Long bone detection method based on guided wave signal and echo signal

Abstract

The invention discloses a long bone detection method based on a guided wave signal and an echo signal, which belongs to the technical field of bone detection, and specifically comprises the following steps: firstly, transmitting voltage to an ultrasonic transceiving probe by adopting a high-voltage pulse transmitting circuit, and respectively receiving the guided wave signal and the echo signal after an axial propagation mode and a pulse echo mode in sequence; secondly, hardware preprocessing is carried out on the received guided wave signals and echo signals; interference of organization signals is removed from the preprocessed guided wave signals through self-adaptive filtering; separating the preprocessed echo signal by using a time window to remove interference; performing mode separation on the guided wave signal subjected to interference removal, and obtaining an axial development parameter under the conditions of a time intercept parameter and a thickness parameter; performing phase analysis and intensity difference analysis on the echo signals subjected to interference removal to obtain transverse development parameters; and finally, fusing the obtained axial development parameters and transverse development parameters to obtain long bone development parameters.

Description

Technical Field

[0001] The present invention belongs to the technical field of bone detection, and in particular to a long bone detection method based on waveguide signals and echo signals. Background Art

[0002] Studies have shown that excessive calcium supplementation can cause edema, sweating, anorexia, nausea, constipation and indigestion, and in severe cases, hypercalciuria. At the same time, excessive calcium supplementation in children may also limit brain development and affect growth. If the blood calcium concentration is too high, calcium will affect vision if it is deposited around the cornea, affect heart function if it is deposited on the heart valve, and aggravate vascular sclerosis if it is deposited on the blood vessel wall. Therefore, it is necessary to test children's bone health conditions such as bone development and bone calcium content. The earliest company to study this type of equipment was Israel's Sunlight Company. As early as the 1990s, the company launched the Omnisenst series of ultrasonic bone strength meters. This series of equipment currently has a relatively complete database of Asian children and occupies most of the market. The device reflects bone development by measuring the ultrasonic sound velocity of the tibia and radius. However, the health of bones is not only related to the sound velocity parameter, but also to the thickness of the long bone cortex. In addition, the ultrasonic sound velocity can be divided into transverse wave velocity and longitudinal wave velocity according to the vibration mode of the ultrasonic wave, and there are also different sound velocity parameters such as phase velocity and group velocity according to the characteristics of the wave. The device only uses a single velocity parameter, which is difficult to fully reflect the different characteristics of bones and has obvious deficiencies. And clinical performance shows that the device is not sensitive enough to the bone condition of children over 3 years old.

[0003] Domestic research in this area is mainly focused on application research and theoretical research. In the application research, clinicians have conducted a lot of research on the clinical application of ultrasonic bone density meter. For example, Shanxi Provincial Maternal and Child Health Hospital tested the bone density of healthy children aged 0-3 in Taiyuan, Capital Medical University tested the bone strength of 10,401 children aged 5 months to 18 in the northern region, and Kunshan Maternal and Child Health Center measured and analyzed the ultrasonic bone density of 813 healthy children in Kunshan. The equipment used by the above research institutes is Omnisenst7000p ultrasonic bone density meter of Sunlight. Clinical results show that the ultrasonic bone density equipment can effectively reflect the bone health of children, but there are also problems such as low detection rate, deviation in overall results and mostly used for foot measurement, which makes patients and doctors feel uncomfortable. Theoretical research is mainly the research on long bone waveguide signal processing. This direction has been a hot topic in the field of medical ultrasound in recent years. Domestic and foreign scholars have done a lot of valuable research work. Foreign research directions focus on measurement methods, signal processing, parameter acquisition and equipment research. At present, the development of equipment in China is slow. Summary of the invention

[0004] The purpose of the present invention is to provide a long bone detection method based on waveguide signals and echo signals to solve the problems of the prior art that the measurement part is single, the accuracy is low, and it is difficult to accurately measure the bones of young children.

[0005] To achieve the above object, the present invention provides a long bone detection method based on a waveguide signal and an echo signal, comprising the following steps:

[0006] Step 1, using a high-voltage pulse transmitting circuit to transmit voltage to the ultrasonic transceiver probe, and receiving the waveguide signal and the echo signal respectively after passing through the axial propagation mode and the pulse echo mode in sequence;

[0007] Step 2: preprocessing the waveguide signal and echo signal received in step 1 by hardware;

[0008] Step 3: Adaptively filter the pre-processed waveguide signal to remove interference from the tissue signal; separate the pre-processed echo signal using a time window to remove interference;

[0009] Step 4: Perform mode separation on the waveguide signal after interference removal in step 3, and then obtain the axial growth parameter S under the conditions of time intercept parameter and thickness parameter. L ;

[0010] Step 5: Perform phase analysis and intensity difference analysis on the echo signal after interference removal in step 3 to obtain the lateral growth parameter S H ;

[0011] Step 6: Fusion the axial development parameters obtained in step 4 and the transverse development parameters obtained in step 5 to obtain long bone development parameters.

[0012] Preferably, the hardware preprocessing in step 2 is specifically as follows: the received waveguide signal and echo signal are sequentially input into a protection circuit, a low-noise amplifier circuit, an anti-aliasing filter, an adjustable gain amplifier, high-speed AD sampling, FPGA and USB communication to convert them into digital signals.

[0013] Preferably, the high-voltage pulse transmitting circuit transmits an ultrasonic pulse wave with an adjustable frequency of 100K-10MHz, and the peak value of the transmitting voltage is 100V.

[0014] Preferably, the adjustable range of the adjustable gain amplifier is 0 to 60 dB.

[0015] Preferably, the sampling rate of the high-speed AD sampling is 40Msps.

[0016] Preferably, in step 3, the calculation expression for removing interference of tissue signals from the preprocessed waveguide signal by adaptive filtering is as follows:

[0017] e(t)=d(t)-y(t)=s(t)+n0(t)-y(t) (1)

[0018] E[e 2 (t)]=E[s 2 (t)]+E[(n0(t)-y(t)) 2 ]+2E[s(t)(n0(t)-y(t))] (2)

[0019] Calculate the minimum value E[e 2 (t)] min , satisfying E[(n0(t)-y(t)) 2 ]=E[(n0(t)-y(t)) 2 ] min , specifically:

[0020] E[e 2 (t)] min =E[s 2 (t)]+E[(n0(t)-y(t)) 2 ] min (3)

[0021] According to equations (3) and (2), the weights of the filter are calculated using the minimum mean square error criterion, and adaptive filtering is performed to obtain s(t);

[0022] Where s(t) is the waveguide signal in the bone, n0(t) is the ultrasonic signal in the tissue, d(t) = s(t) + n0(t), y(t) is the output of the filter under the optimal criterion, y(t) is close to s(t), and e(t) is the error signal between y(t) and s(t).

[0023] Preferably, the specific contents of using a time window to separate and remove interference from the preprocessed echo signal in step 3 are as follows:

[0024] S31, preprocessing: preprocessing the input echo signal, including filtering and noise reduction;

[0025] S32, setting a time window: determining the time window size according to actual conditions. The smaller the time window, the less noise interference is detected.

[0026] S33, start time window: start the selected time window and start running;

[0027] S34, data acquisition and processing: real-time acquisition of echo signal data, setting corresponding thresholds, and analyzing and judging the real-time acquisition of echo signal data; if the set threshold is met, the signal in the current time period is clean; otherwise, the signal in the current time period is a noise point;

[0028] S35, fusion threshold: a fusion threshold is set for the points where noise is identified by the average value fusion method, and the data of these noise points are fused;

[0029] S36, close the time window: when a complete signal cycle ends, close the time window and wait for the start of the next signal cycle.

[0030] Preferably, the specific content of performing mode separation on the waveguide signal after interference removal in step 3 in step 4 is as follows:

[0031] S41. Decompose the waveguide signal by empirical mode decomposition (EMD) and Wiener filter cross-correlation to obtain a set of intrinsic mode functions (IMFs). The calculation expression is as follows:

[0032]

[0033] In the formula, IMF i represents the intrinsic mode function, r(t) is the residual term; each IMF component corresponds to a different frequency band from high to low, and the frequency components contained are also different, and the frequency components are specifically thickness or time intercept;

[0034] S42. Calculate the time-frequency coordinates of the maximum energy point of the intrinsic mode function, obtain the transformation curve of the frequency-thickness product and the group velocity, use the curve fitting function to fit the group velocity dispersion curves of the A0 and S0 modes, and separate the A0 and S0 mode signals from the guided wave signal.

[0035] Preferably, the calculation expression of the axial growth parameter is obtained under the conditions of the time intercept parameter and the thickness parameter as follows:

[0036] S L =F(s1(t),s2(t)) (5)

[0037] In the formula, S L represents the axial growth parameter, s1(t) represents the guided wave signal under the time intercept parameter, and s2(t) represents the guided wave signal under the thickness parameter.

[0038] Preferably, in step 6, the axial development parameters obtained in step 4 and the transverse development parameters obtained in step 5 are fused to obtain the calculation expression of the long bone development parameters as follows:

[0039] S bone =f(S L ,S H ) (6)

[0040] In the formula, S bone represents the long bone development parameter, S L represents the axial growth parameter, S HRepresents the lateral development parameter.

[0041] Therefore, the present invention adopts the above-mentioned long bone detection method based on waveguide signal and echo signal, which has the following beneficial effects:

[0042] (1) Empirical mode decomposition and Wiener filter cross-correlation were proposed to separate signals of different modes, solving the dispersion phenomenon of ultrasonic guided waves during propagation in long bones. This method can effectively separate commonly used ultrasonic mode signals;

[0043] (2) By using a high-voltage pulse transmission circuit to transmit voltage to the ultrasonic transceiver probe, a corresponding matching filter is used during reception to improve the signal-to-noise ratio, thereby reducing the impact of noise and making a more accurate diagnosis; this method also reduces the intensity of the transmitted ultrasonic wave to a certain extent, increases the penetration depth of the ultrasonic wave, and further reduces the effect of the ultrasonic wave on the human body;

[0044] (3) By integrating the measured axial and transverse development parameters of the bones, the overall development of the bones is finally detected, thereby improving the overall detection accuracy;

[0045] (4) The received waveguide signal and echo signal are sequentially input into the protection circuit, low-noise amplifier circuit, anti-aliasing filter, adjustable gain amplifier, high-speed AD sampling, FPGA and USB communication to convert them into digital signals for simple preprocessing, converting complex signals into relatively simple and stable digital signals, making subsequent signal processing easier and faster.

[0046] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is an overall flow chart of a long bone detection method based on a waveguide signal and an echo signal of the present invention;

[0048] Figure 2 The overall architecture diagram of a long bone detection method based on a waveguide signal and an echo signal of the present invention;

[0049] Figure 3 It is a schematic diagram of the principle of measuring the cross section of a long bone in the axial propagation mode according to an embodiment of the present invention;

[0050] Figure 4 Schematic diagram of long bone cross-section measurement using a pulse echo method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. 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.

[0052] See also Figure 1-4 , a long bone detection method based on waveguide signal and echo signal, comprising the following steps:

[0053] Step 1, using a high-voltage pulse transmitting circuit to transmit voltage to the ultrasonic transceiver probe, and receiving the waveguide signal and the echo signal respectively after passing through the axial propagation mode and the pulse echo mode in sequence;

[0054] Step 2: Perform hardware preprocessing on the waveguide signal and echo signal received in step 1; specifically: input the received waveguide signal and echo signal into the protection circuit, low-noise amplifier circuit, anti-aliasing filter, adjustable gain amplifier, high-speed AD sampling, FPGA and USB communication in turn to convert them into digital signals. Among them, the high-voltage pulse transmitting circuit transmits an ultrasonic pulse wave with an adjustable frequency of 100K-10MHz, and the peak value of the transmitting voltage is 100V; the adjustable range of the adjustable gain amplifier is 0 to 60dB; the sampling rate of the high-speed AD sampling is 40Msps; FPGA is used as the control core of the control circuit; USB2.0 interface protocol is used as the data transmission circuit;

[0055] Step 3: Adaptively filter the pre-processed waveguide signal to remove interference from the tissue signal; separate the pre-processed echo signal using a time window to remove interference;

[0056] Among them, the calculation expression for removing the interference of tissue signal by adaptive filtering on the pre-processed waveguide signal is as follows:

[0057] e(t)=d(t)-y(t)=s(t)+n0(t)-y(t) (1)

[0058] E[e 2 (t)]=E[s 2 (t)]+E[(n0(t)-y(t)) 2 ]+2E[s(t)(n0(t)-y(t))] (2)

[0059] Calculate the minimum value E[e 2 (t)] min , satisfying E[(n0(t)-y(t)) 2 ]=E[(n0(t)-y(t)) 2 ] min , specifically:

[0060] E[e 2 (t)] min =E[s 2 (t)]+E[(n0(t)-y(t)) 2 ] min (3)

[0061] According to equations (3) and (2), the weights of the filter are calculated using the minimum mean square error criterion, and adaptive filtering is performed to obtain s(t);

[0062] Where s(t) is the waveguide signal in the bone, n0(t) is the ultrasonic signal in the tissue, d(t) = s(t) + n0(t), y(t) is the output of the filter under the optimal criterion, y(t) is close to s(t), and e(t) is the error signal between y(t) and s(t).

[0063] The specific contents of using time window to separate and remove interference from the preprocessed echo signal are as follows:

[0064] S31, preprocessing: preprocessing the input echo signal, including filtering and noise reduction;

[0065] S32, setting a time window: determining the time window size according to actual conditions. The smaller the time window, the less noise interference is detected.

[0066] S33, start time window: start the selected time window and start running;

[0067] S34, data acquisition and processing: real-time acquisition of echo signal data, setting corresponding thresholds, and analyzing and judging the real-time acquisition of echo signal data; if the set threshold is met, the signal in the current time period is clean; otherwise, the signal in the current time period is a noise point;

[0068] S35, fusion threshold: a fusion threshold is set for the points where noise is identified by the average value fusion method, and the data of these noise points are fused;

[0069] S36, close the time window: when a complete signal cycle ends, close the time window and wait for the start of the next signal cycle.

[0070] Step 4: Perform mode separation on the waveguide signal after interference removal in step 3, and then obtain the axial growth parameter S under the conditions of time intercept parameter and thickness parameter. L ;

[0071] The specific contents of mode separation of the waveguide signal after interference removal in step 3 are as follows:

[0072] S41. Decompose the waveguide signal by empirical mode decomposition (EMD) and Wiener filter cross-correlation to obtain a set of intrinsic mode functions (IMFs). The calculation expression is as follows:

[0073]

[0074] In the formula, IMF i represents the intrinsic mode function, r(t) is the residual term; each IMF component corresponds to a different frequency band from high to low, and the frequency components contained are also different, and the frequency components are specifically thickness or time intercept;

[0075] S42. Calculate the time-frequency coordinates of the maximum energy point of the intrinsic mode function, obtain the transformation curve of the frequency-thickness product and the group velocity, use the curve fitting function to fit the group velocity dispersion curves of the A0 and S0 modes, and separate the A0 and S0 mode signals from the guided wave signal.

[0076] The calculation expression of the axial development parameter under the conditions of time intercept parameter and thickness parameter is as follows:

[0077] S L =F(s1(t),s2(t)) (5)

[0078] In the formula, S L represents the axial growth parameter, s1(t) represents the guided wave signal under the time intercept parameter, and s2(t) represents the guided wave signal under the thickness parameter.

[0079] Step 5: Perform phase analysis and intensity difference analysis on the echo signal after interference removal in step 3 to obtain the lateral growth parameter S H ;

[0080] Step 6: The axial development parameters obtained in step 4 and the transverse development parameters obtained in step 5 are fused to obtain the long bone development parameters. The specific calculation expression is as follows:

[0081] S bone =f(S L ,S H ) (6)

[0082] In the formula, S bone represents the long bone development parameter, S L represents the axial growth parameter, S H Represents the lateral development parameter.

[0083] Therefore, the present invention adopts the above-mentioned long bone detection method based on waveguide signals and echo signals. First, a high-voltage pulse transmitting circuit is used to transmit voltage for the ultrasonic transceiver probe, and after passing through the axial propagation mode and the pulse echo mode in turn, the waveguide signal and the echo signal are received respectively; secondly, the received waveguide signal and the echo signal are preprocessed by hardware; then, the interference of the bone surface tissue signal is removed from the received waveguide signal, and different mode waveguides are obtained through the waveguide mode separation algorithm, and the correlation between different mode waveguides and children's bones is studied. According to the different mode waveguides, the ultrasonic parameters reflecting the axial development are calculated; for the received echo signal, the characteristic parameters related to the transverse direction development status of the bone are proposed through phase analysis and intensity difference analysis, and finally the two are effectively fused to obtain the final accurate bone development ultrasonic parameters; the method proposed by the present invention can measure the long bones of the human limbs, improve the overall measurement accuracy, and is suitable for children of all ages, making up for the gap in children's ultrasonic bone health detection.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A long bone detection method based on waveguide signal and echo signal, characterized in that: The following steps are involved: Step 1, using a high-voltage pulse transmitting circuit to transmit voltage to the ultrasonic transceiver probe, and receiving the waveguide signal and the echo signal respectively after passing through the axial propagation mode and the pulse echo mode in sequence; Step 2: preprocessing the waveguide signal and echo signal received in step 1 by hardware; Step 3, removing the interference of tissue signal by adaptive filtering the pre-processed waveguide signal; The pre-processed echo signal is separated and interference is removed using a time window; Step 4: Perform mode separation on the waveguide signal after interference removal in step 3, and then obtain the axial growth parameter S under the conditions of time intercept parameter and thickness parameter. L ; Step 5: Perform phase analysis and intensity difference analysis on the echo signal after interference removal in step 3 to obtain the lateral growth parameter S H ; Step 6: Fusion the axial development parameters obtained in step 4 and the transverse development parameters obtained in step 5 to obtain long bone development parameters.

2. The long bone detection method based on waveguide signal and echo signal according to claim 1, characterized in that: The hardware preprocessing in step 2 is specifically as follows: the received waveguide signal and echo signal are sequentially input into the protection circuit, low-noise amplifier circuit, anti-aliasing filter, adjustable gain amplifier, high-speed AD sampling, FPGA and USB communication to convert them into digital signals.

3. The long bone detection method based on waveguide signal and echo signal according to claim 2, characterized in that: The high-voltage pulse transmitting circuit transmits 100K-10MHz adjustable frequency ultrasonic pulse waves, and the peak value of the transmitting voltage is 100V.

4. The long bone detection method based on waveguide signal and echo signal according to claim 3, characterized in that: The adjustable gain amplifier has an adjustable range of 0 to 60dB.

5. The long bone detection method based on waveguide signal and echo signal according to claim 4, characterized in that: The sampling rate of high-speed AD sampling is 40Msps.

6. A long bone detection method based on waveguide signal and echo signal according to claim 5, characterized in that: In step 3, the calculation expression for removing the interference of tissue signal by adaptive filtering on the pre-processed waveguide signal is as follows: e(t)=d(t)-y(t)=s(t)+n0(t)-y(t)(1) E[e 2 (t)]=E[s 2 (t)]+E[(n0(t)-y(t)) 2 ]+2E[s(t)(n0(t)-y(t))] (2) Calculate the minimum value E[e 2 (t)] min , satisfying E[(n0(t)-y(t)) 2 ]=E[(n0(t)-y(t)) 2 ] min , specifically: And[and 2 (t)] min =And[s 2 (t)]+E[(n0(t)-y(t)) 2 ] min (3) According to equations (3) and (2), the weights of the filter are calculated using the minimum mean square error criterion, and adaptive filtering is performed to obtain s(t); Where s(t) is the waveguide signal in the bone, n0(t) is the ultrasonic signal in the tissue, d(t) = s(t) + n0(t), y(t) is the output of the filter under the optimal criterion, and e(t) is the error signal between y(t) and s(t).

7. A long bone detection method based on waveguide signal and echo signal according to claim 6, characterized in that: The specific contents of using time window to separate and remove interference from the preprocessed echo signal in step 3 are as follows: S31, preprocessing: preprocessing the input echo signal, including filtering and noise reduction; S32, setting a time window: determining the time window size according to actual conditions. The smaller the time window, the less noise interference is detected. S33, start time window: start the selected time window and start running; S34, data acquisition and processing: real-time acquisition of echo signal data, setting corresponding thresholds, and analyzing and judging the real-time acquisition of echo signal data; if the set threshold is met, the signal in the current time period is clean; otherwise, the signal in the current time period is a noise point; S35, fusion threshold: a fusion threshold is set for the points where noise is identified by the average value fusion method, and the data of these noise points are fused; S36, close the time window: when a complete signal cycle ends, close the time window and wait for the start of the next signal cycle.

8. The long bone detection method based on waveguide signal and echo signal according to claim 7, characterized in that: The specific contents of mode separation of the waveguide signal after interference removal in step 3 in step 4 are as follows: S41. Decompose the waveguide signal by empirical mode decomposition (EMD) and Wiener filter cross-correlation to obtain a set of intrinsic mode functions (IMFs). The calculation expression is as follows: In the formula, IMF i represents the intrinsic mode function, r(t) is the residual term; each IMF component corresponds to a different frequency band from high to low, and the frequency components contained are also different, and the frequency components are specifically thickness or time intercept; S42. Calculate the time-frequency coordinates of the maximum energy point of the intrinsic mode function, obtain the transformation curve of the frequency-thickness product and the group velocity, use the curve fitting function to fit the group velocity dispersion curves of the A0 and S0 modes, and separate the A0 and S0 mode signals from the guided wave signal.

9. The long bone detection method based on waveguide signal and echo signal according to claim 8, characterized in that: The calculation expression of the axial development parameter under the conditions of time intercept parameter and thickness parameter is as follows: S L =F(s1(t),s2(t)) (5) In the formula, S L represents the axial growth parameter, s1(t) represents the guided wave signal under the time intercept parameter, and s2(t) represents the guided wave signal under the thickness parameter.

10. The long bone detection method based on waveguide signal and echo signal according to claim 9, characterized in that: In step 6, the axial development parameters obtained in step 4 and the transverse development parameters obtained in step 5 are fused to obtain the calculation expression of the long bone development parameters as follows: S bone =f(S L ,S H ) (6) In the formula, S bone represents the long bone development parameter, S L represents the axial growth parameter, S H Represents the lateral development parameter.

Images