An ultrasonic fast imaging method combining difference compensator and adaptive denoising

Abstract

The application discloses a kind of combined difference compensator and ultrasonic fast imaging method of adaptive denoising, the method is as follows: one: based on the spatial focusing and diffusion characteristics of ultrasonic sound field propagation establishes ultrasonic imaging degradation model, and extracts the PSF information of local and defocusing imaging area, by Fourier transform conversion to frequency domain, based on ideal imaging condition, the frequency domain difference of local and overall PSF is analyzed, frequency domain difference compensator is designed, high-frequency information is expanded, and overall imaging performance is optimized;Two: design signal-to-noise ratio self-estimation module, analyze the signal-to-noise ratio of input image, and construct adaptive parameterized denoising module combined with decomposition level and sub-band position information, effectively remove noise and enhance image contrast, obtain contrast-enhanced ultrasonic reconstructed image.The method is efficient in calculation, without complex parameter adjustment, can quickly optimize ultrasonic imaging quality, suitable for traditional and portable ultrasonic equipment, and has wide application value in medical imaging field.

Description

Technical Field

[0001] This invention relates to an ultrasound imaging post-processing algorithm, specifically to a rapid ultrasound imaging method that combines a difference compensator and adaptive denoising. Background Technology

[0002] Ultrasound imaging, due to its low cost, portability, ease of operation, and non-invasiveness, has become one of the most commonly used tools in medical diagnosis, and is widely applied in various fields such as cardiology, abdominal surgery, obstetrics, musculoskeletal surgery, and vascular surgery. However, current ultrasound equipment still has limitations in resolution and contrast, which severely restricts its application in refined diagnosis. To improve image quality, researchers have explored image reconstruction techniques such as deep learning, beamforming, and deconvolution, and have achieved some success. However, even with state-of-the-art equipment, ultrasound imaging still presents an inherent trade-off between reconstruction speed, flexibility, and image quality. High-resolution imaging often requires longer processing times, affecting real-time performance; while increasing processing speed may lead to a decrease in image quality. This challenge is particularly prominent in low-performance miniaturized devices, such as wearable and portable ultrasound devices, which are limited by hardware resources and computing power, and urgently need an efficient method that can improve image quality without significantly increasing the computational burden. Therefore, developing a practical technology for rapidly and automatically improving the quality of ultrasound imaging is not only extremely challenging but also has significant clinical application value. Summary of the Invention

[0003] This invention addresses the trade-offs between reconstruction time, imaging performance, and flexibility in ultrasound imaging. By combining the differences between the focused region and the overall point spread function (PSF) information in the frequency domain, and based on preliminary estimation of the signal-to-noise ratio (SNR), decomposition hierarchy, and sub-band position information in the wavelet domain, it provides a rapid ultrasound imaging method that integrates a difference compensator and adaptive denoising. This method is computationally efficient, requires no complex parameter adjustments, and can quickly optimize ultrasound imaging quality. It overcomes the limitations of ultrasound equipment in terms of resolution, SNR, and contrast, and is applicable to both traditional and portable ultrasound devices, possessing broad application value in the field of medical imaging.

[0004] The objective of this invention is achieved through the following technical solution:

[0005] A rapid ultrasound imaging method combining a difference compensator and adaptive denoising includes the following steps:

[0006] Step 1: Based on the spatial focusing and diffusion characteristics of ultrasonic sound field propagation, establish an ultrasonic imaging degradation model, extract PSF information of the focused and overall imaging areas, transform it to the frequency domain through Fourier transform, and use the local focused area as the target ideal imaging condition to analyze the frequency domain difference between the local focused area and the overall PSF. Design a frequency domain difference compensator to expand high-frequency information, optimize the overall imaging performance, and obtain an image with enhanced resolution. The specific steps are as follows:

[0007] Step 11: Under the assumptions of linear propagation and weak scattering, the pressure field received by the ultrasound system is modeled using the first-order Born approximation. Considering the influence of measurement noise, the ultrasound imaging model represents the radio frequency (RF) image as the convolution result of the PSF and the tissue reflection function (TRF), with a noise term added. Its mathematical expression is as follows:

[0008]

[0009] Where x and y represent the horizontal and vertical sampling directions, respectively. This indicates that an RF image was observed. Indicates PSF, Indicates measurement noise. Describes the TRF to be determined. This represents the convolution operation;

[0010] The ultrasound imaging degradation model is reconstructed by decomposing the acquired radiofrequency image into a superposition of the convolution results of the TRF, the focused region PSF, and other defocused region PSFs. The specific expression is as follows:

[0011]

[0012] in, and These are the PSF values ​​for the in-focus area and the PSF values ​​for the out-of-focus area, respectively. and These represent the TRF information for the focused and out-of-focus areas, respectively. The obtained degraded image information;

[0013] Steps 1 and 2: Obtain ultrasound imaging results with higher imaging performance under focused PSF and noise-free imaging conditions:

[0014]

[0015] in, The imaging results, representing the focusing PSF and noise-free imaging conditions, are expressed in the spectral domain as follows: (The Fourier transform of the above equation yields the following expression:)

[0016]

[0017] in, , , They are respectively , , The Fourier transform result, The multiplication symbol is used. Let x and y represent the frequency forms, respectively.

[0018] In the out-of-focus areas where the imaging resolution is poor, the image is represented as follows:

[0019]

[0020] in, The above formula, representing the imaging result of the defocused region, is expressed in the spectral domain as:

[0021]

[0022] in, , , , They are respectively , , , The Fourier transform form;

[0023] Step 13: Based on Step 11 and Step 12, design a frequency domain differential compensator with the PSF under ideal imaging conditions and noise-free imaging results as the objectives. The high-frequency content in the extended far-field region is calculated using the following formula:

[0024]

[0025] The differential compensator is obtained by minimizing the deviation between the ideal imaging result and the calibration output. The optimal solution is as follows:

[0026]

[0027] Where E represents the expectation operation, To obtain the optimal value, we derive the following: By simultaneously differentiating both sides of the above equation and setting the result to zero. Explicit expression:

[0028]

[0029] in, for transpose, As a regularization factor;

[0030] Finally, by integrating the above equations, an image with enhanced resolution is obtained, as shown below:

[0031]

[0032] in, This represents the inverse Fourier transform. This represents a reconstructed image with increased resolution.

[0033] Step 2: Based on the image with enhanced resolution obtained in Step 1, a signal-to-noise ratio (SNR) self-estimation module is designed to analyze the SNR of the input image. An adaptive parametric denoising module is then constructed by combining information from the decomposition level and sub-band position. This effectively removes noise and enhances image contrast, resulting in a contrast-enhanced ultrasound reconstructed image. The specific steps are as follows:

[0034] Step 21: Design a signal-to-noise ratio (SNR) self-estimation module to estimate the overall SNR of the image. This module takes a noisy image and a clean image as input, wherein: a wavelet denoising method is used to estimate the clean image, followed by SNR calculation, as shown below:

[0035]

[0036] Among them, SNR est It is the signal-to-noise ratio estimate. It's a noisy image. This is the estimated clean image; subsequently, the signal-to-noise ratio weighting parameter is adjusted. With SNR est Related, if SNR est An index greater than 25 indicates the image is considered relatively clean. Set to 1; for other cases, adjust based on 25 as the reference center, as shown below:

[0037]

[0038] Where max represents the maximum value operation;

[0039] Step 22: Taking into account the decomposition level and its corresponding sub-band position, design the weighting factor: decomposition level Sub-band position And combined with signal-to-noise ratio weighting adjustment parameters The threshold weighting factor F for the noise is adaptively adjusted as follows:

[0040]

[0041] Where, min represents the minimum value operation, and:

[0042]

[0043]

[0044] Where k and L represent the current decomposition level and the maximum decomposition level, respectively, and s represents the decomposition subband. The weighting factor representing the adjustment threshold is obtained through... By adaptively adjusting the wavelet processing threshold and filtering out noise, a guiding image is obtained. The guiding image is then used to guide Wiener denoising using a dual wavelet basis, achieving a trade-off between edge protection and noise removal, and finally obtaining an ultrasound reconstructed image with enhanced contrast.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] 1. This invention addresses the spatial transformation characteristics in ultrasound imaging by constructing an ultrasound imaging degradation model based on the frequency domain information difference between the focused area and the global PSF, and designing a difference compensator to improve imaging resolution. Simultaneously, through in-depth analysis of noise distribution, a signal-to-noise ratio self-estimation module is constructed, and an adaptive denoising module is designed by combining hierarchical and positional information to effectively remove noise and enhance image contrast.

[0047] 2. This invention proposes a solution that meets the needs of practical applications. It cleverly utilizes the difference in PSF frequency domain information, deeply considers the actual imaging capabilities, proposes a difference compensation method to quickly improve resolution in the frequency domain, and designs an adaptive parameterless denoising model in the wavelet domain, which can quickly and effectively improve the quality of ultrasound imaging.

[0048] 3. While ensuring rapid reconstruction, this invention achieves simultaneous optimization of ultrasound image resolution and contrast, overcoming the difficulty of complex parameter adjustment in traditional methods, and improving the practicality and reliability of ultrasound imaging in medical diagnosis.

[0049] 4. This invention is applicable to various medical imaging devices and complex imaging environments, providing a higher quality and more stable solution for ultrasound imaging. Attached Figure Description

[0050] Figure 1 This is a flowchart of a rapid ultrasound imaging method that combines a difference compensator and adaptive denoising.

[0051] Figure 2 This is a block diagram illustrating the execution structure of a rapid ultrasound imaging method that combines a difference compensator and adaptive denoising.

[0052] Figure 3 Designed for the SNRest module.

[0053] Figure 4 These are the results of the phantom experiment.

[0054] Figure 5 These are results from human trials. Detailed Implementation

[0055] The technical solution of the present invention will be further described below with reference to the accompanying drawings, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention that do not depart from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention.

[0056] This invention establishes a mathematical degradation model for ultrasound imaging based on the focusing and diffusion characteristics of ultrasonic sound field propagation. Building upon this model, it fully considers the information difference in the frequency domain between the PSF of the local focused area and the overall imaging characteristics. Simultaneously, it combines the signal-to-noise ratio (SNR) estimation of the input data with the hierarchical and sub-band position information of wavelet decomposition to provide a rapid ultrasound imaging method that integrates a joint difference compensator and adaptive denoising. Specifically, firstly, a degradation model of the ultrasound imaging system is constructed based on the focusing and diffusion characteristics of the ultrasonic sound field. Then, Fourier transforms are performed on the PSF of the local focused and defocused areas to simplify the calculation process and improve computational efficiency. Based on this, a difference compensator is designed with noise-free ideal imaging in the focused state as the optimization objective. This compensator performs high-frequency information compensation for the defocused areas with poor imaging performance, thereby improving resolution. Furthermore, the resolution-enhanced image is input into a signal-to-noise ratio self-estimation module to evaluate image quality. The noise distribution characteristics are analyzed by combining the hierarchical and sub-band position information of wavelet decomposition, constructing an adaptive denoising model that does not require manual parameter tuning to improve image contrast and optimize imaging effects. Figure 1 and Figure 2 As shown, the specific steps are as follows:

[0057] Step 1: Based on the spatial focusing and diffusion characteristics of ultrasonic sound field propagation, establish an ultrasonic imaging degradation model, extract PSF information of the focused and overall imaging areas, transform it to the frequency domain through Fourier transform, and use the local focused area as the target ideal imaging condition to analyze the frequency domain difference between the local focused area and the overall PSF. Design a frequency domain difference compensator to expand high-frequency information, optimize the overall imaging performance, and obtain an image with enhanced resolution. The specific steps are as follows:

[0058] Step 11: Under the assumptions of linear propagation and weak scattering, the pressure field received by the ultrasound system can be modeled using the first-order Born approximation. Considering the influence of measurement noise, the ultrasound imaging model represents the RF image as the convolution result of PSF and TRF, with a noise term superimposed. Its mathematical expression is as follows:

[0059]

[0060] Where x and y represent the horizontal and vertical sampling directions, respectively. This indicates that an RF image was observed. Indicates PSF, Indicates measurement noise. Describes the TRF to be determined. This represents the convolution operation.

[0061] In the above methods, the PSF is assumed to be spatially invariant. However, this assumption does not hold true in practical applications, leading to distortion in the recovered image. Due to the focusing and diffusion characteristics of ultrasound sound field propagation, the focused region has the highest imaging resolution, while other regions have poorer imaging performance. Based on this, a new ultrasound imaging degradation model is constructed, decomposing the acquired radio frequency image into a superposition of the convolution results of the TRF, the focused region PSF, and the PSF of other defocused regions. The specific expression is as follows:

[0062]

[0063] in, and These are the PSF values ​​for the in-focus area and the PSF values ​​for the out-of-focus area, respectively. and These represent the TRF information for the focused and out-of-focus areas, respectively. This refers to the obtained degraded image information.

[0064] Steps one and two: If, under focused PSF and noise-free imaging conditions, ultrasound imaging results with higher imaging performance can be obtained:

[0065]

[0066] in, The imaging results, representing the focusing PSF and noise-free imaging conditions, can be expressed in the spectral domain as follows: (The Fourier transform of the above equation yields the following expression:)

[0067]

[0068] in, , , They are respectively , , The Fourier transform result, The multiplication symbol is used. Let x and y represent the frequency forms, respectively.

[0069] In the out-of-focus areas where the imaging resolution is poor, the image is represented as follows:

[0070]

[0071] in, The above formula, representing the imaging result of the defocused region, is expressed in the spectral domain as:

[0072]

[0073] in, , , , They are respectively , , , The Fourier transform form of .

[0074] Step 13: Based on Step 11 and Step 12, design a frequency domain differential compensator with the PSF under ideal imaging conditions and noise-free imaging results as the objectives. To expand the high-frequency content in the far-field region, the calculation formula is as follows:

[0075]

[0076] A differential compensator can be obtained by minimizing the deviation between the ideal imaging result and the calibration output. The optimal solution is as follows:

[0077]

[0078] Where E represents the expectation operation, To obtain the optimal value, by simultaneously differentiating both sides of the above equation and setting the result to zero, we can derive the following: Explicit expression:

[0079]

[0080] in, for transpose, As a regularization factor;

[0081] Finally, by integrating the above equations, an image with enhanced resolution can be obtained, as shown below:

[0082]

[0083] in, This represents the inverse Fourier transform. This represents a reconstructed image with increased resolution.

[0084] Step Two: Resolution and contrast are two key indicators in ultrasound imaging, both of which are indispensable. To further suppress noise such as clutter and artifacts to improve contrast, based on the image with enhanced resolution obtained in Step One, a signal-to-noise ratio (SNR) self-estimation module is designed to analyze the SNR of the input image. An adaptive parametric denoising module is then constructed by combining information on the decomposition level and sub-band position to effectively remove noise and enhance image contrast, resulting in an ultrasound reconstructed image with improved contrast. The specific steps are as follows:

[0085] Step Two One: As Figure 3 As shown, a signal-to-noise ratio (SNR) self-estimation module is designed to estimate the overall SNR of an image. This module requires two inputs: a noisy image and a clean image. However, obtaining a noise-free signal in the real world is impractical for ultrasound images. To address this issue, this invention employs a wavelet denoising method with general parameter settings to estimate the clean image, followed by SNR calculation, as follows:

[0086]

[0087] Among them, SNR est It is the signal-to-noise ratio estimate. It's a noisy image. This is the estimated clean image. Subsequently, the present invention adjusts the signal-to-noise ratio weighting parameters. With SNR est Related, specifically: if SNR est An index greater than 25 indicates the image is considered relatively clean. Set it to 1; otherwise, adjust it based on 25 as the center, as shown below:

[0088]

[0089] Here, max represents the operation of finding the maximum value.

[0090] Step 22: The noise distribution is also closely related to the wavelet decomposition level and sub-band position. Low-frequency sub-bands contain more image information, while high-frequency components are more susceptible to noise. Therefore, applying a larger threshold to the high-frequency sub-band can more effectively suppress noise. On the other hand, as the decomposition level increases, the frequency bandwidth of the generated sub-bands gradually narrows, and the corresponding threshold should be appropriately reduced. Based on this analysis, this invention comprehensively considers the decomposition level and its corresponding sub-band position, and designs a weighting factor: decomposition level Sub-band position And combined with signal-to-noise ratio weighting adjustment parameters The threshold weighting factor F for the noise is adaptively adjusted as follows:

[0091]

[0092] Where, min represents the minimum value operation, and:

[0093]

[0094]

[0095] Where k and L represent the current decomposition level and the maximum decomposition level, respectively, and s represents the decomposition subband. The weighting factor representing the adjustment threshold is obtained through... By adjusting the wavelet processing threshold through adaptive adjustment, noise is filtered out in a more robust and flexible manner to obtain a guide image. The guide image is then used in a Wiener denoising framework based on a dual wavelet basis to perform a trade-off between edge protection and noise removal, ultimately resulting in an ultrasound reconstructed image with enhanced contrast.

[0096] Example:

[0097] Step 1: Based on linear propagation theory and the weak scattering assumption, the first-order Born approximation is used to model the signal received by the ultrasound system. Taking into full account the influence of measurement noise, and addressing the impact of spatial transformation characteristics on image quality in traditional ultrasound imaging, the RF image is decomposed into the focal region PSF and the out-of-focus region PSF. These are then convolved with their corresponding TRFs and superimposed to more accurately characterize the spatial transformation effect during imaging. Based on this, optimization modeling is performed in the frequency domain using the focal region PSF and noise-free conditions as optimization objectives. This focuses on addressing the problem of degraded image quality in the out-of-focus region. To this end, a frequency domain differential compensator is designed. By minimizing the deviation between the ideal imaging result and the calibration output, the compensation strategy is optimized to adaptively recover the high-frequency information in the out-of-focus region. Ultimately, this method effectively expands the frequency bandwidth of the out-of-focus region, improves the spatial resolution of ultrasound imaging, and makes the overall image quality closer to the ideal imaging effect, providing more accurate tissue structure information for ultrasound diagnosis.

[0098] Step Two: Based on the resolution-enhanced image obtained in Step One, an adaptive denoising module was further designed to improve contrast. Specifically, an SNRest module was designed to estimate the overall signal-to-noise ratio (SNR) of the image. Since a noise-free reference image cannot be directly obtained for ultrasound images, wavelet denoising was used to preprocess the input image to estimate a clean image and calculate its SNR. Subsequently, the estimated SNR value was compared with the weight adjustment parameters. Correspondingly, when the estimated signal-to-noise ratio (SNR) is higher than a set threshold (e.g., 25), it indicates that the image is relatively clean, and the weight is set to 1; otherwise, it is dynamically adjusted based on the set threshold to enhance the denoising capability for low SNR images. Simultaneously, it incorporates the hierarchical information of the influence of decomposition level and sub-band position on noise, along with positional weight factors. and Specifically, the distribution characteristics of noise differ across different frequency sub-bands. Low-frequency sub-bands contain more image information, while high-frequency sub-bands are more susceptible to noise contamination. Therefore, a larger denoising threshold is applied to high-frequency sub-bands, while the threshold for low-frequency sub-bands is relatively smaller. Furthermore, as the wavelet decomposition level increases, the frequency bandwidth of the sub-bands narrows, requiring a gradual decrease in the corresponding denoising threshold. Finally, the three influencing weighting factors are integrated, and the threshold is adaptively adjusted based on the image information, avoiding manual adjustment or the use of fixed values, thus improving the robustness and flexibility of the method.

[0099] Finally, the method proposed in this invention was verified, and the results are as follows: Figure 4 and Figure 5 As shown in the results, the method proposed in this invention can meet real-time requirements without relying on any parameter adjustments, and performs well in terms of image resolution and contrast.

Claims

1. A rapid ultrasound imaging method combining a difference compensator and adaptive denoising, characterized in that... The method includes the following steps: Step 1: Based on the spatial focusing and diffusion characteristics of ultrasonic sound field propagation, establish an ultrasonic imaging degradation model, extract PSF information of the focused and overall imaging areas, transform it to the frequency domain through Fourier transform, and use the local focused area as the target ideal imaging condition to analyze the frequency domain difference between the local focused area and the overall PSF. Design a frequency domain difference compensator to expand high-frequency information, optimize the overall imaging performance, and obtain an image with enhanced resolution. The specific steps are as follows: Step 11: Under the assumptions of linear propagation and weak scattering, the pressure field received by the ultrasound system is modeled using the first-order Born approximation. Considering the influence of measurement noise, the ultrasound imaging model represents the radio frequency (RF) image as the convolution result of the PSF and the tissue reflection function (TRF), with a noise term added. Its mathematical expression is as follows: Where x and y represent the horizontal and vertical sampling directions, respectively. This indicates that an RF image was observed. Indicates PSF, Indicates measurement noise. Describes the TRF to be determined. This represents the convolution operation; The ultrasound imaging degradation model is reconstructed by decomposing the acquired radiofrequency image into a superposition of the convolution results of the TRF, the focused region PSF, and other defocused region PSFs. The specific expression is as follows: in, and These are the PSF values ​​for the in-focus area and the PSF values ​​for the out-of-focus area, respectively. and These represent the TRF information for the focused and out-of-focus areas, respectively. The obtained degraded image information; Steps 1 and 2: Obtain ultrasound imaging results with higher imaging performance under focused PSF and noise-free imaging conditions: in, The imaging results, representing the focusing PSF and noise-free imaging conditions, are expressed in the spectral domain as follows: (The Fourier transform of the above equation yields the following expression:) in, , , They are respectively , , The Fourier transform result, The multiplication symbol is used. Let x and y represent the frequency forms, respectively. In the out-of-focus areas where the imaging resolution is poor, the image is represented as: in, The above formula, representing the imaging result of the defocused region, is expressed in the spectral domain as: in, , , , They are respectively , , , The Fourier transform form; Step 13: Based on Step 11 and Step 12, design a frequency domain differential compensator with the PSF under ideal imaging conditions and noise-free imaging results as the objectives. The high-frequency content in the extended far-field region is calculated using the following formula: The differential compensator is obtained by minimizing the deviation between the ideal imaging result and the calibration output. The optimal solution is as follows: Where E represents the expectation operation, To obtain the optimal value, we derive the following: By simultaneously differentiating both sides of the above equation and setting the result to zero. Explicit expression: in, for transpose, As a regularization factor; Finally, by integrating the above equations, an image with enhanced resolution is obtained, as shown below: in, This represents the inverse Fourier transform. Represents a reconstructed image with increased resolution; Step 2: Based on the image with enhanced resolution obtained in Step 1, a signal-to-noise ratio self-estimation module is designed to analyze the signal-to-noise ratio of the input image. An adaptive parametric denoising module is constructed by combining the information of decomposition level and sub-band position to effectively remove noise and enhance image contrast, thereby obtaining an ultrasound reconstructed image with improved contrast.

2. The rapid ultrasound imaging method with combined difference compensator and adaptive denoising according to claim 1, characterized in that... The specific steps of step two are as follows: Step 21: Design a signal-to-noise ratio self-estimation module to estimate the overall signal-to-noise ratio of the image. This module takes a noisy image and a clean image as input. Step 22: Taking into account the decomposition level and its corresponding sub-band position, design the weighting factor: decomposition level Sub-band position And combined with signal-to-noise ratio weighting adjustment parameters The threshold weighting factor for the noise is adaptively adjusted as follows: Where min represents the minimum value operation. The weighting factor representing the adjustment of the noise threshold is obtained by... By adaptively adjusting the wavelet processing threshold and filtering out noise, a guiding image is obtained. The guiding image is then used to guide Wiener denoising using a dual wavelet basis, achieving a trade-off between edge protection and noise removal, and finally obtaining an ultrasound reconstructed image with enhanced contrast.

3. The rapid ultrasound imaging method with combined difference compensator and adaptive denoising according to claim 2, characterized in that... In step two, a clean image is estimated using a wavelet denoising method, followed by signal-to-noise ratio calculation, as shown below: Among them, SNR est It is the signal-to-noise ratio estimate. It's a noisy image. This is the estimated clean image; subsequently, the signal-to-noise ratio weighting parameter is adjusted. With SNR est Related, if SNR est If the value is greater than 25, the image is considered relatively clean. Set to 1; for other cases, adjust based on 25 as the reference center, as shown below: Here, max represents the operation of finding the maximum value.

4. The rapid ultrasound imaging method with combined difference compensator and adaptive denoising according to claim 2, characterized in that... In step two, the decomposition level is... Sub-band position The following conditions must be met: Where k and L represent the current decomposition level and the maximum decomposition level, respectively, and s represents the decomposition subband.

Images