Doppler power spectrum envelope extraction method, system, ultrasonic equipment and medium

Abstract

The present invention is applicable to the field of medical equipment technology, and provides a method, system, ultrasonic equipment and medium for extracting the envelope of a Doppler power spectrum. The method includes continuously acquiring the Doppler power spectrum at each moment; weighting the Doppler power spectrum at the current moment and the historical moment to obtain a weighted Doppler power spectrum; denoising the weighted power spectrum by noise estimation to obtain a denoised Doppler power spectrum and performing smoothing, identifying the maximum frequency position of the smoothed Doppler power spectrum obtained by smoothing to obtain a candidate maximum frequency position; correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency position of the Doppler power spectrum at the current moment and the historical moment to obtain its maximum frequency position; connecting the maximum frequency positions of the Doppler power spectrum at each moment to obtain a maximum frequency envelope for display. The present invention solves the problem that the maximum frequency envelope cannot be extracted in real time and accurately under noise interference.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment, and in particular to a method, system, ultrasonic equipment and medium for extracting the envelope of a Doppler power spectrum. Background Art

[0002] Ultrasonic spectrum Doppler technology is a widely used medical diagnostic tool to evaluate the speed and direction of blood flow. Based on the principle of Doppler effect, the movement speed of reflectors such as red blood cells in the blood flow will cause the echo frequency to shift, and this frequency shift is proportional to the movement speed of the reflector. By analyzing the Doppler power spectrum of the ultrasound signal, doctors can obtain blood flow velocity information. The maximum frequency envelope of the Doppler power spectrum (hereinafter referred to as the envelope) refers to the maximum frequency curve in the Doppler power spectrum. Figure 4 As shown, the maximum frequency envelope usually corresponds to the maximum velocity information of blood flow, which has important clinical value in the diagnosis of pathological conditions such as vascular stenosis and valvular disease.

[0003] Among them, the measurement of envelope is of great significance in many fields, including:

[0004] Blood flow velocity assessment: By determining the maximum blood flow velocity through the envelope line, doctors can determine whether there are abnormal blood flow characteristics and help identify diseases such as vascular stenosis, arteriosclerosis or heart valve disease.

[0005] Hemodynamic analysis: The shape and height of the envelope reflect the dynamic characteristics of blood flow, such as pulsation and periodicity, and can be used to evaluate the functional state of the heart and blood vessels, especially in the evaluation of cardiac function and blood circulation.

[0006] Diagnostic reference: The envelope measurement results are often used as a basis for diagnosis, especially in cardiac ultrasound examinations, where the envelope characteristics can provide doctors with clear diagnostic references.

[0007] Treatment monitoring: By continuously monitoring changes in the envelope, doctors can evaluate the effectiveness of treatment and the progression of the disease, and make timely treatment adjustments.

[0008] Although envelope measurement is widely used in clinical practice, existing implementation methods have many limitations. Common solutions can be divided into two categories: threshold-based and image processing-based. Threshold-based methods rely on preset empirical values ​​or equipment parameters to distinguish valid signals from noise by setting fixed thresholds. However, the accuracy of such methods is low, especially through simple fixed threshold judgments, which often result in the maximum frequency value of the envelope being too small and difficult to be consistent with the doctor's visual perception. Although some complex threshold methods attempt to introduce integral curve analysis, noise estimation and other means to improve accuracy, this method has obvious shortcomings, especially in the calculation of the maximum frequency, which is often lower than the actual value and cannot accurately reflect the doctor's visual perception, making its adaptability in different scenarios still poor.

[0009] Another common method is based on image processing, including binarization, edge (gradient) detection, connected domain analysis and other technologies. This method determines the envelope by processing the image features of the power spectrum. However, this method has high computational complexity and long processing time, which makes it difficult to meet the needs of real-time clinical applications. At the same time, the method of distinguishing noise from effective signals is too dependent on the size of the region and has poor robustness.

[0010] In addition, existing denoising methods perform poorly when faced with noise at different gains and cannot effectively adapt to changes in signal strength in various clinical scenarios. When the probe moves, the pressure changes, or the patient moves, signal instability will appear as short-term abnormal brightness changes on the power spectrum. These situations further aggravate the errors of existing algorithms, especially when dealing with envelope mutations. Existing algorithms have difficulty accurately capturing these rapidly changing signal characteristics, further affecting the accuracy of the envelope. Furthermore, some methods rely on users to manually correct the envelope, which reduces the degree of automation and is not conducive to the efficiency and real-time performance of clinical operations.

[0011] In summary, the existing measurement method of the maximum velocity envelope of the Doppler power spectrum has a lot of room for improvement in terms of accuracy, real-time and noise resistance. There is an urgent need for a more effective technical means to overcome the limitations of the existing method. Summary of the invention

[0012] Based on this, the purpose of the present invention is to provide a Doppler power spectrum envelope extraction method, system, ultrasonic equipment and medium, so as to fundamentally solve the existing problem that the maximum frequency envelope cannot be extracted in real time and accurately under noise interference.

[0013] According to an embodiment of the present invention, a method for extracting the envelope of a Doppler power spectrum is provided, wherein:

[0014] Continuously obtain the Doppler power spectrum at each moment;

[0015] Performing weighted processing on the Doppler power spectra at the current moment and at multiple historical moments to obtain a weighted Doppler power spectrum at the current moment, wherein the weighted Doppler power spectrum is a Doppler power spectrum weighted by historical information;

[0016] Performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment, wherein the denoised Doppler power spectrum is a Doppler power spectrum obtained after denoising the weighted Doppler power spectrum;

[0017] The denoised Doppler power spectrum at the current moment is smoothed to obtain a smoothed Doppler power spectrum at the current moment, wherein the smoothed Doppler power spectrum is a Doppler power spectrum obtained by smoothing the denoised Doppler power spectrum;

[0018] The maximum frequency position of the smoothed Doppler power spectrum at the current moment is identified to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment;

[0019] According to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments, the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected to obtain the maximum frequency position of the Doppler power spectrum at the center moment, wherein the center moment is the moment at the center position between the current moment and multiple historical moments, and the maximum frequency position is the frequency position determined after the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected;

[0020] The maximum frequency positions of the corrected Doppler power spectra at each moment are connected to obtain the maximum frequency envelope of the Doppler power spectrum, and the maximum frequency envelope of the Doppler power spectrum is displayed.

[0021] In addition, the method for extracting the envelope of a Doppler power spectrum according to the above embodiment of the present invention may also have the following additional technical features:

[0022] Furthermore, the step of performing weighted processing on the Doppler power spectra at the current moment and the Doppler power spectra at multiple historical moments to obtain the weighted Doppler power spectrum at the current moment includes:

[0023] The standard deviation coefficient is determined based on the scanning speed in Doppler mode;

[0024] Determine the weighting coefficients of the current moment and each historical moment according to the time interval between each historical moment and the current moment and the standard deviation coefficient;

[0025] The weighted Doppler power spectrum is obtained by performing weighted processing according to the Doppler power spectrum at the current moment and each historical moment and the weighting coefficients at the current moment and each historical moment.

[0026] Further, the step of performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising processing on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment includes:

[0027] Calculate the histogram of the weighted Doppler power spectrum at the current moment;

[0028] Accumulating the calculated histograms to obtain an accumulated histogram;

[0029] Calculate the local slope of each point in the cumulative histogram;

[0030] According to the local slope of each point, the pixel value corresponding to the point with the largest local slope is determined as the noise intensity;

[0031] All parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity are denoised to obtain the denoised Doppler power spectrum at the current moment.

[0032] Furthermore, the step of performing maximum frequency position identification on the smoothed Doppler power spectrum at the current moment to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment includes:

[0033] Performing gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment to obtain the gradient value of each frequency;

[0034] According to the gradient values ​​of each frequency, a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​are determined as candidate points;

[0035] Calculate the pixel mean in the local area below each candidate point in turn;

[0036] Determine the confidence level of each candidate point as the maximum frequency position based on the forward gradient value and pixel mean of each candidate point;

[0037] According to the confidence of each candidate point, the candidate point with the largest confidence is determined as the candidate maximum frequency position of the Doppler power spectrum at the current moment.

[0038] Further, the step of correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment includes:

[0039] Determine a central moment located at a central position between the current moment and the multiple historical moments according to the current moment and the multiple historical moments;

[0040] Calculate the coordinate differences between the candidate maximum frequency position of the Doppler power spectrum at the center moment and the candidate maximum frequency position of the Doppler power spectrum at the previous moment and the next moment respectively;

[0041] Determine a coordinate difference value according to the two coordinate differences, and judge whether the coordinate difference value is less than a difference threshold;

[0042] When it is determined that the coordinate difference is less than the difference threshold, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the corresponding candidate maximum frequency position.

[0043] Furthermore, the step of determining whether the coordinate difference is less than a difference threshold value further includes:

[0044] When it is determined that the coordinate difference is not less than the difference threshold, the previous moment and the next moment of the center moment are respectively used as the new center moment;

[0045] Calculate the coordinate differences between the candidate maximum frequency positions of the Doppler power spectra at the two new center moments and the corresponding candidate maximum frequency positions of the Doppler power spectra at the previous moment and the next moment respectively;

[0046] Determine two new coordinate difference values ​​according to the two coordinate differences respectively determined at the two new center moments, and judge whether the two new coordinate difference values ​​are both less than the difference threshold;

[0047] When it is determined that the differences of the two new coordinate differences are both less than the difference threshold, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the corresponding candidate maximum frequency position.

[0048] Furthermore, the step of determining whether the difference values ​​of the two new coordinate differences are both less than the difference threshold value further includes:

[0049] When it is determined that any new coordinate difference is not less than the difference threshold, the mean value between the candidate maximum frequency positions of the Doppler power spectrum at the previous moment and the next moment of the center moment is calculated to obtain the mean maximum frequency position of the Doppler power spectrum at the center moment;

[0050] Traverse each candidate point in the Doppler power spectrum at the center time, and determine whether there is a candidate point whose coordinate difference between each candidate point and the mean maximum frequency position is less than the difference threshold;

[0051] If yes, then determine that the maximum frequency position of the Doppler power spectrum at the center time is the target candidate point that is closest to the difference threshold among the candidate points that are smaller than the difference threshold;

[0052] If not, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the mean maximum frequency position.

[0053] Furthermore, the calculation formula for determining the weighted coefficients of the current moment and each historical moment according to the time interval between each historical moment and the current moment and the standard deviation coefficient is:

[0054] ;

[0055] in, For the moment The weighting coefficient of For the current moment, is the standard deviation coefficient, is an exponential function with the natural base e as base;

[0056] The calculation formula for obtaining the weighted Doppler power spectrum by weighted processing based on the Doppler power spectrum at the current moment and each historical moment and the weighted coefficients at the current moment and each historical moment is:

[0057] ;

[0058] in, For the current moment The weighted Doppler power spectrum of is the number of historical moments, For the moment The Doppler power spectrum.

[0059] Furthermore, the calculation formula for calculating the histogram of the weighted Doppler power spectrum at the current moment is:

[0060] ;

[0061] in, is the pixel value The number of occurrences of is the weighted Doppler power spectrum, is a logical operator, is the pixel value;

[0062] The calculation formula for obtaining the accumulated histogram by accumulating the calculated histogram is:

[0063] ;

[0064] in, is the pixel value The cumulative histogram of ;

[0065] The calculation formula for calculating the local slope of each point in the cumulative histogram is:

[0066] ;

[0067] in, is the pixel value The local slope of is the pixel distance in the histogram;

[0068] The calculation formula for determining the pixel value corresponding to the point with the largest local slope according to the local slope of each point as the noise intensity is:

[0069] ;

[0070] in, is the pixel value corresponding to the noise intensity, is the local slope The k value when the maximum value is obtained;

[0071] The calculation formula for denoising all parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity to obtain the denoised Doppler power spectrum at the current moment is:

[0072] ;

[0073] in, is the denoised Doppler power spectrum, is the weighted Doppler power spectrum.

[0074] Furthermore, the calculation formula for smoothing the denoised Doppler power spectrum at the current moment to obtain the smoothed Doppler power spectrum at the current moment is:

[0075] ;

[0076] ;

[0077] in, For the current moment The smoothed Doppler power spectrum of is the weight array of the Hanning smoothing filter, is the first Point weight value, for The denoised Doppler power spectrum at time instant.

[0078] Furthermore, the calculation formula for obtaining the gradient value of each frequency by performing gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment is:

[0079] ;

[0080] in, For the current moment In frequency The gradient value at is the gradient detection operator, is the gradient detection operator The value of a point, For the current moment The smoothed Doppler power spectrum at frequency The value at

[0081] The calculation formula for determining a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​as candidate points according to the gradient values ​​of each frequency is:

[0082] ;

[0083] ;

[0084] ;

[0085] in, For the gradient value Sort by maximum value. is the frequency position coordinate array after the forward gradient values ​​are sorted from large to small, is the gradient value array after the forward gradient values ​​are sorted from large to small, For the current moment The one with the largest positive gradient The gradient intensity sequence of points; For the current moment The one with the largest positive gradient The frequency position coordinate sequence of points, is the number of valid candidate points, which is also the preset number;

[0086] The calculation formula for sequentially calculating the pixel mean in the local area below each candidate point is:

[0087] ;

[0088] in, For the current moment No. The average pixel value in the local area below the candidate point, is the local range parameter, For the current moment The denoised Doppler power spectrum is The pixel value of each candidate point;

[0089] The calculation formula for determining the confidence level of each candidate point as the maximum frequency position according to the forward gradient value and pixel mean of each candidate point is:

[0090] ;

[0091] ;

[0092] ;

[0093] in, For the current moment Middle The forward gradient confidence of the candidate points, is the threshold of the minimum positive gradient value, For the current moment Middle The pixel mean confidence of candidate points, is the threshold of the minimum pixel mean, For the current moment Middle The comprehensive confidence that the candidate points are the maximum frequency positions;

[0094] The calculation formula for determining the candidate point with the largest confidence as the candidate maximum frequency position of the Doppler power spectrum at the current moment according to the confidence of each candidate point is:

[0095] ;

[0096] in, is the index of the candidate point with the highest confidence. for When the maximum value is obtained value, For the current moment The candidate maximum frequency position of the Doppler power spectrum.

[0097] Another embodiment of the present invention aims to provide a system for extracting the envelope of a Doppler power spectrum, the system comprising:

[0098] An input module is used to continuously obtain the Doppler power spectrum at each moment;

[0099] A weighting module, used for weighting the Doppler power spectra at the current moment and at multiple historical moments to obtain a weighted Doppler power spectrum at the current moment, wherein the weighted Doppler power spectrum is a Doppler power spectrum weighted with historical information;

[0100] A denoising module, used for performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising processing on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment, wherein the denoised Doppler power spectrum is a Doppler power spectrum obtained after denoising the weighted Doppler power spectrum;

[0101] A smoothing module is used to smooth the denoised Doppler power spectrum at the current moment to obtain a smoothed Doppler power spectrum at the current moment, wherein the smoothed Doppler power spectrum is a Doppler power spectrum obtained by smoothing the denoised Doppler power spectrum;

[0102] An identification module, used for identifying the maximum frequency position of the smoothed Doppler power spectrum at the current moment to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment;

[0103] A correction module, used for correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment, wherein the center moment is the moment of the center position between the current moment and multiple historical moments, and the maximum frequency position is the frequency position determined after the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected;

[0104] The display module is used to connect the maximum frequency positions of the corrected Doppler power spectrum at each moment to obtain the maximum frequency envelope of the Doppler power spectrum, and to display the maximum frequency envelope of the Doppler power spectrum.

[0105] Another embodiment of the present invention aims to provide an ultrasound device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the envelope extraction method of the Doppler power spectrum as described above when running the computer program.

[0106] Another embodiment of the present invention aims to provide a medium storing a program, wherein when the program is executed by a processor, the method for extracting the envelope of the Doppler power spectrum as described above is implemented.

[0107] The Doppler power spectrum envelope extraction method provided by the embodiment of the present invention performs weighted processing on the Doppler power spectra at the current moment and multiple historical moments, thereby effectively reducing the influence of instantaneous noise on the frequency position judgment, so that the extraction of the maximum frequency position is smoother and more accurate, ensuring the real-time nature of the envelope and meeting the needs of real-time clinical diagnosis; the noise is estimated and denoised by the histogram local slope method, so that the noise interference can be accurately identified and removed, the quality of the power spectrum signal after denoising is effectively improved, the accuracy of the maximum frequency envelope extraction is ensured, and the maximum frequency envelope can still be stably extracted under the noise background; the denoised Doppler power spectrum is smoothed, so that the identification of the maximum frequency position is smoother and more coherent, and the jitter of the maximum frequency envelope is reduced. It presents smoother dynamic changes; uses data from historical moments to correct the maximum frequency position at the current moment, making the extraction of the envelope more stable and continuous, avoiding frequency position jumps caused by instantaneous abnormal data; through the confidence calculation and correction mechanism of the candidate maximum frequency position, the error fluctuations caused by noise or transient changes in the recognition process are further reduced, thereby improving the accuracy of the envelope; through multi-step processing of the Doppler power spectrum at each moment, including denoising, smoothing, frequency position identification, candidate position correction, etc., it can adapt to complex spectral forms, especially in low signal-to-noise ratio conditions, it can still accurately capture the changing trend of the spectrum and generate a smooth maximum frequency envelope; it solves the existing problem of being unable to extract the maximum frequency envelope in real time and accurately under noise interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0108] Figure 1 It is a flow chart of a method for extracting the envelope of a Doppler power spectrum provided by an embodiment of the present invention;

[0109] Figure 2 It is a structural schematic diagram of a Doppler power spectrum envelope extraction system provided by an embodiment of the present invention;

[0110] Figure 3 is a structural schematic diagram of an ultrasonic device provided by an embodiment of the present invention;

[0111] Figure 4 is a schematic diagram of the maximum frequency envelope of the Doppler power spectrum in the prior art;

[0112] Figure 5 It is a schematic diagram of a weighted Doppler power spectrum in a Doppler power spectrum envelope extraction method provided by an embodiment of the present invention;

[0113] Figure 6 is a histogram of a weighted Doppler power spectrum in a Doppler power spectrum envelope extraction method provided by an embodiment of the present invention;

[0114] Figure 7 It is a cumulative histogram of a weighted Doppler power spectrum in a Doppler power spectrum envelope extraction method provided by an embodiment of the present invention;

[0115] Figure 8 It is a schematic diagram of a denoised Doppler power spectrum in a Doppler power spectrum envelope extraction method provided by an embodiment of the present invention;

[0116] The following specific implementations will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0117] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0118] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0119] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0120] See also Figure 1 , which shows the envelope extraction method of the Doppler power spectrum in the first embodiment of the present invention. For the convenience of explanation, only the part related to the embodiment of the present invention is shown. The envelope extraction method of the Doppler power spectrum provided by the embodiment of the present invention includes:

[0121] Step S10, continuously acquiring the Doppler power spectrum at each moment;

[0122] Among them, in one embodiment of the present invention, Doppler Power Spectrum is an important image or signal expression method used in ultrasonic Doppler technology to describe the speed and direction of blood flow or tissue movement. It is usually obtained through Doppler ultrasonic signal processing, reflecting the signal intensity of different frequency components, and is a key tool for evaluating blood flow velocity, flow direction and flow characteristics in clinical practice. Specifically, the Doppler effect refers to the change in the frequency of sound waves when sound waves move relative to a moving object (such as blood). In ultrasonic Doppler examination, the ultrasonic beam is emitted into the moving blood flow, and the frequency of the reflected signal will shift due to the different speed and direction of the blood flow. This frequency shift is the Doppler frequency shift, which is proportional to the blood flow velocity.

[0123] Therefore, the ultrasound device transmits sound waves and receives reflected waves through the Doppler mode, and then calculates the frequency change (Doppler frequency shift), where the frequency change can be used to infer the speed and direction of blood flow. Among them, in the Doppler power spectrum at each moment, the X-axis (horizontal axis) of the Doppler power spectrum represents the frequency (Frequency, f), which is used to reflect the frequency change of the reflected echo signal, which is related to the blood flow velocity. The higher the frequency means the faster the blood flow velocity. The positive frequency indicates that the blood flow is toward the probe, while the negative frequency indicates that the blood flow is away from the probe. The Y-axis (vertical axis) of the Doppler power spectrum represents the power or signal intensity (Power or Amplitude), which is usually expressed by brightness or color. At this time, the larger the value of the Y-axis, the greater the intensity of the reflected signal at this frequency, which means that there is more blood movement at this frequency. Therefore, in the Doppler power spectrum at each moment, the distribution of the frequency on the X-axis and the corresponding Y-axis power value together constitute the frequency distribution diagram of the blood flow velocity at that moment.

[0124] Furthermore, the Doppler power spectra at each moment are tiled in time in turn to form a two-dimensional image data, that is, a power spectrum graph. At this time, the X-axis (horizontal axis) of the power spectrum graph represents time (t), and the data is updated once at each moment. The X-axis can determine the change trend of blood flow over time. Each time the ultrasound device is updated, the Doppler power spectrum data at a new moment (that is, the frequency distribution at the new moment) will be displayed. The Y-axis (vertical axis) represents the frequency (f) of the reflected echo, which reflects the speed of blood flow. The higher the frequency, the faster the blood flow. The positive frequency indicates that the blood flow is toward the probe (forward flow). The negative frequency indicates that the blood flow is away from the probe (reverse flow). The brightness or color in the power spectrum graph usually indicates the power of the signal, that is, the strength of the reflected signal. The higher the brightness or the darker the color, the greater the signal strength of the corresponding frequency, which means that more blood flows at this frequency / speed. Therefore, the power spectrum can be used to evaluate blood flow velocity, determine blood flow direction, and analyze hemodynamics. Specifically, the blood flow velocity can be calculated through the frequency on the Y-axis, and then the lesions such as vascular stenosis and occlusion can be diagnosed. Since the Doppler power spectrum can be displayed in positive and negative values, positive values ​​usually represent blood flow moving toward the direction of the ultrasound probe, and negative values ​​represent blood flow away from the probe. At the same time, since the signal intensity in the Doppler power spectrum reflects the energy distribution of the blood flow signal, it can be used to analyze blood flow and vascular resistance. Therefore, in clinical use, the Doppler power spectrum can be used to evaluate the function of the heart valve, the degree of vascular stenosis, blood flow abnormalities, etc., and the Doppler power spectrum can be used to detect the blood supply of the fetus and evaluate the placental function.

[0125] Further, when the spectrum Doppler mode is used clinically for blood flow detection, there are two observation modes: real-time mode and freeze mode. In clinical practice, when doctors use the real-time mode for blood flow detection, the Doppler power spectrum will be continuously updated over time, and the Doppler power spectrum will be displayed at the corresponding horizontal axis position (time) on the display. At this time, at each new time point, the generated Doppler power spectrum will only update one column, which represents the frequency distribution at that time. Therefore, in real-time mode, the ultrasound device only processes the Doppler power spectrum (i.e., each column of data) at the current moment each time. In freeze mode, the ultrasound device will save the complete Doppler power spectrum after accumulating data for a certain period of time, and display all the data at one time. At this time, the complete power spectrum diagram has been generated (including multiple time points), but the processing method can still be processed according to the real-time mode, that is, one column of data is output each time for processing. The purpose of the embodiment of the present invention is to support the identification of the maximum frequency position of the Doppler power spectrum in real-time mode, so that data can be processed column by column regardless of whether it is in real-time mode or freeze mode. In freeze mode, although a complete power spectrum is available, the ultrasound device can still output each column of the Doppler power spectrum in batches, simulating the data flow in real-time mode.

[0126] Furthermore, in one embodiment of the present invention, the data input source for acquiring the Doppler power spectrum at each moment may be an ultrasound device, or a cloud server storing the Doppler power spectrum, or other devices storing Doppler power spectrum data.

[0127] Step S20, performing weighted processing on the Doppler power spectra at the current moment and at multiple historical moments to obtain a weighted Doppler power spectrum at the current moment;

[0128] Among them, in one embodiment of the present invention, in the Doppler power spectrum, the Doppler power spectra at adjacent moments should not be very different in theory. However, in practical applications, due to factors such as noise, operating errors and patient movement, the Doppler power spectra at adjacent moments may change irregularly. Therefore, it is necessary to smooth the Doppler power spectrum to reduce the influence of noise, operating errors and external interference, so as to obtain a smoother and more stable Doppler power spectrum. In a specific embodiment of the present invention, the above-mentioned step of weighting the Doppler power spectra of the current moment and multiple historical moments to obtain the weighted Doppler power spectrum of the current moment includes:

[0129] The standard deviation coefficient is determined based on the scanning speed in Doppler mode;

[0130] Determine the weighting coefficients of the current moment and each historical moment according to the time interval between each historical moment and the current moment and the standard deviation coefficient;

[0131] The weighted Doppler power spectrum is obtained by weighting the Doppler power spectrum at the current moment and each historical moment and the weighting coefficients at the current moment and each historical moment. The weighted Doppler power spectrum is a weighted Doppler power spectrum containing historical information.

[0132] Specifically, the amount of Doppler power spectrum data collected per second by the ultrasound device in Doppler mode is different, which is called the scanning speed. When the scanning speed is fast, the time interval for the ultrasound device to collect the Doppler power spectrum is short, and when the scanning speed is slow, the time interval is long. The standard deviation coefficient is a parameter used to determine the importance or influence of historical data in weighted processing. When the scanning speed increases, the Doppler power spectrum data at adjacent moments are more similar, so the historical data can play a greater reference role. At this time, the standard deviation coefficient can be appropriately increased, indicating that the weighted effect of historical data on the data at the current moment increases, that is, the weighting coefficient changes slowly, and the weight of historical data decays slowly; conversely, when the scanning speed decreases, the difference between adjacent data may be greater, and the standard deviation coefficient can be appropriately reduced, indicating that the weighted effect of historical data on the data at the current moment decreases, that is, the influence of historical data on current data is small, and the weight of historical data decays quickly.

[0133] Furthermore, the time interval between the current moment and the historical moment is an important reference factor for the weighting coefficient. The larger the time interval, the lower the reference value of the historical data, and the smaller the weighting coefficient. Therefore, the weighting coefficient must satisfy the characteristics that the weighting coefficient of the data at the current moment is large, and the data at the past moment gradually decays with the increase of the time interval. It can be a Sigmoid function, a Gaussian function, or other types of functions that meet the characteristics. Specifically, in the embodiment of the present invention, the calculation formula for the weighting coefficients of the current moment and each historical moment is determined according to the time interval between each historical moment and the current moment, and the standard deviation coefficient:

[0134] ;

[0135] in, For the moment The weighting coefficient of For the current moment, is the standard deviation coefficient, is an exponential function with the natural base e as the base. Specifically, Indicates the time difference between the current moment and the historical moment. The standard deviation coefficient determined above is used to control the decay speed of the weighted coefficient. The larger the standard deviation coefficient, the slower the decay; the smaller the standard deviation coefficient, the faster the decay. Therefore, the weight of the data at the current moment is relatively large, and as the time interval between the historical moment and the current moment increases, the weight of the historical data gradually decreases.

[0136] Further, weighted processing is performed according to the Doppler power spectrum at the current moment and each historical moment, and the weighting coefficient to obtain a weighted Doppler power spectrum, wherein the weighted Doppler power spectrum is a Doppler power spectrum weighted by historical information. By weighted summing the data at the current moment and the data at several previous historical moments, a Doppler power spectrum weighted by historical information is obtained. The purpose of weighted processing is to smooth the power spectrum and reduce the data instability caused by noise, patient movement or other interference factors. At this time, the calculation formula for obtaining the weighted Doppler power spectrum by weighted processing according to the Doppler power spectrum at the current moment and each historical moment, and the weighting coefficients at the current moment and each historical moment is:

[0137] ;

[0138] in, For the current moment The weighted Doppler power spectrum of is the number of historical moments, that is, the number of historical data involved in weighted processing, For the moment The Doppler power spectrum of Time to current time All Doppler power spectrum data are weighted The weighted Doppler power spectrum is obtained by weighted summation. In the weighted processing, the weighting coefficient is a length of A fixed sequence / array representing the current moment As well as before The weighted coefficients of each historical moment. Each value in the array is only related to the time difference between the corresponding historical moment and the current moment, and has nothing to do with the current specific absolute time point. In other words, the weighted coefficient is a relative coefficient based on the time difference, not an absolute coefficient at a specific moment. If determined, the weighting coefficient It is also determined as a fixed sequence. Therefore, the size of the weighting coefficient depends on the time difference between the historical moment and the current moment. The larger the time difference, the smaller the weighting coefficient. Although the current moment will change as the scan progresses, the calculation rule of the weighting coefficient remains unchanged. As long as the time difference between the historical moment and the current moment is the same, the corresponding weighting coefficient will also be the same. Among them, it should be pointed out that in the actual processing process, the time length of the historical moment that needs to be cached is about 0.2s-0.5s. When the time length is greater than this time range, the correlation between the latest and oldest data is very weak.

[0139] Therefore, in an embodiment of the present invention, the weighted Doppler power spectrum of the current moment is obtained by weighting the Doppler power spectrum of the current moment with the Doppler power spectrum of multiple historical moments, so that interference such as noise and fluctuation can be effectively reduced, and the smoothness and stability of the result can be improved, thereby obtaining a more accurate Doppler power spectrum. Specifically, the Doppler power spectrum at a single moment is easily affected by instantaneous noise. By fusing historical data, the adverse effects of instantaneous fluctuations can be reduced, making the Doppler power spectrum smoother and more stable. By introducing data from multiple historical moments, short-term frequency fluctuations caused by factors such as operating errors or patient movement are reduced, thereby improving the accuracy of signal processing. At the same time, since the Doppler power spectra of adjacent moments usually do not differ greatly, the weighted effect of historical data helps to smooth the data changes at different time points, making the results more coherent. It helps to improve the robustness of the ultrasonic signal, so that parameters such as frequency changes and blood flow velocity are more accurate and reliable in the Doppler mode.

[0140] Step S30, performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising processing on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment;

[0141] Among them, in one embodiment of the present invention, in the Doppler power spectrum, noise will interfere with the accurate detection of the signal, especially the noise will cause obvious interference to the subsequent determination of the maximum frequency position, which may lead to erroneous judgment, so it is necessary to remove the possible noise first. At the same time, under different signal-to-noise ratio (SNR, Signal-to-Noise Ratio) conditions, the degree of interference of noise on the signal will be different. For example, when the signal-to-noise ratio is high, the effective signal is strong and the noise is relatively weak, and the noise interference is small at this time. When the signal-to-noise ratio is low, the signal and noise intensities are close, and the noise interference is significant at this time. At the same time, when the original acquisition signal of the ultrasonic device remains unchanged, the gain (amplification factor) of the ultrasonic device itself will also cause the change of the manifestation of noise on the Doppler power spectrum. When the gain is increased, the pixel difference between the noise and the effective signal on the Doppler power spectrum will be shortened, and it is difficult to simply distinguish the noise and the signal through the pixel value of the Doppler power spectrum, making it more difficult to distinguish the signal and the noise on the Doppler power spectrum. Among them, the common existing threshold method often estimates noise based on the average strength of the signal. This estimation method is not robust and overly dependent on signal strength. It performs poorly when encountering conditions with low signal-to-noise ratios. It also performs unstable when encountering changes in the gain of the ultrasound system, which reduces the accuracy of noise estimation.

[0142] Further, under normal circumstances, the strength of the effective signal in the Doppler power spectrum is higher than the noise strength, and between different frequency positions, the internal difference of the effective signal and the internal difference of the noise should be smaller than the difference between the effective signal and the noise. Therefore, the effective signal and the noise can be understood as two independent point sets, which are distributed in two areas of the histogram respectively. The point set of the effective signal is concentrated in a concentrated area in the histogram domain, and the point set of the noise is concentrated in another area in the histogram domain, and the internal pixel difference of each of the two areas is smaller than the difference between the two areas. Based on the above prior, the embodiment of the present invention is distinguished from the existing threshold method, and from the perspective of the histogram statistics of the Doppler power spectrum, a noise estimation method adapted to different signal-to-noise ratios and different system gains is proposed. By analyzing the characteristics of the effective signal and noise in the Doppler power spectrum, a method is designed to estimate and remove noise more robustly, so that the interference of noise on the signal can be reduced, especially when the signal-to-noise ratio is low or the system gain changes, so as to be more robust. Therefore, after removing the noise, the effective signal, especially the maximum frequency position, can be determined more accurately, ensuring the stability and accuracy of subsequent calculations, thereby avoiding misjudgment caused by noise interference.

[0143] Furthermore, the steps of performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising processing on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment include:

[0144] Calculate the histogram of the weighted Doppler power spectrum at the current moment;

[0145] Accumulating the calculated histograms to obtain an accumulated histogram;

[0146] Calculate the local slope of each point in the cumulative histogram;

[0147] According to the local slope of each point, the pixel value corresponding to the point with the largest local slope is determined as the noise intensity;

[0148] All parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity are denoised to obtain the denoised Doppler power spectrum at the current moment. The denoised Doppler power spectrum is the Doppler power spectrum obtained after the weighted Doppler power spectrum is denoised.

[0149] Specifically, first obtain the weighted Doppler power spectrum at the current moment, and its weighted Doppler power spectrum refers to Figure 5 As shown above, the X-axis of the weighted Doppler power spectrum represents frequency, and the Y-axis represents power or signal strength. Therefore, the pixel values ​​(signal strength values) in the weighted Doppler power spectrum at the current moment can be divided into several bins, and then the number of pixel values ​​in each bin is counted to form a histogram. Figure 6 As shown. The X-axis (horizontal axis) of the histogram represents the range of pixel values, and the Y-axis (vertical axis) represents the frequency of pixel values ​​within the range. Specifically, the histogram is used to show the distribution of signal strength (pixel value) at different frequency positions in the entire Doppler power spectrum, which can reflect the distribution characteristics of signals and noise. The calculation formula for calculating the histogram of the weighted Doppler power spectrum at the current moment is:

[0150] ;

[0151] in, is the pixel value The number of occurrences of is the weighted Doppler power spectrum, is a logical operator, is the pixel value. Represents the interval of pixel value (or grayscale value) (discrete value of 1-256), corresponding to each bin (column) of the histogram. is a Boolean expression that determines whether each data point in the weighted Doppler power spectrum is equal to a certain pixel value (or falls within a certain interval). It is a logical operator, which means "equal to". Its function is to judge the weighted Doppler power spectrum. Is each data point in equal to For example, if the weighted Doppler power spectrum There is a value equal to , the comparison returns True (i.e. 1); otherwise it returns False (i.e. 0). Represents all pixels in the statistically weighted Doppler power spectrum that are equal to The number of points. The pixel value is The number of times (frequency). Since the histogram counts the frequency of each pixel value appearing in the weighted Doppler power spectrum, it can reflect the distribution of the signal at different pixel levels. By calculating the histogram of the weighted Doppler power spectrum, the distribution of effective signals and noise can be clearly seen. Specifically, due to its uniform distribution, the noise has a higher frequency in the low pixel value area, which appears as a relatively gentle and dense peak on the histogram. The effective signal forms a significant peak in the high pixel area of ​​the histogram due to its higher intensity value and relatively concentrated distribution. Therefore, in the low pixel value area, the higher frequency is noise, while the high frequency part in the high pixel value area is the effective signal. By identifying these distribution patterns, the intensity value of the noise can be inferred.

[0152] Furthermore, for each pixel value interval, the pixel value frequencies of all previous intervals are accumulated to obtain a cumulative histogram, referring to Figure 7 As shown. For example, the first point of the cumulative histogram is the same as the ordinary histogram, and the second point is the accumulation of the first and second intervals, and so on, the histogram data is accumulated and summed in sequence. Each position of the cumulative histogram represents the sum of all values ​​from that position forward (that is, from 0 to that position), which is used to reflect the cumulative distribution of pixel values ​​from low to high, helping to better identify noise areas and valid signal areas. The calculation formula for the cumulative histogram obtained by accumulating the calculated histograms is:

[0153] ;

[0154] in, is the pixel value The cumulative histogram of the original data is shown in Figure 1. The cumulative histogram has the characteristics of enhanced noise resistance, distinction between valid signals and noise, and reduced sensitivity to outliers. Specifically, the cumulative histogram is more robust to noise than the original histogram. Because the original histogram directly counts each pixel value, if there are extreme values ​​or noise points in the data, the histogram will directly reflect them, while the cumulative histogram can smooth out some abnormal points through accumulation. Even if the value of a certain noise point is abnormally high, the impact during the accumulation process is relatively small, and it will not cause large fluctuations in the results like the original histogram. At the same time, the cumulative histogram can help to more easily identify the difference between valid signals and noise. In the original histogram, valid signals and noise may be difficult to distinguish in some cases (especially when the signal-to-noise ratio is low), especially in Doppler power spectrum data. If we only look at the original histogram, the boundary between noise and valid signal is not clear, but through the cumulative histogram, the cumulative value of noise points is relatively stable, while the cumulative change of valid signal is large, so only obvious inter-domain differences will cause obvious changes in the cumulative histogram. Therefore, by observing the cumulative histogram, we can see the boundary point between noise and valid signal more clearly. At the same time, the cumulative histogram is not as sensitive to outliers as the original histogram during calculation. For example, there may be a small number of outliers in the valid signal area. If we only look at the original histogram, we may mistakenly think that these outliers are noise or signal. The cumulative characteristics of the cumulative histogram can effectively alleviate the impact of a single outlier and help estimate the noise intensity more stably. Therefore, the above histogram reflects the frequency of each pixel value, while the cumulative histogram is the cumulative sum of the histogram, which can process data more smoothly and robustly. Especially when removing noise, the cumulative histogram helps to better distinguish between valid signals and noise, and can reduce the interference of noise on the final result.

[0155] Further, for each point in the cumulative histogram, the local slope of the local area of ​​each point is obtained by calculating the difference between adjacent points. The local slope of the cumulative histogram can reflect the rate of change of the pixel value frequency. Specifically, the local slope calculates the amplitude of the cumulative histogram change between the current power value or pixel value and its neighboring values. The transition between noise and valid signal will be shown as a slope change in the cumulative histogram. Specifically, the cumulative histogram in the noise area rises more evenly, and the local slope is small and stable. When entering the valid signal area, due to the sudden increase in power value, the local slope of the cumulative histogram will increase significantly, which is manifested as a sharp increase in cumulative sum. This drastic increase is due to the concentration of valid signals and the intensity is much higher than that of noise, so the frequency increases rapidly in these positions, resulting in a sharp increase in the local slope. Therefore, by calculating the local slope near each pixel value, it is possible to identify the place where the slope rises sharply, which is usually the junction of noise and valid signal, that is, the highest intensity of noise. Specifically, the calculation formula for the local slope of each point in the cumulative histogram is as follows:

[0156] ;

[0157] in, is the pixel value The local slope of is the pixel distance in the histogram. Indicates the neighborhood range used in slope calculation. The choice of value is determined by experiment, and the smaller It is more sensitive to noise, and the noise intensity estimated at different times will vary greatly. This will make the noise estimate more stable.

[0158] Further, all calculated local slopes are traversed to find the pixel value corresponding to the maximum value, which is the noise intensity, indicating the upper limit of the noise. Wherein, referring to the above, the local slope represents the rate of change of the cumulative histogram, and the dividing point between noise and effective signals is usually manifested as an obvious slope change on the cumulative histogram. Because the intensity of noise is usually low and densely distributed, while the intensity of effective signals is high and concentrated, in the cumulative histogram, when transitioning from the noise area to the effective signal area, the growth rate of the cumulative value will be significantly accelerated, forming a slope peak. Therefore, the place with the largest local slope is often the dividing point between noise and effective signals, indicating that the influence of noise on the weighted Doppler power spectrum gradually decreases, and the effective signal begins to dominate. Therefore, the point with the largest local slope can be considered as an estimated value of the noise intensity, because after this point, the growth rate of the cumulative frequency accelerates rapidly, indicating that the signal strength has increased significantly. The above-mentioned calculation formula for determining the pixel value corresponding to the point with the largest local slope according to the local slope of each point is the noise intensity is:

[0159] ;

[0160] in, is the pixel value corresponding to the noise intensity, is the local slope The k value when the maximum value is obtained. Figure 7 As shown, the asterisk point is the position of the maximum slope, and the horizontal coordinate of this point is 81, indicating that the estimated noise intensity is 81.

[0161] Furthermore, the weighted Doppler power spectrum at the current moment is traversed. For each pixel value, if its value is lower than the estimated noise intensity, it is set to zero or directly filtered out to obtain the denoised Doppler power spectrum. Figure 8 As shown in FIG. 1 , all parts of the weighted Doppler power spectrum below the noise intensity 81 are denoised to obtain the denoised Doppler power spectrum. Figure 8It can be seen that the noise is well removed and the distribution of the original effective signal is basically kept unchanged. This allows the part of the weighted Doppler power spectrum that is lower than the noise intensity to be removed and the effective signal to be retained to improve the accuracy of subsequent maximum frequency position detection. Therefore, the denoised Doppler power spectrum can effectively eliminate noise while maintaining the distribution of the original effective signal. The calculation formula for denoising all parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity is:

[0162] ;

[0163] in, is the denoised Doppler power spectrum, is the weighted Doppler power spectrum.

[0164] Among them, the method based on the cumulative histogram slope in the embodiment of the present invention can adapt to different signal-to-noise ratios and system gain changes. Regardless of the signal-to-noise ratio, the noise is concentrated in the low power value area, and the effective signal is concentrated in the higher power value area, so that the cumulative value of the noise area will be relatively uniform, while the effective signal area will show a sudden cumulative change. Even if the system gain changes, the relative distribution form of noise and effective signal will not change significantly. Therefore, the embodiment of the present invention does not rely on the specific signal strength value of the signal or noise, but is based on the statistical characteristics of the relative distribution change of the Doppler power spectrum. Therefore, when the gain changes, although the overall signal and noise intensity will increase or decrease at the same time, their relative position and distribution changes can still be accurately identified by the method in this embodiment, which is robust and avoids the problem of excessive reliance on signal strength in existing methods.

[0165] Step S40, smoothing the denoised Doppler power spectrum at the current moment to obtain a smoothed Doppler power spectrum at the current moment;

[0166] Among them, in one embodiment of the present invention, in the denoising process, the interference of noise is reduced, but the denoising process may produce false gradients. These gradient values ​​do not represent the real signal characteristics, but are artifacts in the processing process. At the same time, although the denoising process reduces the noise, strong noise sometimes still remains in the Doppler power spectrum, affecting the subsequent analysis. Among them, the false gradients and the strong noise after denoising are much smaller in scale than the effective signal. Therefore, a suitable smoothing filter can be used to perform neighborhood smoothing on the denoised Doppler power spectrum, which can not only retain the effective signal, but also remove the false gradients and strong noise to obtain the smoothed power spectrum. The smoothing method can use common mean smoothing, Gaussian smoothing, Hanning smoothing, or other filters with smoothing effects. Specifically, in an embodiment of the present invention, Hanning smoothing is used to smooth the denoised Doppler power spectrum. It smoothes the signal by applying the Hanning window, which is suitable for reducing the edge effect and noise in the spectrum. The calculation formula for smoothing the denoised Doppler power spectrum at the current moment to obtain the smoothed Doppler power spectrum at the current moment is: ;

[0167] in, For the current moment The smoothed Doppler power spectrum of is the weight array of the Hanning smoothing filter, is the first Point weight value, for The denoised Doppler power spectrum at time . Specifically, is a five-point Hanning smoothing window used to perform weighted averaging on the denoised Doppler power spectrum data. The specific weight values ​​are: , which is the center point The weight of the points is the largest, and the weights of the surrounding points gradually decrease. The smoothed Doppler power spectrum value , the current moment is taken in the formula The denoised Doppler power spectrum values ​​of the two points before and after (a total of five points) are weighted averaged according to the Hanning window weight. Through this weighted average, the false gradient caused by the denoising process can be effectively smoothed out, while weakening possible strong noise. Specifically, it should be pointed out that when calculating the smoothed Doppler power spectrum at the current moment, it is necessary to update the data over time through sliding window processing, and constantly adjust the historical and current data points so that when new Doppler power spectrum data is available, the smoothing calculation is updated. The weight distribution used for the five-point Hanning smoothing is [1 / 8, 1 / 4, 1 / 2, 1 / 4, 1 / 8]. The weight of the center point is the largest, and the weights of the surrounding points gradually decrease. This weighting method makes the value of the center point have the greatest impact on the smoothing result, while the impact of the surrounding points gradually decreases. The weighted distribution helps to reduce the impact of local noise because it depends more on neighboring points. At the same time, by weighted averaging the Doppler power spectrum data, Hanning smoothing can reduce local fluctuations in the data. This smoothing process can effectively reduce false gradients. Among them, false gradients are usually local spikes or fluctuations caused by noise or imperfect denoising processing, and smoothing can reduce the impact of these fluctuations on the overall result. At the same time, noise usually appears as high-frequency components in the spectrum, and since Hanning smoothing is based on the weighted average of the neighborhood, it has a suppressive effect on high-frequency components (i.e., noise). Through smoothing, the impact of high-frequency noise will be weakened because they receive relatively small weights when the neighborhood is weighted. At the same time, five-point Hanning smoothing takes into account multiple neighboring points when weighting, which helps to retain the characteristics of valid signals because the changes in valid signals are usually smoother rather than local spikes. Therefore, the smoothed result can better reflect the true signal trend while reducing false changes caused by noise.

[0168] Of course, in other embodiments of the present invention, mean smoothing, Gaussian smoothing, or other filters with smoothing effects may also be used, wherein mean smoothing is achieved by averaging all pixel values ​​in the neighborhood of each pixel. Gaussian smoothing uses a Gaussian function as a filter to perform weighted averaging on each pixel, with the weight being determined by the distance from the current pixel. Gaussian smoothing can better retain signal features while removing noise.

[0169] Therefore, in an embodiment of the present invention, smoothing can smooth out the false gradients generated during the denoising process, which are usually much smaller than the gradients of the actual signal. Smoothing can also reduce the influence of strong noise, because the signals of these noises fluctuate greatly in the local area and are not easy to maintain through smoothing operations. At the same time, when applying the smoothing filter, the weight of the center point is the largest, so that the characteristics of the effective signal can be better preserved. Therefore, smoothing can significantly improve the quality of the power spectrum, so that the subsequent maximum frequency position identification can more accurately identify the effective signal without being interfered by false gradients and strong noise, and can reduce the influence of false gradients and strong noise while retaining the effective signal, thereby improving the accuracy and robustness of Doppler power spectrum analysis.

[0170] Step S50, performing maximum frequency position identification on the smoothed Doppler power spectrum at the current moment to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment;

[0171] Among them, in one embodiment of the present invention, in the prior art, the method based on image processing is usually inefficient, and it needs to process the power spectrum data of a longer time period for overall analysis, which cannot meet the clinical real-time needs. In the embodiment of the present invention, by applying the image processing idea to a single power spectrum line (that is, the smoothed Doppler power spectrum obtained after weighted denoising and smoothing of the Doppler power spectrum at the current moment), rather than analyzing the data of all time periods as a whole, the calculation efficiency can be significantly improved. By weighting the historical information, the Doppler power spectrum at the current moment is processed so that the data has a stronger correlation in time and reduces the impact of noise on the analysis. And using gradient information, the effective signal and noise are quickly identified. The effective signal presents a continuous and concentrated distribution feature in the Doppler power spectrum, while the noise is random and discrete. At this time, the target area is judged whether it is a valid signal by the gradient situation in the local area, and the dividing line (i.e., the maximum frequency position) between the effective signal and the noise is distinguished.

[0172] Furthermore, the frequency of the Doppler power spectrum exists in both positive and negative directions, with the central baseline as the boundary, the positive frequency above the baseline, and the reverse frequency below the baseline. In the Doppler power spectrum analysis required by clinical needs, both the positive spectrum (above the baseline) and the reverse spectrum (below the baseline) may need to identify their maximum frequency envelope. Since the frequencies in two directions can be distinguished by the baseline, when the maximum frequency envelope is identified for the positive spectrum above the baseline, the maximum frequency envelope is directly identified in the spectrum above the baseline. When the maximum frequency envelope is identified for the negative spectrum below the baseline, the reverse spectrum below the baseline can be flipped up and down to obtain a positive spectrum similar to the one above. After flipping, the same envelope identification method is applied to obtain the maximum frequency envelope of the reverse spectrum. Finally, the result is flipped back to the original negative direction coordinate to obtain the correct reverse maximum frequency envelope. Therefore, for the spectra in the positive and negative directions, the processing methods of the two are consistent. In the subsequent description, the embodiment of the present invention focuses on the maximum frequency envelope identification method for the positive spectrum / above the baseline.

[0173] Specifically, the step of identifying the maximum frequency position of the smoothed Doppler power spectrum at the current moment to obtain the candidate maximum frequency position of the Doppler power spectrum at the current moment includes:

[0174] Performing gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment to obtain the gradient value of each frequency;

[0175] According to the gradient values ​​of each frequency, a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​are determined as candidate points;

[0176] Calculate the pixel mean in the local area below each candidate point in turn;

[0177] Determine the confidence level of each candidate point as the maximum frequency position based on the forward gradient value and pixel mean of each candidate point;

[0178] According to the confidence of each candidate point, the candidate point with the largest confidence is determined as the candidate maximum frequency position of the Doppler power spectrum at the current moment.

[0179] Among them, firstly, the input smoothed Doppler power spectrum is subjected to gradient detection, wherein the gradient detection is used to identify the change of pixel intensity in the power spectrum data. The larger the gradient, the more drastic the signal change at that position. The effective signal area usually appears as a higher intensity in the smoothed Doppler power spectrum, while the background noise appears as a lower intensity. At this time, by detecting the gradient, it is possible to find the place where the signal intensity changes significantly, that is, the edge in the spectrum, which helps to locate the feature points in the spectrum more accurately (such as the maximum frequency position). The gradient detection specifically calculates the gradient value by differentiating the upper and lower pixels of the smoothed Doppler power spectrum. The above-mentioned calculation formula for the gradient value of each frequency obtained by performing gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment is: ;

[0180] in, For the current moment In frequency The gradient value at is the gradient detection operator, is the gradient detection operator The value of a point, For the current moment The smoothed Doppler power spectrum at frequency Specifically, the power value difference of the surrounding points (such as the adjacent points before and after) is calculated for each frequency point, and a gradient value is obtained by using the gradient detection operator to help calculate the difference, which represents the degree of change of the smoothed Doppler power spectrum.

[0181] Furthermore, for the positive frequency, the signal strength should be higher at the bottom, while the background noise should be lower at the top. Therefore, only the positive gradient value is retained, that is, the negative values ​​in the gradient value are filtered out, and only the positive changes are considered. Then for each moment, the calculated gradient values ​​are sorted, and the frequency position coordinates with the largest gradient values ​​before the preset number are selected as candidate points. The calculation formula for determining the preset number of frequency position coordinates with the positive gradient values ​​arranged in order from large to small as candidate points according to the gradient values ​​of each frequency is: ; ; ;

[0182] in, For the gradient value Sort by maximum value. is the frequency position coordinate array after the forward gradient values ​​are sorted from large to small, is the gradient value array after the forward gradient values ​​are sorted from large to small, For the current moment The one with the largest positive gradient The gradient intensity sequence of points; For the current moment The one with the largest positive gradient The frequency position coordinate sequence of points, is the number of valid candidate points, which is also the preset number. The positions of the candidate points are determined by The function is sorted and the points with the largest gradient values ​​are retained. represents the gradient values ​​of these candidate points, Represents the positions of these candidate points (i.e., the coordinates on the frequency axis of the Doppler power spectrum).

[0183] Furthermore, for the forward maximum frequency position, since the power of the effective signal is large and the intensity of the background noise area is small, the effective signal area of ​​its forward spectrum is usually manifested as a higher intensity value in the Doppler power spectrum, while the background noise is usually distributed in a higher position (lower intensity) of the Doppler power spectrum. At this time, the distribution of the intensity of its neighborhood pixels should be that the effective signal pixel value is large and at the bottom; while the background area pixel value is small and at the top. Therefore, for each candidate point, the pixel mean of the local area below it is calculated. The local mean indicates whether the signal intensity of the area below the candidate point is high. The effective signal is usually at the edge of the strong signal area, so the larger the pixel mean below the candidate point, the more likely it is to be the maximum frequency position. The above calculation formula for calculating the pixel mean in the local area below each candidate point in turn is:

[0184] ;

[0185] in, For the current moment No. The average pixel value in the local area below the candidate point, is the local range parameter, For the current moment The denoised Doppler power spectrum is The pixel values ​​of candidate points; The parameter determines the size of the local region. The parameter value is usually negatively correlated with the pulse repetition frequency (PRF) of the ultrasonic device. When the PRF is higher, the local area is smaller; conversely, when the PRF is lower, the local area is larger.

[0186] Furthermore, the confidence of each candidate point is mainly based on two conditions: gradient value and pixel mean. The larger the gradient value, the greater the possibility that the candidate point is the maximum frequency position. The larger the local mean, the greater the possibility that the candidate point is the maximum frequency position. Therefore, the confidence of each candidate point as the maximum frequency position is determined based on the forward gradient value and pixel mean of each candidate point. The calculation formula is: ; ;

[0187] in, For the current moment Middle The forward gradient confidence of the candidate points, is the threshold of the minimum positive gradient value, For the current moment Middle The pixel mean confidence of candidate points, is the threshold of the minimum pixel mean, For the current moment Middle The candidate point is the comprehensive confidence of the maximum frequency position; when the gradient value of the candidate point is less than the threshold of the minimum positive gradient value , then it is considered that the candidate point cannot be the maximum frequency position, that is, the positive gradient confidence Set to 0. Similarly, when the pixel mean of the candidate point is less than the threshold of the minimum pixel mean , then the candidate point is also excluded, that is, the pixel mean confidence Set to 0.

[0188] Further, from the remaining candidate points, the candidate point with the highest confidence is selected as the candidate maximum frequency position of the Doppler power spectrum at the current moment, that is, the position of the maximum blood flow velocity. The calculation formula for determining the candidate point with the highest confidence as the candidate maximum frequency position of the Doppler power spectrum at the current moment according to the confidence of each candidate point is: ;

[0189] in, is the index of the candidate point with the highest confidence. for When the maximum value is obtained value, For the current moment The candidate maximum frequency position of the Doppler power spectrum.

[0190] Therefore, in an embodiment of the present invention, the most significant change position in the spectrum is determined by calculating the local change of the smoothed Doppler power spectrum through gradient detection, which helps to identify the edge or maximum frequency position of the signal; only the positive gradient is focused on through positive gradient processing to ensure that only the rising part of the effective signal is detected, and the negative gradient caused by noise is excluded; the pixel mean calculation is used to confirm whether the candidate point is in the effective signal area, where the mean of the noise area is low, so the mean of the effective signal point should be high; the possibility of each candidate point is evaluated by combining the gradient strength and the local mean, thereby improving the recognition accuracy. Accurate recognition is achieved by selecting the most likely candidate point as the maximum frequency position of the Doppler power spectrum at the current moment. Compared with the existing methods, it is not necessary to analyze the complete power spectrum data, but to perform fast calculations based on the local information of a single power spectrum line, which improves the calculation efficiency, meets the needs of real-time analysis, and maintains a high degree of accuracy.

[0191] Step S60, correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment;

[0192] Among them, in one embodiment of the present invention, when using ultrasound equipment to collect Doppler power spectrum data in real time, the acquired signal may be unstable due to factors such as the movement of the ultrasound probe, the change of the pressure of the ultrasound probe, the movement of the patient, or the interference of external device signals. The Doppler power spectrum is presented as an abnormal situation of temporary brightening or darkening. This abnormal situation cannot be solved by the above steps. The normal maximum frequency envelope curve should be relatively stable. Even if a mutation occurs, it should be a single mutation. For example, when changing from low-speed blood flow to high-speed blood flow, a single mutation may occur, but multiple jumps cannot occur in a short time. Therefore, based on the above a priori, the acquired maximum frequency envelope can be detected for abnormal points, and then the possible abnormal points can be corrected.

[0193] The step of correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment comprises:

[0194] Determine a central moment located at a central position between the current moment and the multiple historical moments according to the current moment and the multiple historical moments;

[0195] Calculate the coordinate differences between the candidate maximum frequency position of the Doppler power spectrum at the center moment and the candidate maximum frequency position of the Doppler power spectrum at the previous moment and the next moment respectively;

[0196] Determine a coordinate difference value according to the two coordinate differences, and judge whether the coordinate difference value is less than a difference threshold;

[0197] When it is determined that the coordinate difference is less than the difference threshold, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the corresponding candidate maximum frequency position.

[0198] Specifically, first, select a sequence from the candidate maximum frequency positions of the Doppler power spectrum at the current moment and its multiple historical moments, and ensure that the sequence length is greater than or equal to 5. Assuming that the sequence length is 5, the center moment is the third moment of the sequence. Then, for the center moment, calculate the coordinate difference between it and the candidate maximum frequency position of the Doppler power spectrum at the previous moment and the next moment. Specifically, the calculation formula for the coordinate difference at the previous moment is , the calculation formula of the coordinate difference at the next moment is ,in is the coordinate difference at the previous moment, is the coordinate difference at the next moment, is the candidate maximum frequency position of the Doppler power spectrum at the center moment, is the candidate maximum frequency position of the Doppler power spectrum at the moment before the center moment, The candidate maximum frequency position of the Doppler power spectrum at the moment after the center moment, at this time, the difference between the two coordinate differences before and after is calculated to obtain the coordinate difference difference, and the calculation formula for determining the coordinate difference difference based on the two coordinate differences is: ,in is the coordinate difference, Indicates taking the absolute value, at this time the coordinate difference It reflects the change of the central moment relative to the previous and next moments. Then the coordinate difference is Difference threshold with preset value Compare, where the difference threshold is an empirical parameter. If the coordinate difference Less than difference threshold , indicating that the maximum frequency position changes slightly between the two moments before and after. It can be considered that there is no sudden change in the maximum frequency envelope on both sides of the center moment, and the change trend is consistent, so no correction is required. In this case, the candidate maximum frequency position at the center moment is considered reasonable, and the candidate maximum frequency position determined above is directly determined as the maximum frequency position of the Doppler power spectrum at the center moment. The coordinate difference Greater than difference threshold , indicating that the maximum frequency position changes greatly between the two moments before and after. It is believed that the maximum frequency envelope has a sudden change on both sides of the center moment, which requires further processing.

[0199] Furthermore, the step of determining whether the coordinate difference is less than the difference threshold value further includes:

[0200] When it is determined that the coordinate difference is not less than the difference threshold, the previous moment and the next moment of the center moment are respectively used as the new center moment;

[0201] Calculate the coordinate differences between the candidate maximum frequency positions of the Doppler power spectra at the two new center moments and the corresponding candidate maximum frequency positions of the Doppler power spectra at the previous moment and the next moment respectively;

[0202] Determine two new coordinate difference values ​​according to the two coordinate differences respectively determined at the two new center moments, and judge whether the two new coordinate difference values ​​are both less than the difference threshold;

[0203] When it is determined that the differences of the two new coordinate differences are both less than the difference threshold, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the corresponding candidate maximum frequency position.

[0204] Specifically, when it is determined that the coordinate difference is not less than the difference threshold, further search is conducted toward the outside of the center moment, and the previous moment and the next moment are respectively regarded as the new center moment. The above calculation of the coordinate difference is repeated and compared with the difference threshold to determine whether there is any mutation. If the coordinate difference is less than the difference threshold, it is considered that this is a single mutation phenomenon, which may be normal in actual situations, such as a sudden change in blood flow velocity. At this time, no correction is required, and the maximum frequency position of the Doppler power spectrum at the center moment is directly determined as the corresponding candidate maximum frequency position.

[0205] Furthermore, the step of determining whether the difference between the two new coordinates is less than the difference threshold also includes:

[0206] When it is determined that any new coordinate difference is not less than the difference threshold, the mean value between the candidate maximum frequency positions of the Doppler power spectrum at the previous moment and the next moment of the center moment is calculated to obtain the mean maximum frequency position of the Doppler power spectrum at the center moment;

[0207] Traverse each candidate point in the Doppler power spectrum at the center time, and determine whether there is a candidate point whose coordinate difference between each candidate point and the mean maximum frequency position is less than the difference threshold;

[0208] If yes, then determine that the maximum frequency position of the Doppler power spectrum at the center time is the target candidate point that is closest to the difference threshold among the candidate points that are smaller than the difference threshold;

[0209] If not, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the mean maximum frequency position.

[0210] Specifically, when it is determined that any new coordinate difference is not less than the difference threshold, it is determined that a new mutation is detected again, and it is considered that multiple mutations have occurred. Multiple mutations usually indicate abnormal data and do not conform to normal physiological changes. This situation requires special processing to ensure the stability and accuracy of the data. Therefore, at the center moment, when the maximum frequency envelope has multiple abnormal mutations, it needs to be corrected. Among them, since the candidate points themselves are points selected under certain conditions, they are more likely to belong to the true maximum frequency position. Therefore, when correcting, first consider whether there are points that meet the conditions among the candidate points. First, calculate the mean of the candidate maximum frequency positions before and after the moments on both sides of the center moment to obtain the mean maximum frequency position, and then traverse the candidate point coordinates at the center moment to determine whether there are any candidate points whose distance from the mean maximum frequency position is less than the difference threshold. If there is a candidate point, the maximum frequency position of the Doppler power spectrum at the center moment is determined to be the target candidate closest to the difference threshold among the candidate points less than the difference threshold. If there is no candidate point that meets the threshold condition, the maximum frequency position of the Doppler power spectrum at the center moment is directly determined to be the mean maximum frequency position, ensuring the stability of the maximum frequency envelope change. Therefore, in the case of multiple mutations, first consider whether there are points that meet the conditions among the candidate points, which may represent the more realistic maximum frequency position. Selecting the candidate point closest to the mean maximum frequency position can better exclude outliers and ensure that the corrected data is as close to the true value as possible. If there is no point that meets the conditions among the candidate points, the mean maximum frequency position is used as the corrected maximum frequency position. The mean coordinate is the mean of the coordinates of the two moments before and after, which has the effect of smoothing, can reduce the unstable effect caused by mutations, and ensure that the data is more stable and reliable. Therefore, the mean maximum frequency position can be used to provide stable output when there are no other candidate points that meet the conditions. This processing method ensures the accuracy of the maximum frequency position in the Doppler power spectrum and avoids data instability caused by mutations.

[0211] Therefore, in the embodiment of the present invention, the stability and accuracy of the Doppler power spectrum are ensured by detecting and correcting the abnormal points of the maximum frequency envelope; by calculating the coordinate difference, comparing the difference with the threshold, checking the mutation, and finding a suitable point among the candidate points to replace the invalid coordinates, the abnormal phenomenon in the Doppler power spectrum can be effectively identified and corrected, making the real-time data processing of the ultrasonic equipment more reliable.

[0212] Step S70, connecting the maximum frequency positions of the corrected Doppler power spectra at each moment to obtain a maximum frequency envelope of the Doppler power spectrum, and displaying the maximum frequency envelope of the Doppler power spectrum;

[0213] Among them, in one embodiment of the present invention, first, the maximum frequency position of the Doppler power spectrum at each moment is corrected. Through the previous algorithm steps, the corrected maximum frequency position at each moment can be obtained. Then the corrected maximum frequency position of the Doppler power spectrum at each moment is connected into a maximum frequency envelope, and the corrected maximum frequency envelope is superimposed and displayed with the Doppler power spectrum image so that the relationship between the envelope and the spectrum can be intuitively observed. When new data arrives, the envelope is updated in real time so that it can be continuously moved and displayed on the time axis. Therefore, through the above method, the maximum frequency envelope finally extracted can be clearly displayed on the Doppler power spectrum diagram, which is used to evaluate blood flow velocity, analyze hemodynamics, etc.

[0214] In summary, the Doppler power spectrum envelope extraction method in the above-mentioned embodiment of the present invention effectively reduces the influence of instantaneous noise on the frequency position judgment by weighting the Doppler power spectrum at the current moment and multiple historical moments, so that the extraction of the maximum frequency position is smoother and more accurate, ensuring the real-time nature of the envelope and meeting the needs of real-time clinical diagnosis; by estimating and denoising the noise using the histogram local slope method, it is possible to accurately identify and remove noise interference, effectively improve the quality of the power spectrum signal after denoising, ensure the accuracy of the maximum frequency envelope extraction, and enable the maximum frequency envelope to be stably extracted under a noisy background; by smoothing the denoised Doppler power spectrum, it is ensured that the identification of the maximum frequency position is smoother and more coherent, and the jitter of the maximum frequency envelope is reduced. The maximum frequency position of the current moment is corrected by using the data of the historical moment, so that the extraction of the envelope is more stable and continuous, and the frequency position jump caused by instantaneous abnormal data is avoided; the error fluctuation caused by noise or transient changes in the recognition process is further reduced through the confidence calculation and correction mechanism of the candidate maximum frequency position, thereby improving the accuracy of the envelope; through multi-step processing of the Doppler power spectrum at each moment, including denoising, smoothing, frequency position identification, candidate position correction, etc., it can adapt to complex spectrum forms, especially in the case of low signal-to-noise ratio, it can still accurately capture the changing trend of the spectrum and generate a smooth maximum frequency envelope; it solves the existing problem that the maximum frequency envelope cannot be extracted in real time and accurately under noise interference.

[0215] See also Figure 2 , is a schematic diagram of the structure of the Doppler power spectrum envelope extraction system provided by the second embodiment of the present invention. For the sake of convenience, only the part related to the embodiment of the present invention is shown. The system includes:

[0216] An input module 11 is used to continuously obtain the Doppler power spectrum at each moment;

[0217] A weighting module 12 is used to perform weighted processing on the Doppler power spectra at the current moment and the Doppler power spectra at multiple historical moments to obtain a weighted Doppler power spectrum at the current moment, wherein the weighted Doppler power spectrum is a Doppler power spectrum weighted with historical information;

[0218] A denoising module 13 is used to perform noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and perform denoising on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment, wherein the denoised Doppler power spectrum is a Doppler power spectrum obtained after denoising the weighted Doppler power spectrum;

[0219] A smoothing module 14 is used to smooth the denoised Doppler power spectrum at the current moment to obtain a smoothed Doppler power spectrum at the current moment, wherein the smoothed Doppler power spectrum is a Doppler power spectrum obtained by smoothing the denoised Doppler power spectrum;

[0220] An identification module 15 is used to identify the maximum frequency position of the smoothed Doppler power spectrum at the current moment to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment;

[0221] A correction module 16 is used to correct the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment, wherein the center moment is the moment of the center position between the current moment and multiple historical moments, and the maximum frequency position is the frequency position determined after the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected;

[0222] The display module 17 is used to connect the maximum frequency positions of the corrected Doppler power spectra at each moment to obtain the maximum frequency envelope of the Doppler power spectrum, and to display the maximum frequency envelope of the Doppler power spectrum.

[0223] Furthermore, in one embodiment of the present invention, the weighting module 12 includes:

[0224] A standard deviation coefficient determination unit, used to determine the standard deviation coefficient according to the scanning speed in the Doppler mode;

[0225] A weighting coefficient determination unit, used to determine the weighting coefficients of the current moment and each historical moment according to the time interval between each historical moment and the current moment and the standard deviation coefficient;

[0226] The weighted Doppler power spectrum determination unit is used to perform weighted processing according to the Doppler power spectra at the current moment and each historical moment and the weighted coefficients at the current moment and each historical moment to obtain the weighted Doppler power spectrum.

[0227] Furthermore, in one embodiment of the present invention, the denoising module 13 includes:

[0228] A histogram calculation unit, used for calculating a histogram of a weighted Doppler power spectrum at a current moment;

[0229] A cumulative histogram processing unit, used for performing cumulative processing on the calculated histogram to obtain a cumulative histogram;

[0230] A local slope calculation unit, used for calculating the local slope of each point in the cumulative histogram;

[0231] A noise intensity determination unit, used for determining, according to the local slope of each point, that the pixel value corresponding to the point with the largest local slope is the noise intensity;

[0232] The denoised Doppler power spectrum determination unit is used to perform denoising on all parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity to obtain the denoised Doppler power spectrum at the current moment.

[0233] Furthermore, in one embodiment of the present invention, the identification module 15 includes:

[0234] A gradient value determination unit, used to perform gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment to obtain a gradient value of each frequency;

[0235] A candidate point determination unit, used to determine a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​as candidate points according to the gradient values ​​of each frequency;

[0236] A pixel mean calculation unit, used to sequentially calculate the pixel mean in the local area below each candidate point;

[0237] A confidence determination unit, used to determine the confidence of each candidate point as the maximum frequency position according to the forward gradient value and pixel mean of each candidate point;

[0238] The candidate maximum frequency position determination unit is used to determine the candidate point with the highest confidence as the candidate maximum frequency position of the Doppler power spectrum at the current moment according to the confidence of each candidate point.

[0239] Furthermore, in one embodiment of the present invention, the correction module 16 includes:

[0240] A central moment determination unit, used to determine a central moment located at a central position between the current moment and the multiple historical moments according to the current moment and the multiple historical moments;

[0241] A first coordinate difference calculation unit is used to calculate the coordinate differences between the candidate maximum frequency position of the Doppler power spectrum at the center moment and the candidate maximum frequency position of the Doppler power spectrum at the previous moment and the next moment respectively;

[0242] A first threshold judgment unit, used to determine a coordinate difference value according to the two coordinate differences, and to judge whether the coordinate difference value is less than a difference threshold value;

[0243] The first maximum frequency position determining unit is used to determine the maximum frequency position of the Doppler power spectrum at the center moment as the corresponding candidate maximum frequency position when the first threshold judging unit judges that the coordinate difference is less than the difference threshold.

[0244] Furthermore, in one embodiment of the present invention, the correction module 16 further includes:

[0245] A new center moment determination unit, configured to use the previous moment and the next moment of the center moment as the new center moment respectively when the first threshold determination unit determines that the coordinate difference is not less than the difference threshold;

[0246] A second coordinate difference calculation unit is used to calculate the coordinate differences between the candidate maximum frequency positions of the Doppler power spectra at two new center moments and the candidate maximum frequency positions of the Doppler power spectra at the previous moment and the next moment respectively;

[0247] A second threshold judgment unit is used to determine two new coordinate difference values ​​according to two coordinate differences respectively determined at the two new center moments, and to judge whether the two new coordinate difference values ​​are both less than a difference threshold value;

[0248] The second maximum frequency position determining unit is used to determine the maximum frequency position of the Doppler power spectrum at the center moment as the corresponding candidate maximum frequency position when the second threshold judging unit judges that the differences of the two new coordinate differences are both less than the difference threshold.

[0249] Furthermore, in one embodiment of the present invention, the correction module 16 further includes:

[0250] A mean value calculation unit, used for calculating the mean value between the candidate maximum frequency positions of the Doppler power spectrum at the previous moment and the next moment of the center moment to obtain the mean maximum frequency position of the Doppler power spectrum at the center moment when the second threshold value judgment unit judges that any new coordinate difference is not less than the difference threshold value;

[0251] A candidate point determination unit is used to traverse each candidate point in the Doppler power spectrum at the center time, and determine whether there is a candidate point whose coordinate difference between each candidate point and the mean maximum frequency position is less than a difference threshold;

[0252] A third maximum frequency position determination unit is used to determine that the maximum frequency position of the Doppler power spectrum at the center moment is the target candidate point closest to the difference threshold among the candidate points less than the difference threshold when the candidate point determination unit determines that there is a candidate point less than the difference threshold in the coordinate differences between the candidate points and the mean maximum frequency position;

[0253] The fourth maximum frequency position determination unit is used to determine that the maximum frequency position of the Doppler power spectrum at the center moment is the mean maximum frequency position when the candidate point determination unit determines that there is no candidate point whose coordinate difference between each candidate point and the mean maximum frequency position is less than the difference threshold.

[0254] Furthermore, in one embodiment of the present invention, the calculation formula of the weighted coefficient determination unit is:

[0255] ;

[0256] in, For the moment The weighting coefficient of For the current moment, is the standard deviation coefficient, is an exponential function with the natural base e as base;

[0257] The calculation formula of the weighted Doppler power spectrum determination unit is:

[0258] ;

[0259] in, For the current moment The weighted Doppler power spectrum of is the number of historical moments, For the moment The Doppler power spectrum.

[0260] Furthermore, in one embodiment of the present invention, the calculation formula of the histogram calculation unit is:

[0261] ;

[0262] in, is the pixel value The number of occurrences of is the weighted Doppler power spectrum, is a logical operator, is the pixel value;

[0263] The calculation formula of the cumulative histogram calculation unit is:

[0264] ;

[0265] in, is the pixel value The cumulative histogram of ;

[0266] The calculation formula of the local slope calculation unit is: ;

[0267] in, is the pixel value The local slope of is the pixel distance in the histogram;

[0268] The calculation formula of the noise intensity determination unit is: ;

[0269] in, is the pixel value corresponding to the noise intensity, is the local slope The k value when the maximum value is obtained;

[0270] The calculation formula of the denoised Doppler power spectrum determination unit is: ;

[0271] in, is the denoised Doppler power spectrum, is the weighted Doppler power spectrum.

[0272] Furthermore, in one embodiment of the present invention, the calculation formula of the smoothing module 14 is: ;

[0273] in, For the current moment The smoothed Doppler power spectrum of is the weight array of the Hanning smoothing filter, is the first Point weight value, for The denoised Doppler power spectrum at time instant.

[0274] Furthermore, in one embodiment of the present invention, the calculation formula of the gradient value determination unit is: ;

[0275] ;

[0276] in, For the current moment In frequency The gradient value at is the gradient detection operator, is the gradient detection operator The value of a point, For the current moment The smoothed Doppler power spectrum at frequency The value at

[0277] The calculation formula of the candidate point determination unit is: ; ; ;

[0278] in, For the gradient value Sort by maximum value. is the frequency position coordinate array after the forward gradient values ​​are sorted from large to small, is the gradient value array after the forward gradient values ​​are sorted from large to small, For the current moment The one with the largest positive gradient The gradient intensity sequence of points; For the current moment The one with the largest positive gradient The frequency position coordinate sequence of points, is the number of valid candidate points, which is also the preset number;

[0279] The calculation formula of the pixel mean calculation unit is:

[0280] ;

[0281] in, For the current moment No. The average pixel value in the local area below the candidate point, is the local range parameter, For the current moment The denoised Doppler power spectrum is The pixel value of each candidate point;

[0282] The calculation formula of the confidence determination unit is: ;

[0283] ;

[0284] in, For the current moment Middle The forward gradient confidence of the candidate points, is the threshold of the minimum positive gradient value, For the current moment Middle The pixel mean confidence of candidate points, is the threshold of the minimum pixel mean, For the current moment Middle The comprehensive confidence that the candidate points are the maximum frequency positions;

[0285] The calculation formula of the candidate maximum frequency position determination unit is: ;

[0286] in, is the index of the candidate point with the highest confidence. for When the maximum value is obtained value, For the current moment The candidate maximum frequency position of the Doppler power spectrum.

[0287] An embodiment of the present invention provides a Doppler power spectrum envelope extraction system, whose implementation principle and technical effects are the same as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the system embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0288] Another aspect of the present invention also provides an ultrasonic device, see Figure 3 , shown is an ultrasound device in the third embodiment of the present invention, including a memory 20, a processor 10, and a computer program 30 stored in the memory 20 and executable on the processor 10, wherein the processor 10 implements the envelope extraction method of the Doppler power spectrum as described above when executing the computer program 30.

[0289] In some embodiments, the processor 10 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 20, such as executing access restriction programs.

[0290] The memory 20 includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. The memory 20 may be an internal storage unit of the ultrasound device in some embodiments, such as a hard disk of the ultrasound device. The memory 20 may also be an external storage device of the ultrasound device in other embodiments, such as a plug-in hard disk equipped on the ultrasound device, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. Further, the memory 20 may also include both an internal storage unit and an external storage device of the ultrasound device. The memory 20 may be used not only to store application software and various types of data installed in the ultrasound device, but also to temporarily store data that has been output or is to be output.

[0291] It should be pointed out that Figure 3 The structure shown does not constitute a limitation on the ultrasound device. In other embodiments, the ultrasound device may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0292] The embodiment of the present invention further provides a medium on which a program is stored. When the program is executed by a processor, the method for extracting the envelope of the Doppler power spectrum as described in the above method embodiment is implemented.

[0293] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, a "medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0294] More specific examples of media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0295] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0296] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0297] The above-described embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A method for extracting the envelope of a Doppler power spectrum, characterized in that: The method: Continuously obtain the Doppler power spectrum at each moment; Performing weighted processing on the Doppler power spectra at the current moment and at multiple historical moments to obtain a weighted Doppler power spectrum at the current moment, wherein the weighted Doppler power spectrum is a Doppler power spectrum weighted by historical information; Performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment, wherein the denoised Doppler power spectrum is a Doppler power spectrum obtained after denoising the weighted Doppler power spectrum; The denoised Doppler power spectrum at the current moment is smoothed to obtain a smoothed Doppler power spectrum at the current moment, wherein the smoothed Doppler power spectrum is a Doppler power spectrum obtained by smoothing the denoised Doppler power spectrum; The maximum frequency position of the smoothed Doppler power spectrum at the current moment is identified to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment; According to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments, the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected to obtain the maximum frequency position of the Doppler power spectrum at the center moment, wherein the center moment is the moment at the center position between the current moment and multiple historical moments, and the maximum frequency position is the frequency position determined after the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected; The maximum frequency positions of the corrected Doppler power spectra at each moment are connected to obtain the maximum frequency envelope of the Doppler power spectrum, and the maximum frequency envelope of the Doppler power spectrum is displayed; The step of performing maximum frequency position identification on the smoothed Doppler power spectrum at the current moment to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment comprises: Performing gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment to obtain the gradient value of each frequency; According to the gradient values ​​of each frequency, a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​are determined as candidate points; Calculate the pixel mean in the local area below each candidate point in turn; Determine the confidence level of each candidate point as the maximum frequency position based on the forward gradient value and pixel mean of each candidate point; According to the confidence of each candidate point, determine the candidate point with the largest confidence as the candidate maximum frequency position of the Doppler power spectrum at the current moment; The step of correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment comprises: Determine a central moment located at a central position between the current moment and the multiple historical moments according to the current moment and the multiple historical moments; Calculate the coordinate differences between the candidate maximum frequency position of the Doppler power spectrum at the center moment and the candidate maximum frequency position of the Doppler power spectrum at the previous moment and the next moment respectively; Determine a coordinate difference value according to the two coordinate differences, and judge whether the coordinate difference value is less than a difference threshold; When it is determined that the coordinate difference is less than the difference threshold, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the corresponding candidate maximum frequency position; The step of determining whether the coordinate difference is less than the difference threshold value also includes: When it is determined that the coordinate difference is not less than the difference threshold, the previous moment and the next moment of the center moment are respectively used as the new center moment; Calculate the coordinate differences between the candidate maximum frequency positions of the Doppler power spectra at the two new center moments and the candidate maximum frequency positions of the Doppler power spectra at the previous moment and the next moment respectively; Determine two new coordinate difference values ​​according to the two coordinate differences respectively determined at the two new center moments, and judge whether the two new coordinate difference values ​​are both less than the difference threshold; When it is determined that the differences of the two new coordinate differences are both less than the difference threshold, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the corresponding candidate maximum frequency position.

2. The method for extracting the envelope of the Doppler power spectrum according to claim 1, characterized in that: The step of performing weighted processing on the Doppler power spectra at the current moment and the Doppler power spectra at multiple historical moments to obtain the weighted Doppler power spectrum at the current moment comprises: The standard deviation coefficient is determined based on the scanning speed in Doppler mode; Determine the weighting coefficients of the current moment and each historical moment according to the time interval between each historical moment and the current moment and the standard deviation coefficient; The weighted Doppler power spectrum is obtained by performing weighted processing according to the Doppler power spectrum at the current moment and each historical moment and the weighting coefficients at the current moment and each historical moment.

3. The method for extracting the envelope of the Doppler power spectrum according to claim 1, characterized in that: The steps of performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising processing on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment include: Calculate the histogram of the weighted Doppler power spectrum at the current moment; Accumulating the calculated histograms to obtain an accumulated histogram; Calculate the local slope of each point in the cumulative histogram; According to the local slope of each point, the pixel value corresponding to the point with the largest local slope is determined as the noise intensity; All parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity are denoised to obtain the denoised Doppler power spectrum at the current moment.

4. The method for extracting the envelope of the Doppler power spectrum according to claim 1, characterized in that: The step of judging whether the difference values ​​of the two new coordinate differences are both less than the difference threshold value also includes: When it is determined that any new coordinate difference is not less than the difference threshold, the mean value between the candidate maximum frequency positions of the Doppler power spectrum at the previous moment and the next moment of the center moment is calculated to obtain the mean maximum frequency position of the Doppler power spectrum at the center moment; Traverse each candidate point in the Doppler power spectrum at the center time, and determine whether there is a candidate point whose coordinate difference between each candidate point and the mean maximum frequency position is less than the difference threshold; If yes, then determine that the maximum frequency position of the Doppler power spectrum at the center time is the target candidate point that is closest to the difference threshold among the candidate points that are smaller than the difference threshold; If not, the maximum frequency position of the Doppler power spectrum at the center moment is determined as the mean maximum frequency position.

5. The method for extracting the envelope of the Doppler power spectrum according to claim 2, characterized in that: The calculation formula for determining the weighted coefficients of the current moment and each historical moment based on the time interval between each historical moment and the current moment and the standard deviation coefficient is: ; in, For the moment The weighting coefficient of For the current moment, is the standard deviation coefficient, is an exponential function with the natural base e as base; The calculation formula for obtaining the weighted Doppler power spectrum by weighted processing based on the Doppler power spectrum at the current moment and each historical moment and the weighted coefficients at the current moment and each historical moment is: ; in, For the current moment The weighted Doppler power spectrum of is the number of historical moments, For the moment The Doppler power spectrum.

6. The method for extracting the envelope of the Doppler power spectrum according to claim 3, characterized in that: The calculation formula for calculating the histogram of the weighted Doppler power spectrum at the current moment is: ; in, is the pixel value The number of occurrences of is the weighted Doppler power spectrum, is a logical operator, is the pixel value; The calculation formula for obtaining the accumulated histogram by accumulating the calculated histogram is: ; in, is the pixel value The cumulative histogram of ; The calculation formula for calculating the local slope of each point in the cumulative histogram is: ; in, is the pixel value The local slope of is the pixel distance in the histogram; The calculation formula for determining the pixel value corresponding to the point with the largest local slope according to the local slope of each point as the noise intensity is: ; in, is the pixel value corresponding to the noise intensity, is the local slope The k value when the maximum value is obtained; The calculation formula for denoising all parts of the weighted Doppler power spectrum at the current moment that are lower than the noise intensity to obtain the denoised Doppler power spectrum at the current moment is: ; in, is the denoised Doppler power spectrum, is the weighted Doppler power spectrum.

7. The method for extracting the envelope of the Doppler power spectrum according to claim 1, characterized in that: The calculation formula for smoothing the denoised Doppler power spectrum at the current moment to obtain the smoothed Doppler power spectrum at the current moment is: ; ; in, For the current moment The smoothed Doppler power spectrum of is the weight array of the Hanning smoothing filter, is the first Point weight value, for The denoised Doppler power spectrum at time instant.

8. The method for extracting the envelope of the Doppler power spectrum according to claim 1, characterized in that: The calculation formula for obtaining the gradient value of each frequency by performing gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment is: ; ; in, For the current moment In frequency The gradient value at is the gradient detection operator, is the gradient detection operator The value of a point, For the current moment The smoothed Doppler power spectrum at frequency The value at The calculation formula for determining a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​as candidate points according to the gradient values ​​of each frequency is: ; ; ; in, For the gradient value Sort by maximum value. is the frequency position coordinate array after the forward gradient values ​​are sorted from large to small, is the gradient value array after the forward gradient values ​​are sorted from large to small, For the current moment The one with the largest positive gradient The gradient intensity sequence of points; For the current moment The one with the largest positive gradient The frequency position coordinate sequence of points, is the number of valid candidate points, which is also the preset number; The calculation formula for sequentially calculating the pixel mean in the local area below each candidate point is: ; in, For the current moment No. The average pixel value in the local area below the candidate point, is the local range parameter, For the current moment The denoised Doppler power spectrum is The pixel value of each candidate point; The calculation formula for determining the confidence level of each candidate point as the maximum frequency position according to the forward gradient value and pixel mean of each candidate point is: ; ; ; in, For the current moment Middle The forward gradient confidence of the candidate points, is the threshold of the minimum positive gradient value, For the current moment Middle The pixel mean confidence of candidate points, is the threshold of the minimum pixel mean, For the current moment Middle The comprehensive confidence that the candidate points are the maximum frequency positions; The calculation formula for determining the candidate point with the largest confidence as the candidate maximum frequency position of the Doppler power spectrum at the current moment according to the confidence of each candidate point is: ; ; in, is the index of the candidate point with the highest confidence. for When the maximum value is obtained value, For the current moment The candidate maximum frequency position of the Doppler power spectrum.

9. A Doppler power spectrum envelope extraction system, characterized in that: The system comprises: An input module is used to continuously obtain the Doppler power spectrum at each moment; A weighting module, used for weighting the Doppler power spectra at the current moment and at multiple historical moments to obtain a weighted Doppler power spectrum at the current moment, wherein the weighted Doppler power spectrum is a Doppler power spectrum weighted with historical information; A denoising module, used for performing noise estimation on the weighted Doppler power spectrum at the current moment to obtain an estimated noise intensity, and performing denoising processing on the weighted Doppler power spectrum according to the estimated noise intensity to obtain a denoised Doppler power spectrum at the current moment, wherein the denoised Doppler power spectrum is a Doppler power spectrum obtained after denoising the weighted Doppler power spectrum; A smoothing module is used to smooth the denoised Doppler power spectrum at the current moment to obtain a smoothed Doppler power spectrum at the current moment, wherein the smoothed Doppler power spectrum is a Doppler power spectrum obtained by smoothing the denoised Doppler power spectrum; An identification module, used for identifying the maximum frequency position of the smoothed Doppler power spectrum at the current moment to obtain a candidate maximum frequency position of the Doppler power spectrum at the current moment; A correction module, used for correcting the candidate maximum frequency position of the Doppler power spectrum at the center moment according to the candidate maximum frequency positions of the Doppler power spectrum at the current moment and multiple historical moments to obtain the maximum frequency position of the Doppler power spectrum at the center moment, wherein the center moment is the moment of the center position between the current moment and multiple historical moments, and the maximum frequency position is the frequency position determined after the candidate maximum frequency position of the Doppler power spectrum at the center moment is corrected; A display module, used for connecting the maximum frequency positions of the corrected Doppler power spectrum at each moment to obtain a maximum frequency envelope of the Doppler power spectrum, and displaying the maximum frequency envelope of the Doppler power spectrum; The identification module comprises: A gradient value determination unit, used to perform gradient detection on each frequency in the smoothed Doppler power spectrum at the current moment to obtain a gradient value of each frequency; A candidate point determination unit, used to determine a preset number of frequency position coordinates arranged in descending order of forward gradient values ​​as candidate points according to the gradient values ​​of each frequency; A pixel mean calculation unit, used to sequentially calculate the pixel mean in the local area below each candidate point; A confidence determination unit, used to determine the confidence of each candidate point as the maximum frequency position according to the forward gradient value and pixel mean of each candidate point; A candidate maximum frequency position determination unit, used to determine the candidate point with the highest confidence as the candidate maximum frequency position of the Doppler power spectrum at the current moment according to the confidence of each candidate point; The correction modules include: A central moment determination unit, used to determine a central moment located at a central position between the current moment and the multiple historical moments according to the current moment and the multiple historical moments; A first coordinate difference calculation unit is used to calculate the coordinate differences between the candidate maximum frequency position of the Doppler power spectrum at the center moment and the candidate maximum frequency position of the Doppler power spectrum at the previous moment and the next moment respectively; A first threshold judgment unit, used to determine a coordinate difference value according to the two coordinate differences, and to judge whether the coordinate difference value is less than a difference threshold value; A first maximum frequency position determining unit is used to determine the maximum frequency position of the Doppler power spectrum at the center moment as the corresponding candidate maximum frequency position when the first threshold judgment unit judges that the coordinate difference is less than the difference threshold; The correction module also includes: A new center moment determination unit, configured to use the previous moment and the next moment of the center moment as the new center moment respectively when the first threshold determination unit determines that the coordinate difference is not less than the difference threshold; A second coordinate difference calculation unit is used to calculate the coordinate differences between the candidate maximum frequency positions of the Doppler power spectra at two new center moments and the candidate maximum frequency positions of the Doppler power spectra at the previous moment and the next moment respectively; A second threshold judgment unit is used to determine two new coordinate difference values ​​according to two coordinate differences respectively determined at the two new center moments, and to judge whether the two new coordinate difference values ​​are both less than a difference threshold value; The second maximum frequency position determining unit is used to determine the maximum frequency position of the Doppler power spectrum at the center moment as the corresponding candidate maximum frequency position when the second threshold judging unit judges that the differences of the two new coordinate differences are both less than the difference threshold.

10. An ultrasonic device, characterized in that: The ultrasound device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor runs the computer program, the method for extracting the envelope of the Doppler power spectrum according to any one of claims 1 to 8 is executed.

11. A medium storing a program, characterized in that: When the program is executed by a processor, the method for extracting the envelope of a Doppler power spectrum according to any one of claims 1 to 8 is implemented.

Images