A radio frequency starting method and system for medical cyclotron
By performing frequency domain analysis and Kalman filtering on the radio frequency digital signal of medical cyclotron, the analog-to-digital conversion noise problem in the RF start-up stage is solved, and the signal quality and tuning effect are improved.
Patent Information
- Application Number
- CN202510163061.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-14
- Publication Date
- 2025-04-15
- Estimated Expiration
- 2045-02-14
AI Technical Summary
During the RF startup stage, the medical cyclotron has reduced the detection efficiency of the RF digital signal due to analog-to-digital conversion noise problems, which affects the tuning effect.
By performing frequency domain analysis of RF digital signals, noise interference is filtered, and quantization noise is predicted and eliminated by Kalman filtering method to improve the quality of RF digital signals.
It improves the quality of the RF digital signal, enhances the accuracy of the frequency sweeping operation when starting the RF, and improves the tuning effect of the medical cyclotron.
Smart Images

Figure CN119676927B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of radio frequency starting of a medical cyclotron, and in particular to a radio frequency starting method and system for a medical cyclotron. Background Art
[0002] The medical cyclotron is a particle accelerator used to make radioactive drugs. When using it to accelerate particles, it is necessary to ensure that the frequency of the radio frequency circuit is the same as the fixed frequency of the particle confinement magnetic field, and resonance occurs, so that the particles can be accelerated better, the quality factor can be improved, and the secondary electron effect can be avoided. When the radio frequency circuit is started, due to particle collisions and induced currents inside the medical cyclotron, the D-type box of the medical cyclotron will heat up and expand, thereby changing the natural frequency of the confinement magnetic field current. Therefore, the frequency of the radio frequency circuit needs to be adjusted according to the actual heating situation of the medical cyclotron to ensure resonance.
[0003] In the process of reaching the resonance state through frequency sweeping, the RF circuit signal needs to be converted into a digital signal, and the digital signal needs to be detected to determine whether the frequency sweeping process is accurate. Since the frequency of the RF circuit needs to be adjusted continuously, quantization noise is easily generated during the analog-to-digital conversion of the signal, which affects the detection of the RF circuit signal, reduces the search efficiency, and affects the tuning effect of the medical cyclotron. Summary of the invention
[0004] In order to solve the above technical problems, the present application provides a radio frequency starting method and system for a medical cyclotron accelerator, and the technical solutions adopted are as follows:
[0005] In a first aspect, an embodiment of the present application provides a method for starting a radio frequency accelerator of a medical cyclotron, the method comprising the following steps:
[0006] A1, when the radio frequency pulse in the medical cyclotron is started, the frequency is swept, and the radio frequency signal that is continuously adjusted during the frequency sweeping process is converted into a radio frequency digital signal;
[0007] A2, analyzing the noise characteristics formed after conversion into RF digital signals, filtering out the noise interference in the RF digital signals; specifically:
[0008] (1) Calculate the noise intensity at the current moment based on the distribution difference between the main frequency and other frequencies of the local RF digital signal in the frequency domain;
[0009] (2) Divide the RF digital signal into multiple periodic signals, analyze the change of the main frequency of the local RF digital signal in the frequency domain at the corresponding time of the same element position in each periodic signal, and obtain the sweep rate sequence;
[0010] (3) Combine the frequency sweep rate sequence and the main frequency of the local RF digital signal in the frequency domain to predict the predicted digital signal at the current moment, and combine the noise intensity to calculate the Kalman gain at the current moment;
[0011] (4) Combining the predicted digital signal and the Kalman gain, the Kalman filter method is used to filter the RF digital signal at the current moment;
[0012] A3, controlling the starting frequency of the radio frequency circuit in the medical cyclotron according to the radio frequency digital signal filtered at the current moment.
[0013] Preferably, the main frequency is a frequency at which the radio frequency digital signals at the current moment and several moments before the current moment are converted into a signal with the largest amplitude in the spectrum, and the spectrum is the spectrum at the current moment.
[0014] Preferably, the noise intensity calculation process includes:
[0015] For each frequency other than the main frequency in the spectrum, obtain the frequency difference between the main frequency and each frequency, obtain the normalized value of each signal amplitude, and calculate the product result of the frequency difference and the normalized value;
[0016] The cumulative sum of the product results between the main frequency and all other frequencies except the main frequency is taken as the noise intensity at the current moment.
[0017] Preferably, the method for acquiring the sweep rate sequence includes:
[0018] The main frequency of the spectrum at the moment when each element position in each periodic signal corresponds to form the frequency sequence of each periodic signal;
[0019] The first-order derivative sequence of the frequency sequence of each periodic signal is obtained, and the mean sequence of the first-order derivative sequences of all periodic signals is recorded as the sweep rate sequence.
[0020] Preferably, the method for acquiring the predicted digital signal at the current moment is: using the predicted frequency at the current moment as the center frequency of the bandpass filter, using the RF digital signal at the current moment as the input of the bandpass filter, and outputting the predicted digital signal at the current moment.
[0021] Preferably, the calculation process of the predicted frequency at the current moment is:
[0022] Recording the preset time before the current time as the reference time of the current time;
[0023] Obtain the main frequency of the spectrum at the reference time; obtain the element value in the frequency change rate sequence at the reference time, and record it as the sweep frequency change rate at the reference time;
[0024] Obtain the time interval between the current moment and the start moment of the closest periodic signal, and record the element value of the time interval at the corresponding position in the sweep rate sequence as the predicted sweep rate of change at the current moment;
[0025] Among them, the predicted frequency at the current moment is calculated by combining the main frequency and sweep frequency change rate at the reference moment and the predicted sweep frequency change rate at the current moment. The specific calculation formula is: ;in, is the predicted frequency of h at the current moment, is the main frequency at reference time hS, is the sweep frequency change rate at the reference time hS, is the predicted sweep frequency change rate at the current time h, and S is the time interval from the reference time hS to the current time h.
[0026] Preferably, the calculation formula of the Kalman gain at the current moment is: ;in, is the Kalman gain at the current time h, is the minimum function, is the noise intensity at the reference time hS, is the sweep frequency change rate at the reference time hS, is the predicted sweep frequency change rate at the current time h.
[0027] Preferably, the calculation formula of the radio frequency digital signal at the current moment after filtering is: ;in, is the RF digital signal after filtering at the current moment h, is the actual RF digital signal at the current time h, is the predicted digital signal at the current time h, is the Kalman gain at the current time h.
[0028] Preferably, the specific method of controlling the starting frequency of the RF circuit in the medical cyclotron according to the RF digital signal filtered at the current moment is: inputting the RF digital signal filtered at the current moment into a PID controller, and outputting a control signal after the RF circuit frequency is adjusted.
[0029] In a second aspect, an embodiment of the present application further provides a radio frequency starting system for a medical cyclotron, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of any one of the radio frequency starting methods for a medical cyclotron described above are implemented.
[0030] This application has at least the following beneficial effects:
[0031] The present application aims at the problem of analog-to-digital conversion noise in the radio frequency signal obtained by frequency sweeping during the radio frequency start-up phase of a medical cyclotron, and proposes a radio frequency start-up method for a medical cyclotron, including: frequency domain conversion of a local radio frequency digital signal, and calculation of the noise intensity at the current moment according to the distribution difference between the main frequency and other frequencies after the frequency domain conversion, which is used to characterize the size of the quantization noise at the current moment. The radio frequency digital signal in the process of radio frequency start-up of the medical cyclotron is divided into multiple periodic signals, and the frequency change rate corresponding to the same element position of each periodic signal over time is feature extracted to obtain a frequency sweeping rate sequence, which is used to characterize the law of periodic change of the frequency of the radio frequency digital signal. Then, according to the frequency sweeping rate sequence and the main frequency of the local radio frequency digital signal in the frequency domain, the predicted digital signal at the current moment is calculated, and the Kalman gain at the current moment is calculated in combination with the noise intensity, which is used to evaluate whether the noise needs to be filtered at the current moment. Finally, the radio frequency digital signal at the current moment is filtered by the predicted digital signal and the Kalman gain to eliminate the influence of quantization noise on the radio frequency digital signal. This embodiment analyzes the noise generated during the analog-to-digital conversion of the RF digital signal during the RF startup of the medical cyclotron, filters the RF digital signal at the current moment, improves the quality of the RF digital signal, makes the frequency sweeping operation during RF startup more accurate, and improves the tuning effect of the medical cyclotron. BRIEF DESCRIPTION OF THE DRAWINGS
[0032] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0033] Figure 1 A flowchart of a radio frequency activation method for a medical cyclotron provided in one embodiment of the present application;
[0034] Figure 2 A schematic diagram of a filtering process for filtering noise interference in a radio frequency digital signal provided by an embodiment of the present application;
[0035] Figure 3 A schematic diagram of a filtering process provided for an embodiment of the present application. DETAILED DESCRIPTION
[0036] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following, in combination with the accompanying drawings and preferred embodiments, describes in detail the radio frequency activation method and system of a medical cyclotron accelerator proposed in the present application, its specific implementation, structure, features and effects. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0037] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
[0038] The specific scheme of a radio frequency starting method and system for a medical cyclotron provided by the present application is described in detail below with reference to the accompanying drawings.
[0039] An embodiment of the present application provides a radio frequency starting method and system for a medical cyclotron.
[0040] The present application filters the noise generated in the analog-to-digital conversion of radio frequency signals. Since the processing of digital signals has the disadvantage of high latency, the present embodiment records the radio frequency digital signals before the current moment as historical signals, extracts features from the historical signals, and performs real-time filtering on the current signals through the extracted features, thereby reducing the quantization noise of the radio frequency digital signals and solving the problem of poor tuning effect of a medical cyclotron caused by the analog-to-digital conversion noise of the radio frequency signals.
[0041] Specifically, the following radio frequency activation method for a medical cyclotron is provided, please refer to Figure 1 , the method comprises the following steps:
[0042] A1, when the radio frequency pulse in the medical cyclotron is started, the frequency is swept, and the radio frequency signal that is continuously adjusted during the frequency sweep is converted into a radio frequency digital signal; specifically:
[0043] The medical cyclotron accelerator will perform a frequency sweep operation when the radio frequency is started, and the frequency search parameters include: operating frequency, cold cavity resonance frequency, maximum radio frequency frequency, frequency search step, and frequency search interval.
[0044] When the radio frequency is started, this embodiment adjusts the radio frequency frequency of the accelerator to the maximum radio frequency frequency and keeps the frequency unchanged until the cavity approaches thermal equilibrium, that is, when the cavity temperature is stable, and then starts searching for the optimal radio frequency.
[0045] Since the cavity is heated, the actual resonant frequency of the cavity is greater than the resonant frequency of the cold cavity. If the frequency search interval is used as the time interval, the RF frequency is adjusted according to the frequency search step until the RF frequency is the same as the actual resonant frequency of the cavity. It should be noted that since the actual resonant frequency of the cavity is related to the cavity temperature, adjusting the RF frequency will affect the cavity temperature. Therefore, when searching for the optimal RF frequency, the RF frequency and the actual resonant frequency of the cavity decrease synchronously. Usually, the RF frequency decreases faster than the actual resonant frequency of the cavity, so resonance can be achieved through frequency sweeping.
[0046] Therefore, the frequency search parameters are obtained in this embodiment. In this embodiment, the operating frequency is 72.7 MHz, the maximum RF frequency is 73.1 MHz, the cold cavity resonance frequency is 72.9 MHz, the frequency search step is SP = 0.002 MHz, and the frequency search interval is TS = 1 second. At the same time, the maximum RF frequency and the cold cavity resonance frequency are averaged, and the average frequency is recorded as A = 73 MHz.
[0047] Furthermore, the RF digital signal after analog-to-digital conversion is recorded from the beginning of the RF startup. In this embodiment, B=32-bit sampling is used during analog-to-digital conversion, so the sampling frequency is set to MHz, the data is stored in the form of a vector, and the rth element of the vector represents the voltage of the digital signal at time r.
[0048] Meanwhile, when the frequency sweep operation starts with the cyclotron, the start time of the frequency sweep operation is recorded as the start time bg.
[0049] At this point, the above method can be used to obtain the radio frequency signal obtained by frequency sweeping when the radio frequency pulse of the medical cyclotron is started, as well as the converted radio frequency digital signal.
[0050] A2, analyzing the noise characteristics formed after conversion into RF digital signals, filtering out the noise interference in the RF digital signals; specifically:
[0051] For the noise generated after the RF signal is converted into digital form, this embodiment uses a filtering method to eliminate the noise, thereby improving the quality of the RF digital signal and preventing the quantization noise from affecting the detection of the RF digital signal.
[0052] Since the processing of digital signals has the disadvantage of high latency, this embodiment extracts historical signal features and filters the current signal based on the historical signal features. The noise features formed after conversion into RF digital signals are analyzed, and the filtering process diagram of filtering the noise interference in the RF digital signals is shown in FIG. Figure 2 As shown, specifically:
[0053] (1) Based on the distribution difference between the main frequency and other frequencies of the local RF digital signal in the frequency domain, the noise intensity at any moment is calculated; specifically:
[0054] Since the RF analog signal of the medical cyclotron is a sinusoidal waveform, the RF digital signal obtained by scanning the frequency should have a main frequency with a higher signal amplitude and multiple other frequencies with lower signal amplitudes on the spectrum diagram after analog-to-digital conversion. The main frequency and its vicinity contain most of the information of the RF signal, and the remaining other frequencies mainly contain the information of quantization noise.
[0055] Accordingly, the present application calculates the noise intensity at any moment by analyzing the distribution difference between the main frequency and other frequencies of the local RF digital signal in the frequency domain, which is used to characterize the noise intensity contained in other components other than the main frequency component containing the RF signal information.
[0056] In this application, the noise intensity in the local RF digital signal is analyzed at any time. In this embodiment, the local RF digital signal at the current time h is analyzed, specifically:
[0057] The RF digital signal at the current moment h and the previous S moments is obtained, and the obtained RF digital signal is converted into a frequency domain to obtain a spectrum, which is recorded as the spectrum at the current moment. Each frequency in the spectrum corresponds to a signal amplitude. In this embodiment, S= Microseconds, where A is the mean frequency of the spectrum. If S is a decimal, it is rounded. This embodiment uses a fast Fourier transform algorithm to perform frequency domain conversion to obtain the spectrum. The fast Fourier transform algorithm is a commonly used technology in the field of signal processing and will not be described in detail.
[0058] In particular, if the number of moments before the current moment h is less than S, all RF digital signals from the initial start moment bg to the current moment h are used as inputs of the fast Fourier transform algorithm to calculate the spectrum of the current moment h.
[0059] Among them, the frequency with the largest signal amplitude in the spectrum of h at the current moment is found as the main frequency of the spectrum of h at the current moment.
[0060] This embodiment mainly filters the quantization noise in the digital-to-analog conversion process. Since in the spectrum, the RF signal information is mainly contained in the main frequency and its nearby frequencies, the larger the proportion of the main frequency in the spectrum to the entire spectrum, the more concentrated the frequency, and the smaller the quantization noise in the RF digital signal.
[0061] Therefore, this embodiment calculates the noise intensity at the current moment h based on the difference between the main frequency and other frequencies. The more the frequencies are concentrated on the main frequency, the smaller the noise intensity. The calculation formula for the noise intensity at the current moment h is: ; In the formula, is the noise intensity at the current time h, is the frequency number of the h spectrum at the current moment, is the normalized value of the signal amplitude corresponding to the uth frequency in the h spectrum at the current moment, , are the uth frequency and the main frequency in the spectrum h at the current moment, respectively, and norm is a normalization function. This embodiment adopts the maximum and minimum normalization method, which is a well-known technology and will not be described in detail.
[0062] It should be understood that the greater the frequency difference between a frequency component and the main frequency, and the greater the signal proportion of the frequency component, the less the frequency at that moment is concentrated on the main frequency, and the greater the noise intensity.
[0063] (2) Divide the RF digital signal into multiple periodic signals, analyze the change of the main frequency of the local RF digital signal in the frequency domain at the corresponding time of the same element position in each periodic signal, and obtain the sweep rate sequence; specifically:
[0064] Since the best RF frequency is searched by frequency sweeping in this embodiment, the frequency change in the frequency sweeping process is periodically changed with the frequency search interval TS as a period. However, this embodiment aims to extract the frequency domain characteristics of the RF digital signal for filtering to eliminate noise, so this embodiment extracts the periodic frequency change characteristics during the frequency sweeping operation.
[0065] Based on this, this embodiment constructs a sequence with a time length of TS, which is recorded as a frequency sweep rate sequence. The sequence is used to characterize the periodic variation law of the main frequency of the radio frequency digital signal.
[0066] The acquisition process of the sweep rate sequence is as follows:
[0067] Starting from the start time bg of the RF digital signal, the RF digital signal is divided into multiple periodic signals with TS as the period, the time length of each periodic signal is equal to TS, and the time length of the last periodic signal is less than or equal to TS. Assume that the RF digital signal is finally divided into N segments, of which the nth segment is recorded as the nth periodic signal.
[0068] For the mth element position in the nth periodic signal, the main frequency of the spectrum at the time corresponding to the element position is obtained, and the main frequencies of the spectrum at the time corresponding to all element positions in the nth periodic signal are combined to form the frequency sequence of the nth periodic signal.
[0069] Further, the frequency sequence of the nth periodic signal is taken as input, and the first-order derivative calculation method is adopted to output the first-order derivative sequence of the nth periodic signal, which is recorded as the frequency change rate sequence of the nth periodic signal; the first-order derivative calculation method is a commonly used technology in the field of data processing and will not be repeated here.
[0070] In this way, N frequency change rate sequences are obtained, and the average sequence of the N frequency change rate sequences is calculated and recorded as the sweep rate sequence.
[0071] The sweep rate sequence represents the main frequency variation law of the local RF digital signal in the frequency domain within each sweep cycle. For different medical cyclotrons, their sweep rate sequences are different because of the differences in their sweep frequency adjustment devices.
[0072] In particular, when the RF digital signal starts from the start time bg and the time length is less than TS, the frequency sweep rate sequence is not calculated at this time, and the RF digital signal at this time is not filtered.
[0073] (3) Combine the frequency sweep rate sequence and the main frequency of the local RF digital signal in the frequency domain to predict the predicted digital signal at the current moment, and combine the noise intensity to calculate the Kalman gain at the current moment; specifically:
[0074] At the current moment h, due to the disadvantage of high delay in the processing of digital signals, in order to reduce the delay, the main frequency of the RF digital signal at the current moment h cannot be directly obtained. Therefore, this embodiment combines the sweep rate sequence and the main frequency of the local RF digital signal to calculate the predicted frequency at the current moment h, specifically:
[0075] First, record the Sth moment before the current moment h as the reference moment hS of the current moment h; obtain the main frequency of the reference moment , and at the same time obtain the element value in the frequency change rate sequence at the reference time, recorded as the sweep frequency change rate at the reference time In this embodiment, the Sth moment before the current moment h is used as the preset moment before the current moment h.
[0076] Get the time interval between the current moment h and the start moment of the nearest periodic signal, and record the element value of the time interval number at the corresponding position in the sweep rate sequence as the predicted sweep rate of change at the current moment h.
[0077] Since the time from hS to h is short, it is approximately considered that the change of the main frequency between the two moments is a uniform rate change. Therefore, this embodiment calculates the predicted frequency at the current moment h based on the main frequency of the local RF digital signal in the frequency domain and the sweep rate sequence. The specific calculation formula is: ;in, is the predicted frequency of h at the current moment, is the main frequency at reference time hS, is the sweep frequency change rate at the reference time hS, is the predicted sweep frequency change rate at the current time h, and S is the time interval from the reference time hS to the current time h.
[0078] It should be noted that if the time from moment hS to moment h is short, it can be considered that the change of the main frequency between the two moments is a uniform rate change. The predicted frequency of the current moment h is estimated by the main frequency at moment hS and the sweep frequency change rate, so that the predicted frequency of the current moment h is more accurate, avoiding the disadvantage of high delay in the digital signal processing process.
[0079] This embodiment predicts that the main frequency of h at the current moment is , which is the predicted frequency of h at the current moment.
[0080] In order to avoid the delay caused by digital signal processing, the present application optimizes the bandpass filter according to the predicted frequency of the current moment h, and records the signal filtered by the bandpass filter as the predicted digital signal, which is used to characterize the predicted RF digital signal at the current moment h without noise influence.
[0081] In this embodiment, the predicted frequency at the current moment h is used as the center frequency of the bandpass filter, the RF digital signal at the current moment h is used as the input of the bandpass filter, and the predicted digital signal at the current moment h is output. The bandwidth of the bandpass filter is set to , SP is the frequency search step size, and bandpass filter is a commonly used technology in the field of signal processing and will not be elaborated on here.
[0082] This embodiment uses the Kalman filter method to filter the radio frequency digital signal to remove the quantization noise. Since the digital signal processing delay is relatively long, this embodiment uses the historical signal to predict the current signal state.
[0083] This embodiment calculates the predicted digital signal for the current moment h as a priori signal through the Kalman filtering method, and further needs to calculate the Kalman gain of the current moment h. Generally, the greater the noise intensity of the current moment h, the smaller the Kalman gain of the current moment h should be. The smaller the Kalman gain means that the filter is less dependent on the current measurement value to update the state estimate, thereby removing the noise in the RF digital signal. Among them, Kalman filtering is a common technology in the field of signal processing and will not be repeated.
[0084] During the RF startup phase, the system's frequency sweep operation causes the RF signal's frequency to change continuously, which can cause the RF signal's frequency to become unstable, which in turn increases the quantization noise. Therefore, the more drastic the RF signal's frequency change, the greater the noise at the corresponding moment.
[0085] Based on this, this embodiment analyzes the frequency change between time hS and time h. If the signal frequency change speed at time h is faster than that at time hS, the noise intensity at time h is more likely to be greater than that at time hS. Otherwise, the noise intensity at time h is more likely to be less than that at time hS. At the same time, combined with the noise intensity, the calculation formula for calculating the Kalman gain at the current time h is: ;in, is the Kalman gain at the current time h, is the minimum function, is the noise intensity at the reference time hS, is the sweep frequency change rate at the reference time hS, is the predicted sweep frequency change rate at the current time h.
[0086] It should be understood that if the ratio of the predicted frequency sweeping change rate at the current moment h to the frequency sweeping change rate at the moment hS is greater than 1, it means that the frequency change caused by the frequency sweeping is intensified, and the noise intensity at the current moment h is greater than the noise intensity at the moment hS; conversely, if the ratio is less than or equal to 1, it means that the frequency change caused by the frequency sweeping is slowed down or unchanged, and the noise intensity at the current moment h is less than or equal to the noise intensity at the moment hS; therefore, the noise intensity at the moment hS is multiplied by the ratio to represent the predicted noise intensity at the current moment h; in particular, when the noise intensity at the moment hS multiplied by the ratio is greater than 1, it means that the noise of the RF digital signal directly obtained after the analog-to-digital conversion is very serious. At this time, the predicted digital signal is directly used as the filtered RF digital signal, so the Kalman gain is taken as 1 at this time.
[0087] (4) Combining the predicted digital signal and the Kalman gain, the Kalman filter method is used to filter the RF digital signal at the current moment; specifically:
[0088] Finally, the Kalman filter method is used to filter the RF digital signal. The specific method is as follows: Figure 3 shown.
[0089] exist Figure 3 In the method, the predicted digital signal and Kalman gain at the current moment are calculated according to the RF digital signal at the moment hS, and then the RF digital signal at the moment h is combined with the Kalman filter method to complete the denoising of the RF digital signal at the moment h.
[0090] The specific calculation formula for filtering the RF digital signal at the current time h is: ;in, is the RF digital signal after filtering at the current moment h, is the actual RF digital signal at the current time h, is the predicted digital signal at the current time h, is the Kalman gain at the current time h.
[0091] It should be understood that the predicted digital signal is used as the prior estimation value, the actual RF digital signal is used as the actual measurement value, and the Kalman gain is used as the weight to complete the filtering, and the RF digital signal noise caused by the frequency conversion of the sweep frequency is filtered out, so as to obtain a more accurate RF digital signal. Among them, Kalman filtering is a commonly used technology in the field of signal processing, and its detailed calculation process is not repeated in this embodiment.
[0092] A3, controlling the starting frequency of the radio frequency circuit in the medical cyclotron according to the radio frequency digital signal filtered at the current moment, specifically:
[0093] The original radio frequency digital signal is further replaced by the radio frequency digital signal after filtering at the current moment, and the starting frequency of the radio frequency circuit in the medical cyclotron is controlled, specifically:
[0094] The filtered RF digital signal at the current moment is input into the PID controller, and the output is the control signal after the RF circuit frequency is adjusted. By using the control signal to control the RF circuit frequency, it can output an accurate RF signal when the RF of the medical cyclotron is working, avoiding under- or over-modulation of the RF signal, and ensuring the accuracy of frequency control during the frequency sweeping process. Among them, the PID controller is a common technology in the field of digital control and will not be described in detail.
[0095] By ensuring the accuracy of frequency control during the frequency sweep process, the frequency of the radio frequency circuit can be made the same as the fixed frequency of the particle confinement magnetic field, thereby causing resonance, which can better accelerate the particles in the medical cyclotron.
[0096] Finally, this embodiment solves the problem of poor tuning effect of the medical cyclotron due to analog-to-digital conversion noise of the radio frequency signal during the radio frequency startup process of the medical cyclotron.
[0097] Based on the same inventive concept as the above method, an embodiment of the present application also provides a radio frequency starting system for a medical cyclotron, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of a radio frequency starting method for a medical cyclotron described in any one of the above methods are implemented.
[0098] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.
[0099] It should be noted that, unless otherwise specified and limited, terms such as "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such article or device. In the absence of further restrictions, an element defined by the sentence "including one..." does not exclude the existence of other identical elements in the article or device including the element. In addition, the term "and\or" used herein includes any and all combinations of one or more related listed items.
[0100] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not invented by the present application.
[0101] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.
Claims
1. A radio frequency starting method for a medical cyclotron, characterized in that: The method comprises the following steps: A1, when the radio frequency pulse in the medical cyclotron is started, the frequency is swept, and the radio frequency signal that is continuously adjusted during the frequency sweeping process is converted into a radio frequency digital signal; A2, analyzing the noise characteristics formed after conversion into RF digital signals, filtering out the noise interference in the RF digital signals; specifically: (1) Calculate the noise intensity at the current moment based on the distribution difference between the main frequency and other frequencies of the local RF digital signal in the frequency domain; (2) Divide the RF digital signal into multiple periodic signals, analyze the change of the main frequency of the local RF digital signal in the frequency domain at the corresponding time of the same element position in each periodic signal, and obtain the sweep rate sequence; (3) Combine the frequency sweep rate sequence and the main frequency of the local RF digital signal in the frequency domain to predict the predicted digital signal at the current moment, and combine the noise intensity to calculate the Kalman gain at the current moment; (4) Combining the predicted digital signal and the Kalman gain, the Kalman filter method is used to filter the RF digital signal at the current moment; A3, controlling the starting frequency of the radio frequency circuit in the medical cyclotron according to the radio frequency digital signal filtered at the current moment.
2. The radio frequency starting method of a medical cyclotron according to claim 1, characterized in that: The main frequency is the frequency at which the radio frequency digital signals at the current moment and several moments before the current moment are converted into the signal with the largest amplitude in the spectrum, and the spectrum is the spectrum at the current moment.
3. The radio frequency starting method of a medical cyclotron according to claim 2, characterized in that: The calculation process of the noise intensity includes: For each frequency other than the main frequency in the spectrum, obtain the frequency difference between the main frequency and each frequency, obtain the normalized value of each signal amplitude, and calculate the product result of the frequency difference and the normalized value; The cumulative sum of the product results between the main frequency and all other frequencies except the main frequency is taken as the noise intensity at the current moment.
4. The radio frequency starting method of a medical cyclotron according to claim 1, characterized in that: The method for acquiring the sweep rate sequence includes: The main frequency of the spectrum at the moment when each element position in each periodic signal corresponds to form the frequency sequence of each periodic signal; The first-order derivative sequence of the frequency sequence of each periodic signal is obtained, and the mean sequence of the first-order derivative sequences of all periodic signals is recorded as the sweep rate sequence.
5. The radio frequency starting method of a medical cyclotron according to claim 1, characterized in that: The method for acquiring the predicted digital signal at the current moment is: using the predicted frequency at the current moment as the center frequency of the bandpass filter, using the radio frequency digital signal at the current moment as the input of the bandpass filter, and outputting the predicted digital signal at the current moment.
6. A radio frequency starting method for a medical cyclotron as claimed in claim 5, characterized in that: The calculation process of the predicted frequency at the current moment is: Recording the preset time before the current time as the reference time of the current time; Obtain the main frequency of the spectrum at the reference time; obtain the element value in the frequency change rate sequence at the reference time, and record it as the sweep frequency change rate at the reference time; Obtain the time interval between the current moment and the start moment of the closest periodic signal, and record the element value of the time interval at the corresponding position in the sweep rate sequence as the predicted sweep rate of change at the current moment; Among them, the predicted frequency at the current moment is calculated by combining the main frequency and sweep frequency change rate at the reference moment and the predicted sweep frequency change rate at the current moment. The specific calculation formula is: ;in, is the predicted frequency of h at the current moment, is the main frequency at reference time hS, is the sweep frequency change rate at the reference time hS, is the predicted sweep frequency change rate at the current time h, and S is the time interval from the reference time hS to the current time h.
7. A radio frequency starting method for a medical cyclotron as claimed in claim 6, characterized in that: The calculation formula of the Kalman gain at the current moment is: ;in, is the Kalman gain at the current time h, is the minimum function, is the noise intensity at the reference time hS, is the sweep frequency change rate at the reference time hS, is the predicted sweep frequency change rate at the current time h.
8. The radio frequency starting method of a medical cyclotron as claimed in claim 1, characterized in that: The calculation formula of the RF digital signal at the current moment after filtering is: ;in, is the RF digital signal after filtering at the current moment h, is the actual RF digital signal at the current time h, is the predicted digital signal at the current time h, is the Kalman gain at the current time h.
9. The radio frequency starting method of a medical cyclotron according to claim 1, characterized in that: The specific method for controlling the starting frequency of the radio frequency circuit in the medical cyclotron according to the radio frequency digital signal filtered at the current moment is: inputting the radio frequency digital signal filtered at the current moment into the PID controller, and outputting the control signal after the radio frequency circuit frequency is adjusted.
10. A radio frequency starting system for a medical cyclotron, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the radio frequency starting method for a medical cyclotron are implemented as described in any one of claims 1-9.
Citation Information
Patent Citations
Method and device for eliminating amplitude-phase control interference in self-excitation mode
CN101883469A
Medical cyclotron beam measurement signal protection method and system
CN119064980A