude ratio parameter p is 80%, then the first time t_1 and the second time t_2 should satisfy:
[0073] mag(t_1)=mag(t_max1)*0.8+mag(t_min1)*0.2;
[0074] mag(t_2)=mag(t_max1)*0.2+mag(t_min1)*0.8.
[0075] After determining the possible start times of the end-expiratory phase, namely the first time point t_1 and the second time point t_2, the ratio of the difference between the first maximum time point t_max1 and the first time point t_1 to the difference between the first minimum time point t_min1 and the second time point t_2 is calculated. The result is used as the point count characteristic value. That is, the point count characteristic value is calculated using the following formula:
[0076] f2=(t_max1-t_1) / (t_min1-t_2).
[0077] Step S38: Determine the area characteristic value.
[0078] After determining the extreme points and their corresponding time points, the area characteristic values are calculated using the corresponding time points.
[0079] Figure 7 This is a schematic diagram illustrating the area characteristics of a respiratory signal according to an embodiment of the present disclosure, such as... Figure 7 As shown, connecting the first maximum point max1 and the second maximum point max2 forms a first polygon between the first maximum point max1, the second maximum point max2, and the first minimum point min1, with an area denoted as S1. Simultaneously, connecting the first minimum point min1 and the second minimum point min2 forms a second polygon between the first minimum point min1, the second maximum point max2, and the second minimum point min2, with an area denoted as S2.
[0080] In one embodiment of this disclosure, the area characteristic value is calculated using the following formula:
[0081] f3=s(t_max1,t_max2) / s(t_min1,t_min2).
[0082] Where f3 represents the area eigenvalue, and s represents the area of the polygon enclosed by the extreme points. In fact, f3 is... Figure 7 The ratio of the two areas S1 and S2 shown is S1 / S2.
[0083] The embodiments disclosed herein do not restrict the execution order of the above steps S34 to S38; they can be executed in different orders or simultaneously.
[0084] One embodiment of this disclosure also provides an apparatus for determining the respiratory phase of a respiratory signal. Figure 8 This is a schematic diagram of the structure of a device for determining the respiratory phase of a respiratory signal according to an embodiment of the present disclosure, as shown below. Figure 8 As shown, the device includes an extraction module 82 and a determination module 84.
[0085] Extraction module 82 is configured to extract respiratory feature values from the respiratory signal for training the SVM model. These respiratory features include, but are not limited to, distance feature values, point count feature values, and area feature values.
[0086] In one exemplary embodiment of this disclosure, the distance feature value is calculated using the following formula:
[0087] f1=(t_min1–t_max1) / (t_max2–t_min1).
[0088] f1 represents the distance feature value, t_min1 represents the trough point between two adjacent waveforms in the respiratory signal, t_max1 represents the peak point of the first waveform, and t_max2 represents the peak point of the second waveform.
[0089] In one exemplary embodiment of this disclosure, it is assumed that the time points at which the amplitudes of the maximum and minimum points are 80% are the first time t_1 and the second time t_2, respectively, which means that the conditions mag(t_1) = mag(t_max1)*0.8 + mag(t_min1)*0.2 and mag(t_2) = mag(t_max1)*0.2 + mag(t_min1)*0.8 are satisfied.
[0090] After calculating the first time point t_1 and the second time point t_2 using the above formulas, the point count characteristic value is calculated using the formula f2=(t_max1-t_1) / (t_min1-t_2), and the area characteristic value is calculated using the formula f3=s(t_max1,t_max2) / s(t_min1,t_min2). Here, f2 represents the point count characteristic value, f3 represents the area characteristic value, and s(t1,t2) is defined as the area of the closed polygon formed between all time points from time t1 to t2 and the waveform. For example, s(t_max1,t_max2) represents the area of the closed polygon formed between all time points from time t_max1 to t_max2, and s(t_min1,t_min2) represents the area of the closed polygon formed between all time points from time t_min1 to t_min2.
[0091] After extracting these respiratory feature values from the respiratory signal, module 84 uses these respiratory feature values to train the SVM model, i.e., the respiratory signal model. After the SVM model is trained, it can be used to determine the respiratory stage.
[0092] Figure 9 This is a schematic diagram illustrating the classification effect using respiratory feature values according to the embodiments of this disclosure. For example... Figure 9The three axes represent distance, point count, and area features, respectively. Solid square dots represent the peaks of the waveform after the input distance, point count, and area feature values are classified by a classifier. Solid circular dots represent the troughs of the waveform after the input distance, point count, and area feature values are classified by a classifier. This method identifies the end-expiratory phase of the respiratory signal, allowing the magnetic resonance imaging (MRI) system to be triggered at this phase, thus improving the quality of the MRI image and avoiding motion artifacts.
[0093] One illustrative embodiment of this disclosure also provides a magnetic resonance imaging (MRI) method. The MRI method is an image output method. First, the respiratory phase is determined using the method for determining the respiratory phase of a respiratory signal as disclosed in this disclosure. Second, MRI is performed on the examined area of the subject based on the determined respiratory phase. Specifically, the MRI system is triggered to scan at the end of the expiration phase of the respiratory signal, thereby avoiding motion artifacts in the MRI image and enabling the generation of high-precision images of living tissue.
[0094] One illustrative embodiment of this disclosure also provides a magnetic resonance imaging (MRI) system, including the apparatus for determining the respiratory phase and an imaging device provided herein. After a subject is placed in the MRI system, the respiratory phase of the respiratory signal, particularly the end-expiratory phase, is determined using the apparatus provided herein. Next, the MRI system is triggered to emit a magnetic resonance signal for scanning at the beginning of the end-expiratory phase, and an external magnetic field is modified using gradient coils to select a slice of the subject and generate a region encoding of the magnetic resonance signal. Then, the magnetic resonance signal is reconstructed, for example, by Fourier transform, to generate an image of the selected slice for medical diagnosis. The generation and detection of the magnetic resonance signal are achieved using a high-frequency system, which includes a transmitting antenna that injects high-frequency excitation pulses into the body of the subject and a receiving antenna that receives the resonance signal modulated by the subject.
[0095] This disclosure also provides a non-transitory readable storage medium having a program stored thereon, which, when executed, causes a processor to perform the method provided in this disclosure for determining the respiratory phase of a respiratory signal.
[0096] Figure 10 This is a schematic diagram of an example of a computing device 100, which is part of the hardware portion of an apparatus for implementing a method for determining the respiratory phase of a respiratory signal according to an embodiment of this disclosure. Figure 10As shown, the computing device 100 may include a CPU 1010 for overall control, a read-only memory (ROM) 1020 for storing system software, a random access memory (RAM) 1030 for storing write / read data, a storage unit 1040 for storing various programs and data, an input / output unit 1050 as an input / output interface, and a communication unit 1060 for implementing communication functions. Alternatively, the CPU 1010 may be replaced by a processor such as a microprocessor (MCU) or a programmable logic device (FPGA). The input / output unit 1050 may include various interfaces such as input / output interfaces (I / O interfaces), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), and a network interface. Those skilled in the art will understand that... Figure 10 The structure shown is for illustrative purposes only and does not limit the hardware configuration of the main control system. For example, the computing device 100 may also include a... Figure 10 The more or fewer components shown, or having the same Figure 10 The different configurations shown.
[0097] It should be noted that the CPU 1010 described above may include one or more processors, which, in this disclosure, may generally be referred to as "means for determining the breathing phase". The data processing circuitry may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuitry may be a single, independent processing module, or may be wholly or partially integrated into any other component of the computing device 100.
[0098] Storage unit 1040 can be used to store software programs and modules of application software, such as the program instruction / data storage device corresponding to the master control instruction calculation method described in this disclosure. CPU 1010 implements the aforementioned master control instruction calculation method by running the software programs and modules stored in storage unit 1040. Storage unit 1040 may include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, storage unit 1040 may further include memory remotely located relative to CPU 1010, and these remote memories can be connected to computing device 100 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.
[0099] The communication unit 1060 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computing device 100. In one example, the communication unit 1060 includes a network interface controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the communication unit 1060 may be a radio frequency (RF) module, used for wireless communication with the Internet.
[0100] In embodiments of this disclosure, a respiratory signal model is trained based on machine learning methods such as SVM by extracting feature values of respiratory signals, such as Pilot Tone respiratory signals, and further the respiratory phase of the respiratory signal is determined based on the respiratory signal model, thereby achieving the beneficial effects of robustness, high performance, and support for Pilot Tone applications.
[0101] The above description is only a preferred embodiment of this disclosure. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles of this disclosure, and these improvements and modifications should also be considered within the scope of protection of this disclosure.
Claims
1. A method for determining a respiratory phase of a respiratory signal, comprising: Extracting a distance feature value, a point feature value, and an area feature value from the breathing signal, wherein the distance feature value, the point feature value, and the area feature value are used to indicate a waveform change of two adjacent waveforms in the breathing signal; training a respiratory signal model according to the distance eigenvalue, the point number eigenvalue and the area eigenvalue, so as to determine the respiratory phase of the respiratory signal by using the respiratory signal model, The distance characteristic value is a ratio of a width of a second half of a first waveform of the two adjacent waveforms to a width of a first half of a second waveform of the two adjacent waveforms; and / or The point number characteristic value is the ratio of the width of a rising waveform section to the peak point of the first waveform of the two adjacent waveforms to the width of a falling waveform section to the trough point of the first waveform; and / or The area characteristic value is the ratio of the area formed by the two crest points of the two adjacent waveforms and the trough point between the two waveforms to the area formed by the second waveform.
2. The method according to claim 1, wherein: Before extracting the distance eigenvalue, the point number eigenvalue and the area eigenvalue from the breathing signal, the method further includes: Obtaining the two peak points in the two waveforms as the first maximum point and the second maximum point respectively; Obtaining a trough point between the two waveforms as a first minimum point, and obtaining a trough point after the second maximum point of the second waveform as a second minimum point; Determine the time points corresponding to the first maximum point, the first minimum point, the second maximum point, and the second minimum point, respectively, as the first maximum time point, the first minimum time point, the second maximum time point, and the second minimum time point.
3. The method according to claim 2, wherein: Extracting distance feature values from the breathing signal includes: Calculate the ratio of the difference between the first minimum time point and the first maximum time point to the difference between the second maximum time point and the first minimum time point; The calculated ratio is used as the extracted distance feature value.
4. The method according to claim 2, wherein: Extracting point feature values from the breathing signal includes: According to the predetermined amplitude ratio parameter and the first maximum point and the first minimum point, determine a starting time point of a rising waveform of the first waveform to the first maximum point and a starting time point of a falling waveform of the first waveform to the first minimum point, as a first time point and a second time point for determining the end-expiration starting time point of the breathing stage, respectively; Calculate the ratio of the difference between the first maximum time point and the first time point to the difference between the first minimum time point and the second time point; The calculated ratio is used as the extracted point feature value.
5. The method according to claim 4, wherein: The first time point and the second time point are determined according to the following two formulas: mag(t_1)=mag(t_max1)*p+mag(t_min1)*(1-p); mag(t_2)=mag(t_max1)*(1-p)+mag(t_min1)*p; Among them, t_1 is the first time point, t_max1 is the first maximum time point, p is the amplitude ratio parameter, t_min1 is the first minimum time point, mag is the amplitude function, and t_2 is the second time point.
6. The method according to claim 5, wherein: The amplitude ratio parameter p is equal to 80%.
7. The method according to claim 2, wherein: Extracting area feature values from the respiratory signal includes: Calculate the ratio of the area of a first polygon determined by the first maximum point, the second maximum point and the first minimum point to the area of a second polygon determined by the first minimum point, the second maximum point and the second minimum point; The calculated ratio is used as the extracted area feature value.
8. An apparatus for determining a respiratory phase of a respiratory signal, comprising: An extraction module is configured to extract a distance feature value, a point feature value, and an area feature value from the respiratory signal, wherein the distance feature value, the point feature value, and the area feature value are used to indicate a waveform change between two adjacent waveforms in the respiratory signal; a determination module configured to train a respiratory signal model according to the distance eigenvalue, the point number eigenvalue, and the area eigenvalue, so as to determine the respiratory phase of the respiratory signal using the respiratory signal model, The distance characteristic value is a ratio of a width of a second half of a first waveform of the two adjacent waveforms to a width of a first half of a second waveform of the two adjacent waveforms; and / or The point number characteristic value is the ratio of the width of a rising waveform section to the peak point of the first waveform of the two adjacent waveforms to the width of a falling waveform section to the trough point of the first waveform; and / or The area characteristic value is the ratio of the area formed by the two crest points of the two adjacent waveforms and the trough point between the two waveforms to the area formed by the second waveform.
9. The device according to claim 8, wherein: The extraction module comprises: A maximum value point acquisition unit is configured to acquire the two peak points in the two waveforms as a first maximum value point and a second maximum value point respectively; A minimum value point acquisition unit is configured to acquire a trough point between the two waveforms as a first minimum value point, and acquire a trough point after the second maximum value point of the second waveform as a second minimum value point; A time point determination unit is configured to determine the time points corresponding to the first maximum point, the first minimum point, the second maximum point and the second minimum point, respectively, as the first maximum time point, the first minimum time point, the second maximum time point and the second minimum time point.
10. The device according to claim 9, wherein: The extraction module comprises: a distance calculation unit, configured to calculate a ratio of a difference between the first minimum time point and the first maximum time point to a difference between the second maximum time point and the first minimum time point; The distance extraction unit is configured to use the calculated ratio as the extracted distance feature value.
11. The device according to claim 9, wherein: The extraction module comprises: a feature point determination unit, configured to determine, according to a predetermined amplitude ratio parameter and the first maximum point and the first minimum point, a starting time point of a rising waveform of the first waveform to the first maximum point and a starting time point of a falling waveform of the first waveform to the first minimum point, as a first time point and a second time point for determining an end-expiration start time point of the breathing stage, respectively; a feature point calculation unit, configured to calculate a ratio of a difference between the first maximum value time point and the first time point to a difference between the first minimum value time point and the second time point; The feature point extraction unit is configured to use the calculated ratio as the extracted point feature value.
12. The device according to claim 9, wherein: The extraction module comprises: an area calculation unit, configured to calculate a ratio of an area of a first polygon determined by the first maximum point, the second maximum point and the first minimum point to an area of a second polygon determined by the first minimum point, the second maximum point and the second minimum point; The area feature extraction unit is configured to use the calculated ratio as the extracted area feature value.
13. Magnetic resonance imaging method, comprising: Determining the breathing phase of the breathing signal according to the method of any one of claims 1 to 7; The examined part of the subject is subjected to magnetic resonance imaging according to the determined respiratory phase.
14. A magnetic resonance imaging system, comprising: An apparatus according to any one of claims 8 to 12, configured to determine a respiratory phase of a respiratory signal; The imaging device is configured to perform magnetic resonance imaging of a part to be examined of the subject according to the determined respiratory phase. 15 . A non-transitory readable storage medium having a program stored thereon, which, when executed, causes a processor to execute the method according to claim 1 .