Color ultrasonic imaging diagnosis system based on smart phone and tablet computer

Abstract

The invention discloses a color ultrasonic imaging diagnostic system based on a smart phone and a tablet computer, which belongs to the technical field of medical diagnosis and comprises a signal acquisition module, a self-adaptive signal enhancement module and a signal detection module. The signal acquisition module evaluates the tissue condition by calculating a group domain fluctuation value and a wave rate in a group, and ensures that a detection signal which accurately reflects tissue characteristics is acquired; the self-adaptive signal enhancement module comprehensively uses a band elimination filter, a self-adaptive least mean square filtering algorithm and a short-time Fourier transform technology to effectively remove noise and enhance useful signals; the signal detection module evaluates the quality of a detection enhancement signal according to a signal-to-noise ratio and a mean square error, transmits the signal to an image processor for imaging for diagnosis if the signal reaches the standard, and automatically adjusts parameters of the self-adaptive enhancement module to re-optimize the signal if the signal does not reach the standard, thereby ensuring that the signal entering an imaging link is reliable, effectively improving the diagnosis accuracy and reliability, and improving the diagnosis accuracy and reliability. And the portable development of the ultrasonic imaging technology is promoted.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical diagnosis, and particularly relates to a color ultrasonic imaging diagnosis system based on a smart phone and a tablet computer. Background Art

[0002] In the field of medical diagnosis, ultrasonic imaging technology is widely used due to its advantages such as radiation-free and simple operation; traditional ultrasonic imaging equipment is large in volume and expensive in price, which limits its use in grass-roots medical units and mobile medical scenarios; with the popularization of mobile devices such as smart phones and tablet computers, their powerful computing and display capabilities provide the possibility for the portable development of ultrasonic imaging technology; However, there are many problems with the existing ultrasonic imaging systems based on mobile devices; on the one hand, it is difficult to dynamically adjust the acquisition frequency according to the actual situation of human tissues during the signal acquisition process, resulting in the acquired detection signals being unable to accurately reflect tissue characteristics and affecting the diagnostic accuracy; on the other hand, the acquired signals are easily interfered by noise, and the subsequent signal processing effect is not good, resulting in poor final imaging quality, increasing the difficulty of doctor diagnosis and the risk of misdiagnosis. For this reason, we propose a color ultrasonic imaging diagnosis system based on a smart phone and a tablet computer. Summary of the Invention

[0003] The purpose of the present invention is to provide a color ultrasonic imaging diagnosis system based on a smart phone and a tablet computer to solve the problems raised in the above background art.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: A color ultrasonic imaging diagnosis system based on a smart phone and a tablet computer, comprising: a signal acquisition module, an adaptive signal enhancement module, and a signal detection module; The signal acquisition module is used for performing ultrasonic detection and receiving detection signals, and analyzing the received detection signals to dynamically adjust the acquisition frequency. The detection signals include: reflection signals and scattering signals; The adaptive signal enhancement module removes noise from the detection signals, then enhances the detection signals, and finally extracts the time-frequency characteristics of the enhanced signals, improves the gain of key frequency components, and removes noise frequency components to obtain detection enhanced signals; The signal detection module evaluates the quality of the detection enhanced signals according to the signal-to-noise ratio and mean square error of the detection enhanced signals. If the standard is met, it is transmitted to the image processor for imaging for diagnosis. If the standard is not met, it is retransmitted to the adaptive enhancement module to re-optimize the detection enhanced signals.

[0005] Preferably, the specific process of the signal acquisition module for performing ultrasonic detection is as follows: When performing medical diagnosis, connect the ultrasonic probe to the interface of a smartphone or tablet, and then drive the ultrasonic probe to work; mark the human tissue area for ultrasonic diagnosis as the diagnostic tissue area, preset the initial acquisition frequency, and the ultrasonic probe emits ultrasonic waves to the diagnostic tissue area according to the initial acquisition frequency and receives the detection signal of the diagnostic tissue area; After the ultrasonic detection of the diagnostic tissue area is completed, if the ultrasonic probe display shows high-frequency diagnostic information, re-perform ultrasonic detection on the diagnostic tissue area; If the ultrasonic probe display shows acquisition frequency matching information, send the detection signal of the diagnostic tissue area to the adaptive signal enhancement module.

[0006] Preferably, the specific process of the high-frequency diagnostic information or acquisition frequency matching information on the ultrasonic probe display is as follows: Evenly divide each diagnostic tissue area into several diagnostic areas, preset the diagnostic duration, obtain the reflection signal intensity at each acquisition moment within the diagnostic duration for each diagnostic area, and then sum the reflection signal intensities corresponding to all acquisition moments within the diagnostic duration of the diagnostic area and divide by the number of acquisitions to obtain the diagnostic area signal intensity. By summing the signal intensities corresponding to all diagnostic areas and dividing by the number of diagnostic areas, obtain the average diagnostic area signal intensity. According to the signal intensity corresponding to each diagnostic area and the average diagnostic area signal intensity, use the standard deviation calculation formula to calculate and obtain the signal fluctuation value of the diagnostic tissue area, denoted as the group domain fluctuation value ZY; Within the preset diagnostic duration, for each diagnostic area, collect the reflection signals of the diagnostic area at the initial acquisition frequency to obtain a series of discrete time-domain signal samples, denoted as Xn, where n = 1, 2,..., N; where N is the number of collected samples; perform a fast Fourier transform on the collected time-domain signal Xn to convert it into a frequency-domain signal Xk, k = 1, 2,..., N; the frequency-domain signal Xk is in complex form. For each frequency component k, its corresponding amplitude Ak is the modulus of Xk; use the formula: Fk = (Fs / N)k to obtain the actual frequency Fk corresponding to each frequency component k, where Fs is the initial acquisition frequency; use the formula: , to obtain the center frequency FZ; By summing the center frequencies of all diagnostic areas and dividing by the number of diagnostic areas, obtain the center average rate. According to the center average rate and the center frequency corresponding to each diagnostic area, use the standard deviation calculation formula to calculate and obtain the group center wave rate ZL; After normalizing the group domain fluctuation value ZY and the group center wave rate ZL, use the formula: CT = ZY × a1 + ZL × a2 to obtain the acquisition frequency adjustment value CT, where a1 and a2 are preset weight coefficients; Preset a frequency acquisition adjustment threshold. If the frequency acquisition adjustment value corresponding to the diagnosed tissue area is greater than the preset frequency acquisition adjustment threshold, generate high-frequency diagnostic information and send it to the ultrasonic probe display for display. At this time, the staff controls the ultrasonic probe to re-transmit ultrasonic waves to the diagnosed area according to the updated frequency, and sends the detection signal corresponding to the updated frequency in the diagnosed area to the adaptive signal enhancement module; If the frequency acquisition adjustment value corresponding to the diagnosed tissue area is less than or equal to the preset frequency acquisition adjustment threshold, generate frequency acquisition matching information and display it on the ultrasonic probe display.

[0007] Preferably, the method for obtaining the updated frequency is as follows: By taking the difference between the frequency acquisition adjustment value and the preset frequency acquisition adjustment threshold, obtain a frequency acquisition update value. Preset multiple frequency acquisition update value intervals, and each frequency acquisition update value interval corresponds to an updated frequency. By matching the frequency acquisition update value with multiple frequency acquisition update value intervals, output the updated frequency corresponding to the frequency acquisition update value interval; where each updated frequency is greater than the initial acquisition frequency.

[0008] Preferably, the specific process of the adaptive signal enhancement module for removing noise from the detection signal is as follows: Receive the detection signal from the signal acquisition module , n = 1, 2, ……, Ns; Ns is the number of samples. Perform a fast Fourier transform on the detection signal to convert the time-domain signal into a frequency-domain signal ; Obtain the sampling frequency Fc of the detection signal and the number of sampling points Z, where the sampling frequency is the acquisition frequency corresponding to the detection signal in the signal acquisition module; Use the formula PF = Fc / Z to obtain the frequency resolution PF; For the actual frequency Fi corresponding to each frequency component = PF × i, where i is the label of the frequency component, i = 1, 2, ……, Z; Analyze each frequency component in the frequency-domain signal one by one, record each frequency value and the corresponding amplitude Ai, and organize them into a frequency-amplitude pair list; Perform a square operation on the amplitude Ai of each frequency component to obtain the power spectrum Pi. Preset a noise interference power threshold Py, and traverse the power spectrum. For each frequency component, if Pi < Py, and the power values corresponding to the adjacent [X] frequency components are also less than the noise interference power threshold; where [X] is the preset number of consecutive frequency components to be detected; Mark the entire continuous frequency interval starting from the first frequency component that satisfies Pi less than the power threshold Py and the power values corresponding to its adjacent [X] frequency components are also less than Py, and ending at the frequency component that no longer satisfies this condition as the noise frequency interval; After identifying the noise frequency interval, select a band-stop filter to process the frequency-domain signal. Let the starting frequency component label of the noise frequency interval be , the end frequency component label is , and their corresponding actual frequencies are and ; The frequency response of the band-stop filter is defined as follows: , multiply the frequency-domain signal by the frequency response of the band-stop filter to obtain the filtered frequency-domain signal ; perform an inverse fast Fourier transform on the filtered frequency-domain signal to convert it back to the time-domain signal .

[0009] Preferably, the specific process of the adaptive signal enhancement module for enhancing the detection signal is as follows: Initialize the coefficient vector of the filter , and set its length to , and initialize each element to 0, that is: ; discretize the time-domain signal after band-stop filtering to obtain , where n = 1, 2,..., Ns; and use it as the input signal of the filter; at the same time, preset the desired response signal ; At the nth moment, the input signal vector of the filter , where M is the order of the filter or the length of the input signal vector, which determines the number of signal history values participating in the current operation; T is the transpose symbol, which converts the original row vector inside the brackets into a column vector; The output signal of the filter is calculated as: , calculate the error between the filter output signal and the desired response signal to obtain the error signal , that is: ; According to the least mean square criterion, use the error signal to adjust the coefficients of the filter, and the update formula is: , where is the preset step size factor, and the enhanced signal is obtained after adaptive filtering.

[0010] Preferably, the specific process of the adaptive signal enhancement module for performing time-frequency feature extraction on the enhanced signal, enhancing the gain of key frequency components, and removing noise frequency components to obtain the detection enhanced signal is as follows: Through the Hann window function , initially set the window function length L. According to the short-time Fourier transform formula: , perform a transformation on the signal, where m is the starting index of the time window and k is the frequency index. is the number of points for the Fourier transform. Preset the moving step size, move the time window according to the preset moving step size, and calculate a new value each time it moves, thereby obtaining a complete time-frequency distribution matrix. Calculate the energy distribution of the time-frequency domain signal Preset the energy threshold , and find the frequency interval that satisfies , then the frequency bandwidth ; ; Within each time window m, find the frequency index corresponding to the maximum energy, that is , and the corresponding peak frequency is: ; Traverse all frequency components. If it is found that the energy of some frequency components is weak but judged to be of great significance for diagnosis based on medical knowledge or experience, that is, the frequency components related to specific diseases, preset the gain factor , and use the formula: to obtain the adjusted time-frequency domain signal ; According to the energy distribution characteristics, for the frequency components with extremely low energy and clearly belonging to noise, set their corresponding values to 0 to further remove noise interference and improve the signal quality; use the inverse short-time Fourier transform to convert the optimized time-frequency domain signal back to the time domain to obtain the detection enhanced signal , and send the detection enhanced signal to the signal detection module.

[0011] Preferably, the signal detection module evaluates the quality of the detection enhanced signal based on the signal-to-noise ratio and mean square error of the detection enhanced signal. If it meets the standard, the specific process of transporting it to the image processor for imaging for diagnosis is as follows: Obtain the signal-to-noise ratio and mean square error corresponding to the detection enhanced signal . The specific process is as follows: Use the formula: to obtain the signal-to-noise ratio SNR; where is the preset ideal noise-free signal; use the formula: to obtain the mean square error MSE. Preset the signal-to-noise ratio threshold and the mean square error threshold, and compare the signal-to-noise ratio and the mean square error corresponding to the detected enhanced signal with the corresponding thresholds respectively. If the signal-to-noise ratio is greater than or equal to the corresponding threshold and the mean square error is less than or equal to the corresponding threshold, it is determined that the signal quality meets the standard. The detected enhanced signal is output and sent to the image processor of the smartphone or tablet for ultrasonic imaging. After imaging, professional doctors make a diagnosis based on medical knowledge and clinical experience.

[0012] Preferably, the processing method for the detected enhanced signal whose quality assessment fails to meet the standard is as follows: If the signal-to-noise ratio is less than the corresponding threshold, subtract the signal-to-noise ratio from the corresponding threshold and take the absolute value to obtain the signal-to-noise deviation value. Preset the step size increase coefficient, multiply the signal-to-noise deviation value by the preset step size increase coefficient to obtain the step size factor increase value; according to the step size factor increase value, increase the step size factor in the adaptive enhancement module, and send the detected enhanced signal back to the filtering and enhancement step in the adaptive enhancement module for re-filtering processing; If the mean square error is greater than the corresponding threshold, subtract the mean square error from the corresponding threshold to obtain the mean square deviation value. Preset the mean square reduction coefficient, multiply the mean square deviation value by the preset mean square reduction coefficient to obtain the mean square error reduction value. According to the mean square error reduction value, reduce the gain factor in the adaptive enhancement module, and send the detected enhanced signal back to the adaptive enhancement module again to recalculate the energy distribution, frequency bandwidth, and peak frequency characteristic value of the time-frequency domain signal, and re-evaluate and optimize the signal.

[0013] Compared with the prior art, the beneficial effects of the present invention are: (1) This color ultrasonic imaging diagnosis system based on a smartphone and a tablet can dynamically adjust the acquisition frequency according to the interface complexity of the diagnosed tissue area and the signal frequency characteristic differences through the signal acquisition module; evaluate the tissue condition by calculating the group domain fluctuation value and the group medium wave rate. Then, when the tissue interface is complex or the structure is uneven, the acquisition frequency is timely increased to ensure capturing more tissue detail information, effectively solving the problem that the traditional acquisition method cannot accurately match the tissue characteristics, improving the pertinence and effectiveness of signal acquisition, and laying a foundation for subsequent accurate diagnosis.

[0014] (2) This color ultrasonic imaging diagnosis system based on a smartphone and a tablet uses a variety of signal processing means through the adaptive signal enhancement module, such as using a band-stop filter to remove noise, enhancing the useful signal with the adaptive least mean square filtering algorithm, and extracting time-frequency characteristics and optimizing the signal through short-time Fourier transform; these operations effectively reduce the noise interference, highlight the key frequency components, improve the signal quality, and improve the poor signal processing effect of the traditional system, making the final imaging clearer and reducing the difficulty and misdiagnosis risk of doctor diagnosis.

[0015] (3) The color ultrasound imaging diagnosis system based on a smart phone and a tablet computer calculates the signal-to-noise ratio and mean square error of the detected enhanced signal through a signal detection module, compares them with a preset threshold value to evaluate the signal quality. If the signal quality meets the standard, it is directly transmitted to an image processor for imaging and diagnosis. If it does not meet the standard, the parameters of the adaptive enhancement module are adjusted according to different situations. For example, when the signal-to-noise ratio is low, the step factor is increased for re-filtering; when the mean square error is large, the gain factor is decreased for re-evaluating and optimizing the signal, ensuring the reliable signal quality entering the imaging link and avoiding misdiagnosis and missed diagnosis problems caused by poor signals, and greatly improving the accuracy and reliability of diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of the present invention. SPECIFIC EMBODIMENTS

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1 Please refer to Figure 1 , the present invention provides a color ultrasound imaging diagnosis system based on a smart phone and a tablet computer, including: a signal acquisition module, an adaptive signal enhancement module, and a signal detection module; The signal acquisition module is used to perform ultrasonic detection and receive the detection signal, and dynamically adjust the acquisition frequency by analyzing the received detection signal. The detection signal includes: a reflection signal and a scattering signal. The specific process is as follows: During medical diagnosis, connect the ultrasonic probe to the interface of the smart phone or tablet computer, and then drive the ultrasonic probe to work; mark the human tissue area for ultrasonic diagnosis as the diagnosis tissue area, preset the initial acquisition frequency, and the ultrasonic probe emits ultrasonic waves to the diagnosis tissue area according to the initial acquisition frequency and receives the detection signal of the diagnosis tissue area; After the ultrasonic detection of the diagnosis tissue area is completed, if the high-frequency diagnosis information is displayed on the ultrasonic probe display, the ultrasonic detection of the diagnosis tissue area is performed again; If the acquisition frequency matching information is displayed on the ultrasonic probe display, the detection signal of the diagnosis tissue area is sent to the adaptive signal enhancement module; The specific process of the high-frequency diagnosis information or acquisition frequency matching information on the ultrasonic probe display is as follows: For each piece of diagnostic tissue area, it is evenly divided into several diagnostic regions, and a preset diagnostic duration is set. For each diagnostic region, the reflection signal intensities at each acquisition moment within the diagnostic duration are obtained. Subsequently, the sum of the reflection signal intensities corresponding to all acquisition moments within the diagnostic duration of the diagnostic region is divided by the number of acquisitions to obtain the diagnostic region signal intensity. By summing up the signal intensities corresponding to all diagnostic regions and dividing by the number of diagnostic regions, the average signal intensity of the diagnostic region is obtained. According to the signal intensity corresponding to each diagnostic region and the average signal intensity of the diagnostic region, the standard deviation calculation formula is used for calculation to obtain the signal fluctuation value of the diagnostic tissue region, denoted as the group region fluctuation value ZY. The larger the group region fluctuation value, the more complex the regional interface of the diagnostic tissue, and the higher the acquisition frequency is required for acquisition to ensure that more tissue detail information can be captured; Within the preset diagnostic duration, for each diagnostic region, the reflection signal of the diagnostic region is acquired at the initial acquisition frequency to obtain a series of discrete time-domain signal samples, denoted as Xn, where n = 1, 2, ……, N; where N is the number of acquired samples; the acquired time-domain signal Xn is subjected to a fast Fourier transform to convert it into a frequency-domain signal Xk, k = 1, 2, ……, N; the frequency-domain signal Xk is in complex form. For each frequency component k, the corresponding amplitude Ak is the modulus of Xk; using the formula: Fk = (Fs / N)k, the actual frequency Fk corresponding to each frequency component k is obtained, where Fs is the initial acquisition frequency; using the formula: , the center frequency FZ is obtained; By summing up the center frequencies of all diagnostic regions and dividing by the number of diagnostic regions, the center average rate is obtained. According to the center average rate and the center frequency corresponding to each diagnostic region, the standard deviation calculation formula is used for calculation to obtain the group center wave rate ZL; the larger the group center wave rate, the greater the degree of dispersion of the center frequencies of each diagnostic region relative to the center average rate; this means that the reflection signal frequency characteristics of different diagnostic regions are more significantly different, which may reflect the uneven internal structure of the diagnostic tissue and requires a higher acquisition frequency for acquisition; After normalizing the group region fluctuation value ZY and the group center wave rate ZL, using the formula: CT = ZY × a1 + ZL × a2, the acquisition frequency adjustment value CT is obtained, where a1 and a2 are preset weight coefficients; A preset acquisition frequency adjustment threshold is set. If the acquisition frequency adjustment value corresponding to the diagnostic tissue region is greater than the preset acquisition frequency adjustment threshold, high-frequency diagnostic information is generated and sent to the ultrasonic probe display for display. At this time, the staff controls the ultrasonic probe to re-transmit ultrasonic waves to the diagnostic region according to the updated frequency, and sends the detection signal corresponding to the updated frequency in the diagnostic region to the adaptive signal enhancement module; If the acquisition frequency adjustment value corresponding to the diagnostic tissue region is less than or equal to the preset acquisition frequency adjustment threshold, acquisition frequency matching information is generated and displayed on the ultrasonic probe display; The acquisition method of the update frequency is as follows: By taking the difference between the frequency acquisition adjustment value and the preset frequency acquisition adjustment threshold, the frequency acquisition update value is obtained. Multiple frequency acquisition update value intervals are preset, and each frequency acquisition update value interval corresponds to an update frequency. By matching the frequency acquisition update value with multiple frequency acquisition update value intervals, the update frequency corresponding to the frequency acquisition update value interval is output; each update frequency is greater than the initial acquisition frequency; It should be noted that according to the actual situation of the diagnosed tissue area, such as the complexity of the interface and the difference in signal frequency characteristics, the acquisition frequency is dynamically adjusted; when the tissue interface is complex or the structure is uneven, the acquisition frequency can be increased in a timely manner to ensure that more tissue detail information is captured and the diagnostic accuracy is improved; through multi-dimensional analysis of the reflection signal intensity and frequency characteristics, including calculating the group domain fluctuation value, the center frequency, and the group median wave rate, etc., the condition of the diagnosed tissue area is comprehensively evaluated, providing a scientific basis for the adjustment of the acquisition frequency.

[0019] The adaptive signal enhancement module receives the detection signal of the signal acquisition module, removes the noise from the detection signal, then enhances the detection signal, and finally extracts the time-frequency characteristics of the enhanced signal, improves the gain of the key frequency components, removes the noise frequency components, obtains the detection enhanced signal, and sends the detection enhanced signal to the signal detection module. The specific process is as follows: Receive the detection signal from the signal acquisition module , n = 1, 2,..., Ns; Ns is the number of samples. Perform a fast Fourier transform on the detection signal to convert the time-domain signal into a frequency-domain signal ; Obtain the sampling frequency Fc of the detection signal and the number of sampling points Z, where the sampling frequency is the acquisition frequency corresponding to the detection signal in the signal acquisition module; use the formula PF = Fc / Z to obtain the frequency resolution PF; For the actual frequency Fi corresponding to each frequency component = PF × i, where i is the label of the frequency component, i = 1, 2,..., Z; Analyze each frequency component in the frequency-domain signal one by one, record each frequency value and the corresponding amplitude Ai, and organize them into a frequency-amplitude pair list; Perform a square operation on the amplitude Ai of each frequency component to obtain the power spectrum Pi, where the power spectrum intuitively reflects the energy distribution of the signal at the frequency; By observing the power spectrum, it is possible to clearly distinguish which frequency bands concentrate more energy and which frequency bands have less energy; Usually, the frequency bands with concentrated energy contain the key information of the signal, and this information is of great significance for identifying tissue characteristics and detecting lesions in ultrasonic diagnosis; while the frequency bands with lower energy may be mixed with noise or interference signals; Preset the noise interference power threshold \(P_y\), and traverse the power spectrum. For each frequency component, if \(P_i < P_y\) and the power values of the adjacent \([X]\) frequency components are also less than the noise interference power threshold, where \([X]\) is the detection number of preset continuous frequency components; Mark the entire continuous frequency interval starting from the first frequency component that satisfies \(P_i\) corresponding to the frequency component being less than the power threshold \(P_y\) and the power values of its adjacent \([X]\) frequency components also being less than \(P_y\), and ending at the frequency component that no longer satisfies this condition as the noise frequency interval; After determining the noise frequency interval, select a band-stop filter to process the frequency-domain signal. Let the starting frequency component label of the noise frequency interval be and the ending frequency component label be , and their corresponding actual frequencies be and ; The frequency response of the band-stop filter is defined as follows: , multiply the frequency-domain signal by the frequency response of the band-stop filter to obtain the filtered frequency-domain signal ; perform an inverse fast Fourier transform on the filtered frequency-domain signal to convert it back to the time-domain signal ; To further improve the signal quality, use the adaptive least mean square filtering algorithm to process the time-domain signal . The specific process is as follows: Initialize the coefficient vector of the filter, set its length to , and initialize each element to 0, that is: ; Discretize the time-domain signal after band-stop filtering to obtain , \(n = 1, 2, \cdots, N_s\); and use it as the input signal of the filter; At the same time, preset the desired response signal , where the desired response signal can be determined by taking the average value of multiple measurements. Assume that \(K\) measurements are made, and the signal obtained each time is , then ; where \(n\) is the index of the time series, representing discrete time points; \(n = 1, 2, \cdots, L\); \(L\) represents the total number of the time series; At the \(n\)th moment, the input signal vector of the filter, where \(M\) is the order of the filter or the length of the input signal vector, which determines the number of signal history values participating in the current operation; \(T\) is the transpose symbol, which converts the original row vector inside the brackets into a column vector; The output signal of the filter is: , calculate the error between the filter output signal and the desired response signal , that is: , namely: ; According to the least mean square criterion, use the error signal to adjust the coefficients of the filter, and the update formula is: , where is the preset step size factor, and the enhanced signal is obtained after adaptive filtering; the step size factor can be analyzed according to the maximum eigenvalue of the autocorrelation matrix of the input signal; It should be noted that when using the adaptive least mean square filtering algorithm, first initialize the coefficients of the filter, use the original signal as the input of the filter, calculate the error between the filter output signal and the desired response signal, and according to the least mean square criterion, use the error signal to adjust the coefficients of the filter, so that the output signal of the filter gradually approaches the desired response signal, thereby enhancing the useful signal and suppressing the interference signal; during the filtering process, monitor the change of the signal in real time and dynamically adjust the filter coefficients to adapt to different signal characteristics; Perform time-frequency feature extraction on the signal after adaptive filtering, and use the short-time Fourier transform to convert the signal into the time-frequency domain. The time-frequency features include: energy distribution, frequency bandwidth, and peak frequency; the specific process is: Through the Hanning window function , initially set the window function length L, and according to the short-time Fourier transform formula: , transform the signal, where m is the starting index of the time window, k is the frequency index, is the number of points of the Fourier transform, preset the moving step size, move the time window according to the preset moving step size, and calculate a new value each time, so as to obtain the complete time-frequency distribution matrix; Calculate the energy distribution of the time-frequency domain signal , where by analyzing the distribution of at different times and frequencies, the concentrated area of the signal energy on the time-frequency plane can be intuitively understood; for example, the area where the energy is concentrated may correspond to the main frequency components and active time periods of the signal; Preset the energy threshold , find the frequency interval that satisfies , then the frequency bandwidth , where the frequency bandwidth can reflect the richness of the signal frequency components. A wider bandwidth may indicate that the signal contains multiple frequency components and has a complex structure; Within each time window m, find the frequency index corresponding to the maximum energy , that is , and the corresponding peak frequency is: , where the peak frequency can be used to characterize the main frequency feature of the signal at that moment, and is of great significance for identifying the ultrasonic signal features of specific tissues; Traverse all frequency components. If it is found that the energy of some frequency components is weak but judged to be of great significance for diagnosis based on medical knowledge or experience, that is, the frequency components related to specific diseases, preset the gain factor , and use the formula: , to obtain the adjusted time-frequency domain signal ; According to the energy distribution characteristics, for the frequency components with extremely low energy and obviously belonging to noise, set their corresponding values to 0 to further remove noise interference and improve the signal quality; use the inverse short-time Fourier transform to convert the optimized time-frequency domain signal back to the time domain to obtain the detection enhanced signal , and send the detection enhanced signal to the signal detection module; It should be noted that the enhanced signal is converted to the time-frequency domain through the short-time Fourier transform, and the energy distribution, frequency bandwidth, and peak frequency of the signal are extracted; according to the extracted features, the signal is further optimized; that is: if it is found that the energy of some frequency components is weak but of great significance for diagnosis, the gain adjustment algorithm can be used to appropriately increase the gain of these frequency components to highlight these key features; for the frequency components with extremely low energy and obviously belonging to noise, set their corresponding values to 0 to further remove noise interference and improve the signal quality.

[0020] The signal detection module evaluates the quality of the detection enhanced signal according to the signal-to-noise ratio and mean square error of the detection enhanced signal. If it meets the standard, it is sent to the image processor for imaging for diagnosis. If it does not meet the standard, it is sent back to the adaptive enhancement module to re-optimize the detection enhanced signal. The specific process is as follows: Obtain the signal-to-noise ratio and mean square error corresponding to the detection enhanced signal . The specific process is as follows: Use the formula: , to obtain the signal-to-noise ratio SNR; where is the preset ideal noise-free signal; use the formula: , to obtain the mean square error MSE; Preset the signal-to-noise ratio threshold and the mean square error threshold, and compare the signal-to-noise ratio and the mean square error corresponding to the detected enhanced signal with the corresponding thresholds respectively. If the signal-to-noise ratio is greater than or equal to the corresponding threshold and the mean square error is less than or equal to the corresponding threshold, it is determined that the signal quality meets the standard, and the detected enhanced signal is output and sent to the image processor of the smartphone or tablet for ultrasonic imaging. After imaging, a professional doctor makes a diagnosis based on medical knowledge and clinical experience. The processing method for the signal quality not meeting the standard is as follows: If the signal-to-noise ratio is less than the corresponding threshold, subtract the signal-to-noise ratio from the corresponding threshold and take the absolute value to obtain the signal-to-noise deviation value. Preset the step size increase coefficient, multiply the signal-to-noise deviation value by the preset step size increase coefficient to obtain the step size factor increase value. According to the step size factor increase value, increase the step size factor in the adaptive enhancement module, and send the detected enhanced signal back to the filtering and enhancement step in the adaptive enhancement module for re-filtering processing. If the mean square error is greater than the corresponding threshold, subtract the mean square error from the corresponding threshold to obtain the mean square deviation value. Preset the mean square reduction coefficient, multiply the mean square deviation value by the preset mean square reduction coefficient to obtain the mean square error reduction value. According to the mean square error reduction value, reduce the gain factor in the adaptive enhancement module, and send the detected enhanced signal back to the adaptive enhancement module again to recalculate the energy distribution, frequency bandwidth, and peak frequency eigenvalue of the time-frequency domain signal, and re-evaluate and optimize the signal.

[0021] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A color ultrasonic imaging diagnosis system based on a smart phone and a tablet, comprising: A signal acquisition module, an adaptive signal enhancement module, and a signal detection module, characterized in that: The signal acquisition module is used to perform ultrasonic detection and receive detection signals, and analyze the received detection signals to dynamically adjust the acquisition frequency. The detection signals include: reflected signals and scattered signals; The adaptive signal enhancement module removes noise from the detection signals, then enhances the detection signals, and finally extracts the time-frequency features of the enhanced signals, boosts the gain of key frequency components, and removes noise frequency components to obtain detection enhanced signals; The signal detection module evaluates the quality of the detection enhanced signals based on the signal-to-noise ratio and mean square error of the detection enhanced signals. If the standard is met, it is transmitted to the image processor for imaging for diagnosis. If not, it is resent to the adaptive enhancement module to re-optimize the detection enhanced signals.

2. The color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 1, wherein: The specific process of the signal acquisition module performing ultrasonic detection is as follows: During medical diagnosis, connect the ultrasonic probe to the interface of a smartphone or tablet, and then drive the ultrasonic probe to work; mark the human tissue area for ultrasonic diagnosis as the diagnostic tissue area, preset the initial acquisition frequency, and the ultrasonic probe emits ultrasonic waves to the diagnostic tissue area according to the initial acquisition frequency and receives the detection signals of the diagnostic tissue area; After the ultrasonic detection of the diagnostic tissue area is completed, if the ultrasonic probe display shows high-frequency diagnostic information, re-perform ultrasonic detection on the diagnostic tissue area; If the ultrasonic probe display shows acquisition frequency matching information, send the detection signals of the diagnostic tissue area to the adaptive signal enhancement module.

3. A color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 2, characterized in that: The specific process of the high-frequency diagnostic information or acquisition frequency matching information on the ultrasonic probe display is as follows: Divide the diagnostic tissue area into multiple diagnostic areas, calculate the standard deviation of the signal intensity in each area within the preset time period, i.e., the group domain fluctuation value ZY, and the standard deviation of the central frequency, i.e., the group mid-wave rate ZL; After normalizing the group domain fluctuation value ZY and the group mid-wave rate ZL, use the formula: CT = ZY × a1 + ZL × a2 to obtain the acquisition frequency adjustment value CT, where a1 and a2 are preset weight coefficients; If the acquisition frequency adjustment value CT exceeds the preset threshold, display high-frequency diagnostic information and update the acquisition frequency; otherwise, display acquisition frequency matching information and transmit the signal to the adaptive enhancement module.

4. A color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 3, characterized in that: The method for obtaining the updated frequency is as follows: By taking the difference between the acquisition frequency adjustment value and the preset acquisition frequency adjustment threshold, obtain the acquisition frequency update value. Preset multiple acquisition frequency update value intervals, and each acquisition frequency update value interval corresponds to an update frequency. By matching the acquisition frequency update value with multiple acquisition frequency update value intervals, output the update frequency corresponding to the acquisition frequency update value interval; each update frequency is greater than the initial acquisition frequency.

5. The color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 4, wherein: The specific process of the adaptive signal enhancement module removing noise from the detection signals is as follows: Perform a fast Fourier transform on the detection signals and calculate the power spectrum; Traverse the power spectrum and mark consecutive [X] frequency intervals below the noise power threshold as noise frequency intervals; Use a band-stop filter to filter out the noise frequency intervals and convert back to the time-domain signal through inverse Fourier transform.

6. The color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 5, wherein: The specific process of the adaptive signal enhancement module enhancing the detection signals is as follows: Initialize the coefficient vector of the filter , whose length is set to , and initialize each element to 0, that is: ; Discretize the time-domain signal after band-stop filtering to obtain , where n = 1, 2,..., Ns; and use it as the input signal of the filter; At the same time, preset the desired response signal ; At the nth moment, the input signal vector of the filter , where M is the order of the filter or the length of the input signal vector, which determines the number of signal history values involved in the current operation; T is the transpose symbol that converts the original row vector within the brackets into a column vector; Output signal of the filter The calculation method is as follows: , calculate the output signal of the filter and the desired response signal to obtain the error between them, resulting in an error signal , that is: ; According to the least mean square criterion, the error signal is used to adjust the coefficients of the filter, and the update formula is: , where is a preset step size factor. After adaptive filtering, the enhanced signal is obtained.

7. The color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 6, wherein: The specific process of the adaptive signal enhancement module for extracting the time-frequency features of the enhanced signal, enhancing the gain of key frequency components, and removing noise frequency components to obtain the detected enhanced signal is as follows: Through the Hanning window function , initially set the window function length L. According to the short-time Fourier transform formula: , transform the signal, where m is the starting index of the time window, k is the frequency index, is the number of Fourier transform points, preset the moving step size, move the time window according to the preset moving step size, and calculate a new value each time it moves, so as to obtain the complete time-frequency distribution matrix; Calculate the energy distribution of the time-frequency domain signal Find the preset energy threshold Find the frequency interval that satisfies Then the frequency bandwidth is ;​​ Within each time window m, find the frequency index corresponding to the maximum energy , that is , and the corresponding peak frequency is: ; Traverse all frequency components. If it is found that the energy of some frequency components is weak but judged to be of great significance for diagnosis based on medical knowledge or experience, that is, the frequency components related to specific diseases, preset the gain factor , and use the formula: to obtain the adjusted time-frequency domain signal ; According to the energy distribution characteristics, for the frequency components with extremely low energy and obviously belonging to noise, the corresponding value is set to 0 to further remove noise interference and improve the signal quality; the optimized time-frequency domain signal is converted back to the time domain by using the inverse short-time Fourier transform to obtain the detection enhanced signal , and the detection enhanced signal is sent to the signal detection module.

8. The color ultrasonic imaging diagnosis system based on a smart phone and a tablet according to claim 7, characterized in that: The signal detection module evaluates the quality of the detected enhanced signal based on the signal-to-noise ratio and mean square error of the detected enhanced signal. If the standard is met, the specific process of transmitting it to the image processor for imaging for diagnosis is as follows: Obtain the detection enhanced signal The corresponding signal-to-noise ratio and mean square error. The specific process is as follows: Using the formula: , the signal-to-noise ratio SNR is obtained; where is a preset ideal noise-free signal; Using the formula: , the mean square error MSE is obtained; Preset the signal-to-noise ratio threshold and mean square error threshold, compare the signal-to-noise ratio and mean square error corresponding to the detected enhanced signal with the corresponding thresholds respectively. If the signal-to-noise ratio is greater than or equal to the corresponding threshold and the mean square error is less than or equal to the corresponding threshold, it is determined that the signal quality meets the standard, the detected enhanced signal is output, and sent to the image processor of the smartphone or tablet for ultrasonic imaging. After imaging, professional doctors make a diagnosis based on medical knowledge and clinical experience.

9. A color ultrasound imaging diagnosis system based on a smart phone and a tablet according to claim 8, characterized in that: The processing method for the detected enhanced signal whose quality assessment fails to meet the standard is as follows: If the signal-to-noise ratio is less than the corresponding threshold, subtract the signal-to-noise ratio from the corresponding threshold and take the absolute value to obtain the signal-to-noise deviation value. Preset the step size increase coefficient, multiply the signal-to-noise deviation value by the preset step size increase coefficient to obtain the step size factor increase value; according to the step size factor increase value, increase the step size factor in the adaptive enhancement module, and send the detected enhanced signal back to the filtering enhancement step in the adaptive enhancement module for re-filtering processing; If the mean square error is greater than the corresponding threshold, subtract the mean square error from the corresponding threshold to obtain the mean square deviation value. Preset the mean square reduction coefficient, multiply the mean square deviation value by the preset mean square reduction coefficient to obtain the mean square error reduction value. According to the mean square error reduction value, reduce the gain factor in the adaptive enhancement module, and send the detected enhanced signal back to the adaptive enhancement module again to recalculate the energy distribution, frequency bandwidth, and peak frequency eigenvalue of the time-frequency domain signal, and re-evaluate and optimize the signal.

Images