Adaptive filtering method based on dynamic noise environment recognition and ultrasonic equipment

Abstract

The invention relates to the technical field of ultrasonic equipment, in particular to a self-adaptive filtering method based on dynamic noise environment recognition and ultrasonic equipment. The method comprises the following steps: S1, acquiring environmental noise data acquired by a sensor of ultrasonic equipment, and performing spectral analysis on the environmental noise data; s2, extracting noise characteristic parameters of the environmental noise data, classifying the noise characteristic parameters by adopting a machine learning model, and identifying the type of a noise environment; s3, mapping the extracted noise characteristic parameters to filtering parameters in real time, and constructing a filter by adopting the filtering parameters; and S4, carrying out filtering processing on the ultrasonic data of the ultrasonic equipment through the constructed filter. The invention provides a self-adaptive filtering method based on dynamic noise environment recognition, which can dynamically adjust filtering parameters according to a noise environment detected in real time and improve the signal quality and image definition of handheld ultrasonic equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic equipment, and in particular to an adaptive filtering method based on dynamic noise environment recognition and ultrasonic equipment. Background Art

[0002] Handheld ultrasound devices have gradually become an important tool in medical diagnosis due to their portability and ease of use. However, due to the diverse and complex environments in which they are used, ultrasound signals are often interfered with by different types of noise, affecting image quality and diagnostic accuracy.

[0003] Existing ultrasound equipment usually uses fixed-parameter filtering algorithms for noise suppression, which cannot effectively cope with changes in different noise environments. This leads to poor filtering effects in some cases and may even introduce additional artifacts. In addition, existing technologies lack the ability to identify and dynamically adjust the noise environment in real time, and cannot provide stable image quality in complex and changing usage environments. Summary of the invention

[0004] The purpose of the present invention is to provide an adaptive filtering method based on dynamic noise environment recognition to solve the above technical problems;

[0005] Another object of the present invention is to provide an ultrasonic device to solve the above technical problems.

[0006] The technical problem solved by the present invention can be achieved by adopting the following technical solutions:

[0007] An adaptive filtering method based on dynamic noise environment recognition, comprising:

[0008] Step S1, obtaining environmental noise data collected by a sensor of an ultrasonic device, and performing spectrum analysis on the environmental noise data;

[0009] Step S2, extracting noise characteristic parameters of the environmental noise data, classifying the noise characteristic parameters using a machine learning model, and identifying the type of the noise environment;

[0010] Step S3, mapping the extracted noise characteristic parameters to filter parameters in real time, and constructing a filter using the filter parameters;

[0011] Step S4: filtering the ultrasound data of the ultrasound device through the constructed filter.

[0012] Preferably, step S3 comprises,

[0013] Step S31, extracting the noise characteristic parameters, the noise characteristic parameters at least including peak frequency, bandwidth and intensity, and mapping the noise characteristic parameters to the filtering parameters;

[0014] Step S32, establishing the filter by using the filtering parameters obtained by mapping;

[0015] Step S33: an adaptive adjustment method is adopted to adjust the filter using the filter coefficients updated in real time.

[0016] Preferably, the mapping formula of the filtering parameters in step S31 is:

[0017] (f c ,N)=M(f p ,B w ,I n )

[0018] Among them, (f c , N) represents the filtering parameters, f c is the cut-off frequency of the filter, and N is the order of the filter; (fp, B w ,I n ) represents the noise characteristic parameter, f p is the peak frequency, B w is the bandwidth, I n is the intensity; M represents the mapping function.

[0019] Preferably, in step S32, the filter adopts a Butterworth low-pass filter, and the formula is:

[0020]

[0021] Wherein, H(s) represents the filter, s represents the complex frequency variable, N represents the order of the filter, ω c represents the cut-off angular frequency, ω c =2πf c .

[0022] Preferably, in step S33, the adaptive adjustment formula of the filter is:

[0023]

[0024] e(n)=d(n)-y(n)

[0025] w i (n+1)=w i (n)+μe(n)x(ni)

[0026] Where y(n) represents the filter output, x(n) represents the input signal, d(n) represents the desired signal, e(n) represents the error signal, μ represents the step size factor, and w i (n) represents the filter coefficient, and N represents the order of the filter.

[0027] Preferably, step S32 further includes determining whether the type of the noise environment has changed, and if so, adjusting the mapping function M, and calculating the cutoff frequency f of the filter with the adjusted mapping function. c and order N; if unchanged, the cutoff frequency f of the filter is calculated using the current mapping function c and order N.

[0028] Preferably, it also includes step S34, calculating the signal-to-noise ratio of the detection signal after filtering by the filter, judging whether the signal-to-noise ratio reaches the expected value, if not, adjusting the step size factor μ and the filter coefficient, and returning to execute step S33; if reached, executing step S4.

[0029] Preferably, the method further includes step S5, calculating the signal-to-noise ratio of the ultrasonic data after filtering by the filter, evaluating the filtering effect of the filter, and optimizing the filtering parameters in step S3 according to the evaluation result.

[0030] An ultrasonic device is provided with a signal processing module for implementing the adaptive filtering method based on dynamic noise environment recognition, the signal processing module comprising:

[0031] A noise detection unit, used to obtain the environmental noise data and perform spectrum analysis on the environmental noise data;

[0032] A noise identification unit, connected to the noise detection unit, configured to extract noise characteristic parameters of the environmental noise data, classify the noise characteristic parameters, and identify the type of the noise environment;

[0033] A parameter adjustment unit is connected to the noise identification unit and is used to map the extracted noise characteristic parameters to the filtering parameters and construct the filter using the filtering parameters.

[0034] Preferably, it also includes an effect evaluation unit, which is connected to the parameter adjustment unit and is used to calculate the signal-to-noise ratio of the filter signal filtered by the filter and evaluate the filtering effect. The parameter adjustment unit adjusts the filter according to the evaluation result.

[0035] Beneficial effects of the present invention: Due to the adoption of the above technical scheme, the present invention provides an adaptive filtering method based on dynamic noise environment recognition, which can dynamically adjust the filtering parameters according to the noise environment detected in real time, thereby improving the signal quality and image clarity of the handheld ultrasound device. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the steps of an adaptive filtering method based on dynamic noise environment recognition in an embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of step S3 in an embodiment of the present invention;

[0038] Figure 3 Schematic diagram of the architecture of the signal processing module in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0041] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0042] An adaptive filtering method based on dynamic noise environment recognition, such as Figure 1 , Figure 2 As shown, including

[0043] Step S1, obtaining environmental noise data collected by a sensor of an ultrasonic device, and performing spectrum analysis on the environmental noise data;

[0044] Step S2, extracting noise characteristic parameters of environmental noise data, classifying the noise characteristic parameters using a machine learning model, and identifying the type of noise environment;

[0045] Step S3, mapping the extracted noise characteristic parameters to the filtering parameters in real time, and constructing a filter using the filtering parameters;

[0046] Step S4: filtering the ultrasonic data of the ultrasonic device through the constructed filter.

[0047] Specifically, the present invention provides an adaptive filtering method based on dynamic noise environment recognition, which can dynamically adjust filtering parameters according to the noise environment detected in real time, thereby improving the signal quality and image clarity of a handheld ultrasound device.

[0048] Specifically, in step S1, sensors in the handheld ultrasound device, such as microphones, accelerometers, etc., are used to collect environmental noise data in real time, and spectrum analysis is performed on the collected environmental noise data to identify frequency components and intensities of the environmental noise data.

[0049] More specifically, step S1 also includes preprocessing the extracted environmental noise data, such as removing DC components, normalizing, etc., dividing the preprocessed environmental noise data into frames, performing spectrum analysis on each frame of data, and extracting characteristic parameters such as peak frequency, bandwidth, and intensity. It should be noted that the extracted characteristic data needs to be standardized to ensure that the characteristic values ​​are on the same scale.

[0050] In step S2, the main characteristic parameters of the noise, such as peak frequency, bandwidth, intensity, etc., are extracted. The present invention uses a machine learning algorithm model of K-means clustering to classify the noise characteristics. K-means clustering divides the environmental noise data into several clusters to identify different types of noise environments.

[0051] In a preferred embodiment, step S3 includes:

[0052] Step S31, extracting noise characteristic parameters, the noise characteristic parameters at least including peak frequency, bandwidth and intensity, and mapping the noise characteristic parameters to filtering parameters;

[0053] Step S32, establishing a filter by using the filter parameters obtained by mapping;

[0054] Step S33, using an adaptive adjustment method to adjust the filter using the filter coefficients updated in real time. In a preferred embodiment, the mapping formula of the filter parameters in step S31 is:

[0055] (f c ,N)=M(f p ,B w ,I n )

[0056] Among them, (f c ,N) represents the filtering parameters, f c is the cutoff frequency of the filter, N is the order of the filter; (fp,B w ,I n ) represents the noise characteristic parameter, f p is the peak frequency, B w is the bandwidth, I n is the intensity; M represents the mapping function.

[0057] Specifically, the present invention extracts noise characteristic parameters, such as peak frequency f p Bandwidth B w , Strength I n , is mapped to the filter parameters used to design the filter through the mapping function,

[0058] The mapping formula is:

[0059] (f c,N)=M(f p ,B w ,I n )

[0060] Among them, (f c ,N) represents the filtering parameters, f c is the cutoff frequency of the filter, and N is the order of the filter;

[0061] The present invention selects the filtering parameters of the filter to be suitable for the identified noise environment type based on the mapping relationship between the noise characteristic parameters and the filtering parameters, facilitates the design of different filter types by selecting different filtering parameters according to the type of the noise environment, and improves the adaptive filtering ability of the filter to noise.

[0062] In a preferred embodiment, in step S32, the filter adopts a Butterworth low-pass filter, and the formula is:

[0063]

[0064] Where H(s) represents the filter, s represents the complex frequency variable, N represents the order of the filter, ω c represents the cut-off angular frequency, ω c =2πf c .

[0065] Specifically, the present invention adopts a low-pass Butterworth filter design method. The Butterworth filter has a smooth frequency response and no ripples. The key to the design method of the Butterworth filter is the filter cutoff frequency f c and the filter order N;

[0066] The present invention maps the noise characteristic parameters to the filter parameters, and the filter parameters obtained by mapping include the filter cutoff frequency f c and filter order N to design Butterworth filter,

[0067] The formula is,

[0068]

[0069] Where H(s) represents the filter, s represents the complex frequency variable, N represents the order of the filter, ω c represents the cut-off angular frequency, ω c =2πf c .

[0070] The present invention uses noise characteristic parameters, such as peak frequency f p Bandwidth B w , Strength I n, mapped to the filter parameters used to design the filter through the mapping function, such as the filter cutoff frequency f c and the order N of the filter, a Butterworth low-pass filter that meets specific requirements can be designed so that the filter has a smooth frequency response and good attenuation characteristics at the cutoff frequency.

[0071] The filter cutoff frequency f c It indicates the frequency at which the filter begins to significantly attenuate the signal. The order N of the filter indicates the steepness of the filter response. The higher the order, the faster the frequency response of the filter changes near the cutoff frequency.

[0072] In a preferred embodiment, in step S33, the adaptive adjustment formula of the filter is:

[0073]

[0074] e(n)=d(n)-y(n)

[0075] w i (n+1)=w i (n)+μe(n)x(ni)

[0076] Where y(n) represents the filter output, x(n) represents the input signal, d(n) represents the desired signal, e(n) represents the error signal, μ represents the step size factor, and w i (n) represents the filter coefficient, and N represents the order of the filter.

[0077] Specifically, an LMS (Least Mean Squares) adaptive filter is used to perform filtering using filter parameters updated in real time.

[0078] The update formula of the LMS algorithm is:

[0079]

[0080] e(n)=d(n)-y(n)

[0081] w i (n+1)=w i (n)+μe(n)x(ni)

[0082] Where y(n) represents the filter output, x(n) represents the input signal, d(n) represents the desired signal, e(n) represents the error signal, μ represents the step size factor, and w i (n) represents the filter coefficient, and N represents the order of the filter.

[0083] The present invention adopts LMS (Least Mean Squares) adaptive filter. LMS is a method for optimizing filter parameters by minimizing errors. The optimization steps of the present invention are as follows:

[0084] Step 1: Input signal and output filtering: y(n) represents the filter output, and x(n) represents the input signal. Filter coefficient w i (n) is used to perform weighted summation on the input signal x(n) to generate the output signal y(n);

[0085] Step 2, error calculation: The error signal e(n) is the difference between the expected signal d(n) and the filtered output y(n), that is, e(n) = d(n) - y(n). The error signal e(n) reflects the deviation between the current output of the filter and the expected signal.

[0086] Step 3, coefficient update: filter coefficient w i (n) is updated in real time according to the error e(n) and the input signal x(ni). The update formula is,

[0087] w i (n+1)=w i (n)+μe(n)x(ni)

[0088] Among them, μ represents the step size factor, which is used to control the amplitude of each update. The choice of step size factor μ will affect the convergence speed and stability of the algorithm.

[0089] The present invention continuously adjusts the filter coefficients through the LMS algorithm to minimize the output error, so that the filter performance gradually approaches the desired target.

[0090] In a preferred embodiment, step S32 further includes determining whether the type of the noise environment has changed, and if so, adjusting the mapping function M, and calculating the filter cutoff frequency f with the adjusted mapping function. c and order N; if it does not change, the filter cutoff frequency f is calculated using the current mapping function c and order N.

[0091] Specifically, the present invention dynamically adjusts the filter parameters according to the changes in the real-time noise environment. Whenever a change in the noise environment is detected, the mapping function M is recalculated and the filter cutoff frequency f is updated. c and order N, and redesign the filter.

[0092] In a preferred embodiment, step S34 is also included, calculating the signal-to-noise ratio of the detection signal after filtering, and judging whether the signal-to-noise ratio reaches the expected value. If not, the step size factor μ and the filter coefficient are adjusted, and the process returns to step S33; if so, step S4 is executed.

[0093] Specifically, the present invention evaluates the filtering effect by calculating the signal-to-noise ratio (SNR) of the filtered signal. If the SNR does not reach the expected value, the step size factor μ and the filtering parameters are adjusted to form a closed-loop feedback mechanism to further optimize the filtering effect.

[0094] In a preferred embodiment, the method further includes step S5, calculating the signal-to-noise ratio of the ultrasonic data after filtering by the filter, evaluating the filtering effect of the filter, and optimizing the filtering parameters in step S3 according to the evaluation result.

[0095] Specifically, the present invention adopts an adaptive filtering parameter adjustment method to evaluate the filtering effect by calculating the signal-to-noise ratio (SNR) of the ultrasonic data after filtering, and further optimizes the filtering parameters according to the evaluation results to form a closed-loop feedback mechanism.

[0096] An ultrasonic device, such as Figure 3 As shown, there is a signal processing module 1 for implementing an adaptive filtering method based on dynamic noise environment recognition, and the signal processing module 1 includes:

[0097] A noise detection unit 2 is used to obtain environmental noise data and perform spectrum analysis on the environmental noise data;

[0098] The noise identification unit 3 is connected to the noise detection unit 2 and is used to extract noise characteristic parameters of the environmental noise data, classify the noise characteristic parameters, and identify the type of the noise environment;

[0099] The parameter adjustment unit 4 is connected to the noise identification unit 3 and is used to map the extracted noise characteristic parameters to the filtering parameters and construct the filter using the filtering parameters.

[0100] Specifically, the present invention integrates an adaptive filtering method based on dynamic noise environment recognition into the signal processing module 1 of the ultrasound device. The present invention is mainly applicable to handheld ultrasound devices, and can accelerate the method operation through built-in hardware such as FPGA chips or ASIC integrated circuits to improve operating efficiency.

[0101] The ultrasonic device of the present invention can dynamically adjust the filtering parameters according to the real-time noise environment, effectively suppress different types of noise, and improve the quality and clarity of ultrasonic images. Compared with the traditional fixed parameter filtering method, the present invention shows better adaptability and stability in complex and changeable use environments.

[0102] In a preferred embodiment, it also includes an effect evaluation unit 5, which is connected to the parameter adjustment unit 4 and is used to calculate the signal-to-noise ratio of the filter signal and evaluate the filtering effect. The parameter adjustment unit 4 adjusts the filter according to the evaluation result.

[0103] In a specific embodiment, the present invention integrates an adaptive filtering method based on dynamic noise environment recognition into a handheld ultrasound device for use in hospital emergency rooms and outdoor rescue scenarios. Through the adaptive filtering method of the present invention, the ultrasound device can automatically identify and suppress noise interference in the environment, improve the clarity of the ultrasound image, and help doctors make diagnoses quickly and accurately in complex environments. Experimental results show that the signal-to-noise ratio of ultrasound images of the device using the algorithm of the present invention is improved by more than 30%.

[0104] The above description is only a preferred embodiment of the present invention, and does not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. An adaptive filtering method based on dynamic noise environment recognition, characterized in that: include, Step S1, obtaining environmental noise data collected by a sensor of an ultrasonic device, and performing spectrum analysis on the environmental noise data; Step S2, extracting noise characteristic parameters of the environmental noise data, classifying the noise characteristic parameters using a machine learning model, and identifying the type of the noise environment; Step S3, mapping the extracted noise characteristic parameters to filter parameters in real time, and constructing a filter using the filter parameters; Step S4: filtering the ultrasound data of the ultrasound device through the constructed filter.

2. The adaptive filtering method based on dynamic noise environment recognition according to claim 1, characterized in that: Step S3 comprises, Step S31, extracting the noise characteristic parameters, the noise characteristic parameters at least including peak frequency, bandwidth and intensity, and mapping the noise characteristic parameters to the filtering parameters; Step S32, establishing the filter by using the filtering parameters obtained by mapping; Step S33: an adaptive adjustment method is adopted to adjust the filter using the filter coefficients updated in real time.

3. The adaptive filtering method based on dynamic noise environment recognition according to claim 2 is characterized in that: The mapping formula of the filtering parameters in step S31 is: (f c ,N)=M(f p ,B w ,I n ) Among them, (f c , N) represents the filtering parameters, f c is the cut-off frequency of the filter, and N is the order of the filter; (fp, B w ,I n ) represents the noise characteristic parameter, f p is the peak frequency, B w is the bandwidth, I n is the intensity; M represents the mapping function.

4. The adaptive filtering method based on dynamic noise environment recognition according to claim 2 is characterized in that: In step S32, the filter adopts a Butterworth low-pass filter, and the formula is: Wherein, H(s) represents the filter, s represents the complex frequency variable, N represents the order of the filter, ω c represents the cut-off angular frequency, ω c =2πf c .

5. The adaptive filtering method based on dynamic noise environment recognition according to claim 2, characterized in that: In step S33, the adaptive adjustment formula of the filter is: e(n)=d(n)-y(n) w i (n+1)=w i (n)+μe(n)x(n-i) Among them, y(n) represents the filter output, x(n) represents the input signal, d(n) represents the desired signal, e(n) represents the error signal, μ represents the step size factor, and w i (n) represents the filter coefficient, and N represents the order of the filter.

6. The adaptive filtering method based on dynamic noise environment recognition according to claim 3 is characterized in that: Step S32 also includes determining whether the type of the noise environment has changed. If so, adjusting the mapping function M to calculate the cutoff frequency f of the filter using the adjusted mapping function. c and order N; if unchanged, the cutoff frequency f of the filter is calculated using the current mapping function c and order N.

7. The adaptive filtering method based on dynamic noise environment recognition according to claim 5, characterized in that: The method further includes step S34, calculating the signal-to-noise ratio of the detection signal after filtering by the filter, and judging whether the signal-to-noise ratio reaches the expected value. If not, adjusting the step size factor μ and the filter coefficient, and returning to step S33; if reached, executing step S4.

8. The adaptive filtering method based on dynamic noise environment recognition according to claim 3, characterized in that: The method further includes step S5, calculating the signal-to-noise ratio of the ultrasonic data after filtering by the filter, evaluating the filtering effect of the filter, and optimizing the filtering parameters in step S3 according to the evaluation result.

9. An ultrasonic device, characterized in that: A signal processing module (1) is provided for implementing the adaptive filtering method based on dynamic noise environment recognition as claimed in any one of claims 1 to 8, wherein the signal processing module (1) comprises: A noise detection unit (2), used for acquiring the environmental noise data and performing spectrum analysis on the environmental noise data; A noise identification unit (3), connected to the noise detection unit (2), is used to extract noise characteristic parameters of the environmental noise data, classify the noise characteristic parameters, and identify the type of the noise environment; A parameter adjustment unit (4) is connected to the noise identification unit (3) and is used to map the extracted noise characteristic parameters to the filtering parameters and construct the filter using the filtering parameters.

10. The ultrasonic device according to claim 9, characterized in that It also includes an effect evaluation unit (5), which is connected to the parameter adjustment unit (4) and is used to calculate the signal-to-noise ratio of the filter signal and evaluate the filtering effect. The parameter adjustment unit (4) adjusts the filter according to the evaluation result.

Images