Portable intelligent non-alcoholic fatty liver detection method and system

Abstract

The invention relates to the field of data processing, and provides a portable intelligent non-alcoholic fatty liver detection method and system, an attenuation coefficient and a backscattering coefficient are optimized and extracted through a linear least square method and a one-dimensional convolutional neural network, the detection accuracy is further improved, quantitative relation analysis is performed through a nonlinear regression model, and the detection accuracy is further improved. The ultrasonic fat fraction is quantitatively analyzed based on an energy absorption method, energy needed by detection is saved, the requirement for portable detection is met, different ultrasonic coefficients are combined, the analysis accuracy is improved, a liver comprehensive verification model is designed for simulation verification processing, intelligent verification and intelligent evaluation are achieved, and the detection efficiency is improved. According to the non-alcoholic fatty liver disease detection method, the dependence on professional knowledge is avoided, the detection result is more in line with the actual situation, the optimized result is visually displayed, the detection result is more intuitive and clearer, the detection convenience is improved, and the accuracy and convenience of the non-alcoholic fatty liver disease detection method are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and particularly to a portable intelligent non-alcoholic fatty liver detection method and system. Background Art

[0002] As one of the most common liver diseases, non-alcoholic fatty liver is very important for early detection and quantitative evaluation. However, the current detection of non-alcoholic fatty liver often relies on expensive and complex imaging devices, such as large-scale nuclear magnetic resonance devices and ultrasonic devices. These expensive and complex imaging devices require professional personnel to operate, resulting in obvious limitations in the regional popularity and detection convenience of non-alcoholic fatty liver detection. The portability of non-alcoholic fatty liver detection has become an important part.

[0003] In the prior art, the detection of non-alcoholic fatty liver usually obtains ultrasonic data based on ultrasonic detection for quantitative analysis. However, the existing ultrasonic detection methods often obtain ultrasonic imaging data according to the shear stress method, resulting in relatively high energy required for the overall detection and insufficient clarity of ultrasonic imaging. For portable devices, it is difficult to maintain long-term use. Moreover, for the operators of portable devices, the lack of intelligence makes it impossible to intuitively and clearly obtain the detection results and change trends, so that the detection still depends on the professional knowledge of the operators, affecting the accuracy of portable detection.

[0004] Therefore, how to design a non-alcoholic fatty liver detection method that can meet the requirements of portable detection while improving the accuracy and intelligence of detection has become an urgent problem to be solved. Summary of the Invention

[0005] Based on this, a portable intelligent non-alcoholic fatty liver detection method and system proposed by the present invention optimally extracts attenuation coefficients and backscattering coefficients through linear least squares method and one-dimensional convolutional neural network to obtain more accurate liver fat state information, further improving the accuracy of detection. Then, a nonlinear regression model is designed for quantitative relationship analysis, and the ultrasonic fat fraction is quantitatively analyzed based on the energy absorption method, saving the energy required for detection, meeting the requirements of portable detection, combining different ultrasonic coefficients and improving the accuracy of analysis. Also, a liver comprehensive verification model is designed for simulation verification processing, realizing the intelligent verification and intelligent evaluation of the analysis results, avoiding the dependence on professional techniques and professional knowledge, making the detection results more in line with the actual situation, and performing visual display processing on the optimized results, making the detection results more intuitive and clear, and improving the convenience of detection. The present invention improves the accuracy and convenience of the non-alcoholic fatty liver detection method.

[0006] A portable intelligent non-alcoholic fatty liver detection method proposed by the present invention includes: Obtaining basic ultrasonic signal parameters and extracting an ultrasonic comprehensive coefficient, where the ultrasonic comprehensive coefficient includes an attenuation coefficient and a backscattering coefficient, and the extraction of the ultrasonic comprehensive coefficient is based on the linear least squares method and a one-dimensional convolutional neural network; Performing quantitative relationship analysis based on the ultrasonic comprehensive coefficient to obtain an ultrasonic fat fraction, where the quantitative relationship analysis is based on the energy absorption method, and the ultrasonic fat fraction is calculated based on a non-linear regression model; Performing B-mode ultrasonic imaging processing according to the field programmable gate array architecture to obtain a target basic ultrasonic image; Performing simulation verification processing according to a liver comprehensive verification model to obtain a target optimized ultrasonic image, where the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model; Generating a visual evaluation report based on the ultrasonic comprehensive coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image, where the visual evaluation report is based on an image segmentation model, and the image segmentation model includes morphological correction optimization and connected domain analysis optimization.

[0007] In summary, according to the above-mentioned portable intelligent non-alcoholic fatty liver detection method, through linear least squares method and one-dimensional convolutional neural network, the attenuation coefficient and backscattering coefficient are optimally extracted to obtain more accurate liver fat status information, further improving the accuracy of detection. Then, by designing a non-linear regression model for quantitative relationship analysis and quantitatively analyzing the ultrasonic fat fraction based on the energy absorption method, the energy required for detection is saved, meeting the requirements of portable detection. Also, different ultrasonic coefficients are combined to improve the accuracy of analysis. Moreover, a liver comprehensive verification model is designed for simulation verification processing, realizing the intelligent verification and intelligent evaluation of the analysis results, avoiding the dependence on professional technologies and knowledge, making the detection results more in line with the actual situation. The optimized results are visually displayed, making the detection results more intuitive and clear, and improving the convenience of detection. The present invention improves the accuracy and convenience of the non-alcoholic fatty liver detection method. Specifically, basic ultrasonic signal parameters are obtained and the ultrasonic comprehensive coefficient is extracted. The ultrasonic comprehensive coefficient includes the attenuation coefficient and the backscattering coefficient. The extraction of the ultrasonic comprehensive coefficient is based on the linear least squares method and one-dimensional convolutional neural network to obtain more accurate liver fat status information and further improve the accuracy of detection. Quantitative relationship analysis is performed according to the ultrasonic comprehensive coefficient to obtain the ultrasonic fat fraction. The quantitative relationship analysis is based on the energy absorption method. The ultrasonic fat fraction is calculated based on a non-linear regression model. The ultrasonic fat fraction is quantitatively analyzed based on the energy absorption method, saving the energy required for detection and meeting the requirements of portable detection. Also, different ultrasonic coefficients are combined to improve the accuracy of analysis. B-mode ultrasonic imaging processing is performed according to the field programmable gate array architecture to obtain the target basic ultrasonic image. Simulation verification processing is performed according to the liver comprehensive verification model to obtain the target optimized ultrasonic image. The liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model, realizing the intelligent verification and intelligent evaluation of the analysis results, avoiding the dependence on professional technologies and knowledge, making the detection results more in line with the actual situation. A visual evaluation report is generated according to the ultrasonic comprehensive coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image. The visual evaluation report is based on an image segmentation model. The image segmentation model includes morphological correction optimization and connected domain analysis optimization, making the detection results more intuitive and clear and improving the convenience of detection. The present invention improves the accuracy and convenience of the non-alcoholic fatty liver detection method.

[0008] Further, the step of obtaining the basic ultrasonic signal parameters and extracting the ultrasonic comprehensive coefficient specifically includes: Obtain the basic ultrasonic signal parameters, and perform data interception on the depth range in the basic ultrasonic signal parameters to determine the depth axis region; Analyze the basic ultrasonic signal parameters according to a sliding window to obtain envelope signal parameters, adjust the sliding window according to the target dynamics, solve the envelope signal parameters in the sliding window of the depth axis region, determine the depth value at the center of the sliding window of the depth axis region, and obtain a two-dimensional matrix according to the depth value and the frequency value of the basic ultrasonic signal parameters; Estimate the attenuation coefficient and the backscattering coefficient for the depth value and the frequency value according to the weak scattering Born approximation, and then solve and extract according to the linear least squares method to obtain the attenuation coefficient and the backscattering coefficient; Input the attenuation coefficient and the backscattering coefficient into a one-dimensional convolutional neural network for coefficient enhancement extraction.

[0009] Further, the step of inputting the attenuation coefficient and the backscattering coefficient into a one-dimensional convolutional neural network for coefficient enhancement extraction specifically includes: Perform a time axis-depth axis conversion on the basic ultrasonic signal parameters to intercept the time axis information of the RF signal segment in the basic ultrasonic signal parameters, and perform region of interest tracking adjustment according to the cross-correlation algorithm to dynamically adjust the sliding window of the time axis region; Input the attenuation coefficient and the backscattering coefficient into a one-dimensional convolutional neural network, and the one-dimensional convolutional neural network is constructed based on the historical relevant detection parameters of the detection target; Extract basic ultrasonic features according to depthwise separable convolution, divide the basic ultrasonic features into multiple feature groups of different scales according to the feature scale, perform channel attention enhancement processing on each feature group to obtain multiple attention enhancement feature groups of different scales, and fuse the attention enhancement feature groups step by step according to the order of feature scales from low to high to obtain multi-scale ultrasonic features; Input the multi-scale ultrasonic features into the attenuation coefficient branch and the backscattering coefficient branch respectively for coefficient extraction to obtain the enhanced attenuation coefficient and the enhanced backscattering coefficient respectively.

[0010] Further, the step of performing quantitative relationship analysis on the ultrasonic comprehensive coefficient to obtain the ultrasonic fat fraction specifically includes: Perform Savitzky-Golay smoothing processing on the ultrasonic comprehensive coefficient to eliminate respiratory motion artifacts; Perform depth weighting correction on the ultrasonic comprehensive coefficient according to the sound field depth distribution in the region of interest in the basic ultrasonic signal parameters; Input the original value and the interaction term of the ultrasonic comprehensive coefficient after depth weighting correction into a nonlinear regression model to obtain the ultrasonic fat fraction, and the specific algorithm of the nonlinear regression model is as follows: , , Among them, UDFF represents the original ultrasonic fat fraction, , , represent fitting coefficients, AC represents the attenuation coefficient in the ultrasonic comprehensive coefficient, BSC represents the backscattering coefficient in the ultrasonic comprehensive coefficient, (AC·BSC) represents the interaction term of the ultrasonic comprehensive coefficient, c represents the constant term, represents the ultrasonic fat fraction after standardized mapping, represents the mapping coefficient, represents the mapping constant.

[0011] Further, the step of performing simulation verification processing according to the liver comprehensive verification model to obtain the target optimized ultrasonic image specifically includes: Performing simulation verification processing according to the liver comprehensive verification model, and the liver comprehensive verification model includes a basic simulation model and a multi - coefficient verification model; The basic simulation model includes a basic scatterer model and a global attenuation constraint. The basic scatterer model includes regular scatterers and diffuse scatterers. The distribution of the regular scatterers is based on the chi - square distribution rule, the amplitude of the regular scatterers is based on the uniform distribution rule, the distribution and amplitude of the diffuse scatterers are both based on the uniform distribution rule. The regular scatterers are used to adjust the isotropy of the spatial distribution of biological soft tissue scattering points to map the one - dimensional scattering point distribution into three - dimensional space, and the distribution mapping is based on the Hilbert space - filling curve; The multi - coefficient verification model includes a basic scatterer model, a global attenuation constraint, and a backscattering constraint. The basic scatterer model of the multi - coefficient verification model is an elastic sphere model. The elastic sphere model adjusts the size and distribution of the scatterers based on the intervention theory model and calculates the scattering intensity of a single elastic sphere model. The number density of the basic scatterer model of the multi - coefficient verification model is calculated based on the liver biological characteristic information. The scattering intensity is linearly superimposed based on the scattering intensities of all elastic sphere models, and the scattering intensity is convolved with the spatial pulse of the ultrasonic transducer to obtain the ultrasonic radio frequency echo signal; Performing simulation verification optimization according to the ultrasonic radio frequency echo signal to obtain the target optimized ultrasonic image.

[0012] Further, the step of generating a visualization evaluation report according to the ultrasonic comprehensive coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image specifically includes: Pre - processing the target optimized ultrasonic image, and the pre - processing includes speckle noise suppression processing and contrast enhancement processing. The speckle noise suppression processing is based on non - local mean filtering, and the contrast enhancement processing is based on the locally contrast - limited adaptive histogram equalization algorithm; Segment the preprocessed target optimized ultrasound image according to the image segmentation model to obtain a liver region mask, and then perform interference structure removal processing on the liver region mask, where the interference structures include blood vessel structures and mass structures; Perform post-processing on the liver region mask, where the post-processing includes morphological correction optimization and connected component analysis optimization. Perform closing operation according to the morphological correction optimization algorithm to fill the hole regions, and then perform maximum region retention according to the connected component analysis optimization to obtain an optimized segmentation result of the liver region; Perform heat map color mapping on the optimized segmentation result of the liver region according to the ultrasound fat fraction and the ultrasound comprehensive coefficient, and then convert the optimized segmentation result of the liver region into a color layer through transparency fusion to be superimposed on the target optimized ultrasound image to obtain a visualization display result; Perform parameter and evaluation annotation on the visualization display result according to the ultrasound fat fraction and the ultrasound comprehensive coefficient to obtain a visualization evaluation report.

[0013] A portable intelligent non-alcoholic fatty liver detection system proposed by the present invention includes: A coefficient extraction module for obtaining basic ultrasound signal parameters and extracting an ultrasound comprehensive coefficient, where the ultrasound comprehensive coefficient includes an attenuation coefficient and a backscattering coefficient, and the extraction of the ultrasound comprehensive coefficient is based on the linear least squares method and a one-dimensional convolutional neural network; A quantitative relationship analysis module for performing quantitative relationship analysis according to the ultrasound comprehensive coefficient to obtain an ultrasound fat fraction, where the quantitative relationship analysis is based on the energy absorption method, and the ultrasound fat fraction is calculated based on a non-linear regression model; An ultrasound imaging module for performing B-mode ultrasound imaging processing according to the field programmable gate array architecture to obtain a target basic ultrasound image; A simulation verification module for performing simulation verification processing according to a liver comprehensive verification model to obtain a target optimized ultrasound image, where the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model; A visualization display module for generating a visualization evaluation report according to the ultrasound comprehensive coefficient, the ultrasound fat fraction, and the target optimized ultrasound image, where the visualization evaluation report is based on an image segmentation model, and the image segmentation model includes morphological correction optimization and connected component analysis optimization.

[0014] Further, the ultrasound imaging module specifically includes: an ultrasound transducer, a front-end signal transceiver processor, and a back-end ultrasound image processor; The front - end signal transceiver processor includes a transmitting front - end unit, a receiving front - end unit, a beam processing unit, and a signal processing unit. The transmitting front - end unit is used to generate high - voltage excitation pulses, focus digital electronics, adjust beam scanning, and control aperture size. The receiving front - end unit is used to amplify echo signals at the front - end, perform analog - to - digital conversion, and high - speed transmission. The beam processing unit is used for digital signal processing, variable - aperture reception, dynamic apodization, and beam synthesis. The signal processing unit is used for gain control, dynamic filtering, envelope detection, secondary sampling, logarithmic compression, coordinate transformation processing, and transmits the finally generated signal to the back - end ultrasonic image processor.

[0015] The present invention also provides a storage medium that stores one or more programs. When the program is executed by a processor, it implements the portable intelligent non - alcoholic fatty liver detection method as described above.

[0016] The present invention also provides a computer device, which includes a memory and a processor, wherein: The memory is used to store computer programs; When the processor executes the computer program stored in the memory, it implements the portable intelligent non - alcoholic fatty liver detection method as described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of the portable intelligent non - alcoholic fatty liver detection method proposed in the first embodiment of the present invention; Figure 2 It is a schematic structural diagram of the portable intelligent non - alcoholic fatty liver detection system proposed in the second embodiment of the present invention.

[0018] The following specific embodiments will further illustrate the present invention in conjunction with the above - mentioned drawings. SPECIFIC EMBODIMENTS

[0019] For ease of understanding of the present invention, the present invention will be described more comprehensively with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0020] It should be noted that when an element is referred to as being "fixedly installed on" another element, it can be directly on the other element or there may also be an intermediate element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be an intermediate element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are only for the purpose of illustration.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs. The terms used in the specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0022] Please refer to Figure 1 , which shows the flowchart of the portable intelligent non-alcoholic fatty liver detection method proposed in the first embodiment of the present invention. This portable intelligent non-alcoholic fatty liver detection method includes steps S01 to S05, where: Step S01: Obtain basic ultrasonic signal parameters and extract ultrasonic comprehensive coefficients; It should be noted that in this embodiment, the ultrasonic comprehensive coefficients include attenuation coefficients and backscattering coefficients. The extraction of the ultrasonic comprehensive coefficients is based on the linear least squares method and a one-dimensional convolutional neural network. The basic ultrasonic signal parameters are obtained, and the data in the depth range of the basic ultrasonic signal parameters is intercepted to determine the depth axis region; Analyze the basic ultrasonic signal parameters according to a sliding window to obtain envelope signal parameters. Adjust the sliding window according to the target dynamics, solve the envelope signal parameters in the sliding window of the depth axis region. The center of the sliding window in the depth axis region determines the depth value, and a two-dimensional matrix is obtained according to the depth value and the frequency value of the basic ultrasonic signal parameters; Estimate the attenuation coefficient and backscattering coefficient according to the weak scattering Born approximation for the depth value and the frequency value, and then solve and extract according to the linear least squares method to obtain the attenuation coefficient and backscattering coefficient; The principle formula for estimating the attenuation coefficient and backscattering coefficient is as follows: , where represents the weak scattering Born approximation, represents the combined effect value of the electrical excitation and the transducer, represents the diffraction effect value, represents the cumulative attenuation coefficient, represents the backscattering coefficient, represents the frequency value, represents the depth value; Input the attenuation coefficient and backscattering coefficient into a one-dimensional convolutional neural network for coefficient enhancement extraction; Perform a time axis-depth axis conversion on the basic ultrasonic signal parameters to intercept the time axis information of the RF signal segment in the basic ultrasonic signal parameters, and perform region of interest tracking adjustment according to the cross-correlation algorithm to dynamically adjust the sliding window of the time axis region; Input the attenuation coefficient and the backscattering coefficient into a one-dimensional convolutional neural network, which is constructed based on the historical relevant detection parameters of the detection target; Extract basic ultrasonic features according to depthwise separable convolution, divide the basic ultrasonic features into multiple feature groups with different scales according to the feature scale, perform channel attention enhancement processing on each of the feature groups to obtain multiple attention-enhanced feature groups with different scales, and fuse the attention-enhanced feature groups step by step according to the order of the feature scales from low to high to obtain multi-scale ultrasonic features; Input the multi-scale ultrasonic features into an attenuation coefficient branch and a backscattering coefficient branch respectively for coefficient extraction to obtain an enhanced attenuation coefficient and an enhanced backscattering coefficient respectively.

[0023] Step S02: Perform quantitative relationship analysis according to the ultrasonic comprehensive coefficient to obtain the ultrasonic fat fraction; It should be noted that in this embodiment, the quantitative relationship analysis is based on the energy absorption method, the ultrasonic fat fraction is calculated based on a non-linear regression model, and Savitzky-Golay smoothing processing is performed on the ultrasonic comprehensive coefficient to eliminate respiratory motion artifacts; Perform depth weighting correction on the ultrasonic comprehensive coefficient according to the sound field depth distribution in the region of interest in the basic ultrasonic signal parameters; Input the original value and the interaction term of the depth-weighting corrected ultrasonic comprehensive coefficient into the non-linear regression model to obtain the ultrasonic fat fraction. The specific algorithm of the non-linear regression model is as follows: , , where UDFF represents the original ultrasonic fat fraction, , , represent fitting coefficients, AC represents the attenuation coefficient in the ultrasonic comprehensive coefficient, BSC represents the backscattering coefficient in the ultrasonic comprehensive coefficient, (AC·BSC) represents the interaction term of the ultrasonic comprehensive coefficient, c represents the constant term, represents the ultrasonic fat fraction after standardized mapping, represents the mapping coefficient, represents the mapping constant.

[0024] Step S03: Perform B-mode ultrasonic imaging processing according to the field-programmable gate array architecture to obtain the target basic ultrasonic image; It should be noted that in this embodiment, the B-mode ultrasonic imaging processing is based on an ultrasonic imaging module, and the ultrasonic imaging module specifically includes: an ultrasonic transducer, a front-end signal transceiver processor, and a back-end ultrasonic image processor; The front-end signal transceiver processor includes a transmitting front-end unit, a receiving front-end unit, a beam processing unit, and a signal processing unit. The transmitting front-end unit is used to generate high-voltage excitation pulses, focus digital electronics, adjust beam scanning, and control aperture size. The receiving front-end unit is used to amplify echo signals at the front end, perform analog-to-digital conversion, and high-speed transmission. The beam processing unit is used for digital signal processing, variable-aperture reception, dynamic apodization, and beam synthesis. The signal processing unit is used for gain control, dynamic filtering, envelope detection, secondary sampling, logarithmic compression, coordinate transformation processing, and transmits the finally generated signal to the back-end ultrasonic image processor.

[0025] Step S04: Perform simulation verification processing according to the liver comprehensive verification model to obtain the target optimized ultrasonic image; It should be noted that in this embodiment, the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model. When performing simulation verification processing according to the liver comprehensive verification model, the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model; The basic simulation model includes a basic scatterer model and a global attenuation constraint. The basic scatterer model includes regular scatterers and diffuse scatterers. The distribution of the regular scatterers is based on the chi-square distribution rule, the amplitude of the regular scatterers is based on the uniform distribution rule, the distribution and amplitude of the diffuse scatterers are both based on the uniform distribution rule. The regular scatterers are used to adjust the isotropy of the spatial distribution of biological soft tissue scatter points to map the one-dimensional scatter point distribution to three-dimensional space, and the distribution mapping is based on the Hilbert space-filling curve; The multi-coefficient verification model includes a basic scatterer model, a global attenuation constraint, and a backscattering constraint. The basic scatterer model of the multi-coefficient verification model is an elastic sphere model. The elastic sphere model adjusts the size and distribution of scatterers based on the intervention theory model and calculates the scattering intensity of a single elastic sphere model. The number density of the basic scatterer model of the multi-coefficient verification model is calculated based on the liver biological characteristic information. The scattering intensity is linearly superimposed based on the scattering intensities of all elastic sphere models, and the scattering intensity is convolved with the spatial pulse of the ultrasonic transducer to obtain the ultrasonic radio frequency echo signal; Perform simulation verification optimization according to the ultrasonic radio frequency echo signal to obtain the target optimized ultrasonic image.

[0026] Step S05: Generate a visualization evaluation report according to the ultrasonic comprehensive coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image; It should be noted that in this embodiment, the visual evaluation report is based on an image segmentation model. The image segmentation model includes morphological correction optimization and connected component analysis optimization, and preprocesses the target optimized ultrasound image. The preprocessing includes speckle noise suppression processing and contrast enhancement processing. The speckle noise suppression processing is based on non-local mean filtering, and the contrast enhancement processing is based on an adaptive histogram equalization algorithm with local contrast limitation; Perform segmentation processing on the preprocessed target optimized ultrasound image according to the image segmentation model to obtain a liver region mask, and then perform interference structure removal processing on the liver region mask. The interference structures include blood vessel structures and mass structures; Perform post-processing on the liver region mask. The post-processing includes morphological correction optimization and connected component analysis optimization. Perform closing operation according to the morphological correction optimization algorithm to fill the hole regions, and then perform maximum region retention according to the connected component analysis optimization to obtain an optimized segmentation result of the liver region; Perform heat map color mapping on the optimized segmentation result of the liver region according to the ultrasound fat fraction and the ultrasound comprehensive coefficient, and then convert the optimized segmentation result of the liver region into a color layer through transparency fusion to be superimposed on the target optimized ultrasound image to obtain a visual display result; Perform parameter and evaluation annotation on the visual display result according to the ultrasound fat fraction and the ultrasound comprehensive coefficient to obtain a visual evaluation report.

[0027] In summary, according to the above-mentioned portable intelligent non-alcoholic fatty liver detection method, through linear least squares method and one-dimensional convolutional neural network, the attenuation coefficient and backscattering coefficient are optimally extracted to obtain more accurate liver fat status information, further improving the accuracy of detection. Then, by designing a non-linear regression model for quantitative relationship analysis and quantitatively analyzing the ultrasonic fat fraction based on the energy absorption method, the energy required for detection is saved, meeting the requirements of portable detection. Also, different ultrasonic coefficients are combined to improve the accuracy of analysis. Moreover, a liver comprehensive verification model is designed for simulation verification processing, realizing the intelligent verification and intelligent evaluation of the analysis results, avoiding the dependence on professional technologies and knowledge, making the detection results more in line with the actual situation. The optimized results are subjected to visual display processing, making the detection results more intuitive and clear, and improving the convenience of detection. The present invention improves the accuracy and convenience of the non-alcoholic fatty liver detection method. Specifically, basic ultrasonic signal parameters are obtained and ultrasonic comprehensive coefficients are extracted. The ultrasonic comprehensive coefficients include the attenuation coefficient and the backscattering coefficient. The extraction of the ultrasonic comprehensive coefficients is based on the linear least squares method and one-dimensional convolutional neural network to obtain more accurate liver fat status information and further improve the accuracy of detection. Quantitative relationship analysis is performed according to the ultrasonic comprehensive coefficients to obtain the ultrasonic fat fraction. The quantitative relationship analysis is based on the energy absorption method. The ultrasonic fat fraction is calculated based on a non-linear regression model. The ultrasonic fat fraction is quantitatively analyzed based on the energy absorption method, saving the energy required for detection and meeting the requirements of portable detection. Also, different ultrasonic coefficients are combined to improve the accuracy of analysis. B-mode ultrasonic imaging processing is performed according to the field programmable gate array architecture to obtain the target basic ultrasonic image. Simulation verification processing is performed according to the liver comprehensive verification model to obtain the target optimized ultrasonic image. The liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model, realizing the intelligent verification and intelligent evaluation of the analysis results, avoiding the dependence on professional technologies and knowledge, making the detection results more in line with the actual situation. A visual evaluation report is generated according to the ultrasonic comprehensive coefficients, the ultrasonic fat fraction, and the target optimized ultrasonic image. The visual evaluation report is based on an image segmentation model. The image segmentation model includes morphological correction optimization and connected domain analysis optimization, making the detection results more intuitive and clear, and improving the convenience of detection. The present invention improves the accuracy and convenience of the non-alcoholic fatty liver detection method.

[0028] Please refer to Figure 2 , which shows the structural schematic diagram of the portable intelligent non-alcoholic fatty liver detection system proposed in the second embodiment of the present invention. The system includes: The coefficient extraction module 10 is used to obtain the basic ultrasonic signal parameters and extract the comprehensive ultrasonic coefficient, where the comprehensive ultrasonic coefficient includes the attenuation coefficient and the backscattering coefficient, and the extraction of the comprehensive ultrasonic coefficient is based on the linear least squares method and the one-dimensional convolutional neural network; The quantitative relationship analysis module 20 is used to perform quantitative relationship analysis based on the comprehensive ultrasonic coefficient to obtain the ultrasonic fat fraction, where the quantitative relationship analysis is based on the energy absorption method, and the ultrasonic fat fraction is calculated based on the non-linear regression model; The ultrasonic imaging module 30 is used to perform B-mode ultrasonic imaging processing according to the field programmable gate array architecture to obtain the target basic ultrasonic image; The simulation verification module 40 is used to perform simulation verification processing according to the comprehensive liver verification model to obtain the target optimized ultrasonic image, where the comprehensive liver verification model includes the basic simulation model and the multi-coefficient verification model; The visualization display module 50 is used to generate a visualization evaluation report based on the comprehensive ultrasonic coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image, where the visualization evaluation report is based on the image segmentation model, and the image segmentation model includes morphological correction optimization and connected component analysis optimization.

[0029] The present invention also proposes a computer storage medium, on which one or more programs are stored, and when the program is executed by a processor, it implements the above-mentioned portable intelligent non-alcoholic fatty liver detection method.

[0030] The present invention also proposes a computer device, including a memory and a processor, where the memory is used to store a computer program, and the processor is used to execute the computer program stored on the memory to implement the above-mentioned portable intelligent non-alcoholic fatty liver detection method.

[0031] Those skilled in the art can understand that the logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0032] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0033] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, the multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: a discrete logic circuit having logic gate circuits for implementing logical functions on data signals, an application-specific integrated circuit having appropriate combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0034] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0035] The above-described embodiments merely represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but should not be construed as limiting the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims

1. A portable intelligent non-alcoholic fatty liver detection method, characterized in that, Including: Obtain basic ultrasonic signal parameters and extract ultrasonic comprehensive coefficients, where the ultrasonic comprehensive coefficients include attenuation coefficient and backscattering coefficient, and the extraction of the ultrasonic comprehensive coefficients is based on linear least squares method and one-dimensional convolutional neural network; Conduct quantitative relationship analysis based on the ultrasonic comprehensive coefficients to obtain ultrasonic fat fraction, where the quantitative relationship analysis is based on energy absorption method, and the ultrasonic fat fraction is calculated based on a non-linear regression model; Conduct B-mode ultrasonic imaging processing according to the field programmable gate array architecture to obtain a target basic ultrasonic image; Conduct simulation verification processing according to the liver comprehensive verification model to obtain a target optimized ultrasonic image, where the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model; Conduct a visual evaluation report based on the ultrasonic comprehensive coefficients, the ultrasonic fat fraction and the target optimized ultrasonic image, where the visual evaluation report is based on an image segmentation model, and the image segmentation model includes morphological correction optimization and connected domain analysis optimization.

2. The portable intelligent non-alcoholic fatty liver detection method according to claim 1, characterized in that, The step of obtaining basic ultrasonic signal parameters and extracting ultrasonic comprehensive coefficients specifically includes: Obtain basic ultrasonic signal parameters, and perform data interception on the depth range in the basic ultrasonic signal parameters to determine the depth axis region; Analyze the basic ultrasonic signal parameters according to a sliding window to obtain envelope signal parameters, adjust the sliding window according to the target dynamics, solve the envelope signal parameters in the sliding window of the depth axis region, determine the depth value by the center of the sliding window of the depth axis region, and obtain a two-dimensional matrix according to the depth value and the frequency value of the basic ultrasonic signal parameters; Estimate the attenuation coefficient and backscattering coefficient according to the weak scattering Born approximation for the depth value and the frequency value, and then solve and extract according to the linear least squares method to obtain the attenuation coefficient and backscattering coefficient; Input the attenuation coefficient and backscattering coefficient into a one-dimensional convolutional neural network for coefficient enhancement extraction.

3. The portable intelligent non-alcoholic fatty liver detection method according to claim 2, characterized in that, The step of inputting the attenuation coefficient and backscattering coefficient into a one-dimensional convolutional neural network for coefficient enhancement extraction specifically includes: Perform time axis-depth axis conversion on the basic ultrasonic signal parameters to intercept the time axis information of the RF signal segment in the basic ultrasonic signal parameters, and adjust the sliding window of the region of interest according to the cross-correlation algorithm to dynamically adjust the sliding window of the time axis region; Input the attenuation coefficient and backscattering coefficient into a one-dimensional convolutional neural network, and the one-dimensional convolutional neural network is constructed based on the historical relevant detection parameters of the detection target; Extract basic ultrasonic features according to depthwise separable convolution, divide the basic ultrasonic features into multiple feature groups of different scales according to the feature scale, perform channel attention enhancement processing on each feature group to obtain multiple attention enhancement feature groups of different scales, and fuse the attention enhancement feature groups step by step according to the order of feature scales from low to high to obtain multi-scale ultrasonic features; Input the multi-scale ultrasonic features into the attenuation coefficient branch and the backscattering coefficient branch respectively for coefficient extraction to obtain an enhanced attenuation coefficient and an enhanced backscattering coefficient respectively.

4. The portable intelligent non-alcoholic fatty liver detection method according to claim 1, characterized in that, The step of performing quantitative relationship analysis based on the ultrasonic comprehensive coefficient to obtain the ultrasonic fat fraction specifically includes: Performing Savitzky-Golay smoothing on the ultrasonic comprehensive coefficient to eliminate respiratory motion artifacts; Performing depth weighting correction on the ultrasonic comprehensive coefficient according to the sound field depth distribution within the region of interest in the basic ultrasonic signal parameters; Inputting the original value and interaction term of the depth-weighted corrected ultrasonic comprehensive coefficient into a non-linear regression model to obtain the ultrasonic fat fraction. The specific algorithm of the non-linear regression model is as follows: ， ， Among them, UDFF represents the original ultrasonic fat fraction, , , represent fitting coefficients, AC represents the attenuation coefficient in the ultrasonic comprehensive coefficient, BSC represents the backscattering coefficient in the ultrasonic comprehensive coefficient, (AC·BSC) represents the interaction term of the ultrasonic comprehensive coefficient, c represents the constant term, represents the ultrasonic fat fraction after standardized mapping, represents the mapping coefficient, represents the mapping constant.

5. The portable intelligent non-alcoholic fatty liver detection method according to claim 1, characterized in that, The step of performing simulation verification processing based on the liver comprehensive verification model to obtain the target optimized ultrasonic image specifically includes: Performing simulation verification processing according to the liver comprehensive verification model, where the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model; The basic simulation model includes a basic scatterer model and a global attenuation constraint. The basic scatterer model includes regular scatterers and diffuse scatterers. The distribution of the regular scatterers is based on the chi-square distribution rule, the amplitude of the regular scatterers is based on the uniform distribution rule, the distribution and amplitude of the diffuse scatterers are both based on the uniform distribution rule. The regular scatterers are used to adjust the isotropy of the spatial distribution of biological soft tissue scattering points to map the one-dimensional scattering point distribution to three-dimensional space, and the distribution mapping is based on the Hilbert space-filling curve; The multi-coefficient verification model includes a basic scatterer model, a global attenuation constraint, and a backscattering constraint. The basic scatterer model of the multi-coefficient verification model is an elastic sphere model. The elastic sphere model adjusts the size and distribution of scatterers based on the intervention theory model and calculates the scattering intensity of a single elastic sphere model. The number density of the basic scatterer model of the multi-coefficient verification model is calculated based on the liver biological characteristic information. The scattering intensity is linearly superimposed based on the scattering intensities of all elastic sphere models, and the scattering intensity is convolved with the spatial pulse of the ultrasonic transducer to obtain the ultrasonic radio frequency echo signal; Performing simulation verification optimization based on the ultrasonic radio frequency echo signal to obtain the target optimized ultrasonic image.

6. The portable intelligent non-alcoholic fatty liver detection method according to claim 1, wherein The step of generating a visual evaluation report based on the ultrasonic comprehensive coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image specifically includes: Performing preprocessing on the target optimized ultrasonic image. The preprocessing includes speckle noise suppression processing and contrast enhancement processing. The speckle noise suppression processing is based on non-local means filtering, and the contrast enhancement processing is based on the adaptive histogram equalization algorithm with local contrast limitation; Performing segmentation processing on the preprocessed target optimized ultrasonic image according to an image segmentation model to obtain a liver region mask, and then performing interference structure removal processing on the liver region mask. The interference structures include blood vessel structures and mass structures; Performing postprocessing on the liver region mask. The postprocessing includes morphological correction optimization and connected component analysis optimization. Closing operations are performed according to the morphological correction optimization algorithm to fill the hole regions, and then the largest region is retained according to the connected component analysis optimization to obtain the optimized segmentation result of the liver region; Perform a heat map color mapping on the optimized segmentation result of the liver region according to the ultrasonic fat fraction and the ultrasonic comprehensive coefficient, and then convert the optimized segmentation result of the liver region into a color layer through transparency fusion to be superimposed on the target optimized ultrasonic image to obtain a visual display result; Perform parameter and evaluation annotation on the visual display result according to the ultrasonic fat fraction and the ultrasonic comprehensive coefficient to obtain a visual evaluation report.

7. A portable intelligent non-alcoholic fatty liver detection system, characterized in that, Including: A coefficient extraction module for obtaining basic ultrasonic signal parameters and extracting an ultrasonic comprehensive coefficient, where the ultrasonic comprehensive coefficient includes an attenuation coefficient and a backscattering coefficient, and the extraction of the ultrasonic comprehensive coefficient is based on the linear least squares method and a one-dimensional convolutional neural network; A quantitative relationship analysis module for performing quantitative relationship analysis according to the ultrasonic comprehensive coefficient to obtain an ultrasonic fat fraction, where the quantitative relationship analysis is based on the energy absorption method, and the ultrasonic fat fraction is calculated based on a nonlinear regression model; An ultrasonic imaging module for performing B-mode ultrasonic imaging processing according to the field programmable gate array architecture to obtain a target basic ultrasonic image; A simulation verification module for performing simulation verification processing according to a liver comprehensive verification model to obtain a target optimized ultrasonic image, where the liver comprehensive verification model includes a basic simulation model and a multi-coefficient verification model; A visual display module for generating a visual evaluation report according to the ultrasonic comprehensive coefficient, the ultrasonic fat fraction, and the target optimized ultrasonic image, where the visual evaluation report is based on an image segmentation model, and the image segmentation model includes morphological correction optimization and connected domain analysis optimization.

8. The portable intelligent non-alcoholic fatty liver detection system according to claim 7, wherein The ultrasonic imaging module specifically includes: an ultrasonic transducer, a front-end signal transceiver processor, and a back-end ultrasonic image processor; The front-end signal transceiver processor includes a transmitting front-end unit, a receiving front-end unit, a beam processing unit, and a signal processing unit. The transmitting front-end unit is used to generate high-voltage excitation pulses, focus digital electronics, beam scanning adjustment, and aperture size control. The receiving front-end unit is used to amplify the echo signal at the front end, perform analog-to-digital conversion, and high-speed transmission. The beam processing unit is used for digital signal processing, variable aperture reception, dynamic apodization, and beam synthesis. The signal processing unit is used for gain control, dynamic filtering, envelope detection, secondary sampling, logarithmic compression, and coordinate transformation processing, and transmits the finally generated signal to the back-end ultrasonic image processor.

9. A storage medium, characterized in that, The storage medium stores one or more programs, and when the program is executed by a processor, it implements the portable intelligent non-alcoholic fatty liver detection method according to any one of claims 1-6.

10. A computer device, characterized in that, The computer device includes a memory and a processor, where: The memory is used to store a computer program; When the processor executes the computer program stored on the memory, it implements the portable intelligent non-alcoholic fatty liver detection method according to any one of claims 1-6.

Images