Method, system and computer readable storage medium for low signal noise suppression of ct scanning devices

Abstract

The present application provides a method for suppressing low signal noise of a CT scanning device, comprising the following steps: acquiring data collected by the CT scanning device, and performing low signal threshold T LS The low signal channel needing filtering is screened out; the range of the fixed spatial neighborhood of the low signal channel is acquired; the upper threshold and the lower threshold of the low signal channel to be processed are acquired; the signal to be processed is selected in the fixed spatial neighborhood according to the upper threshold and the lower threshold of the low signal channel; different weights are assigned to a plurality of the signals to be processed according to their sizes for weighted calculation, so as to obtain the final linear attenuation coefficient μ0 of the low signal channel; and the signal channel is filtered by using the final linear attenuation coefficient μ0, so as to obtain the low signal noise after suppression.

Description

[0001] Technology Neighborhood

[0002] This invention relates to the field of CT scan signal processing, and more particularly to a method, system, and computer-readable storage medium for suppressing low signal noise in a CT scan device. Background Technology

[0003] Computed tomography (CT) is a device that uses rotating X-rays to irradiate an object and then processes the data to obtain a cross-sectional image. X-ray photons pass through the irradiated object, reach the detector, and are collected. After collection, they undergo a series of conversions, transforming the photon signal into an electronic signal that is received. The received electronic signal contains noise, primarily composed of electronic noise and photon noise (Poisson noise). As X-ray penetration attenuation increases, fewer X-ray photons reach the detector, resulting in a lower signal strength and more significant noise, leading to a lower signal-to-noise ratio. Significant noise at low signal strength can cause severe streak artifacts in the reconstructed CT image, seriously impacting clinical diagnosis.

[0004] US Patent 8965144B2 proposes a method for suppressing low-signal noise in the projection domain. This method uses adaptive neighborhood Gaussian filtering for low signals. First, it evaluates the noise level of each low-signal data point in the projection domain, and then determines the variance of the Gaussian distribution and the corresponding Gaussian smoothing kernel based on this noise level. The variance of the Gaussian distribution characterizes the smoothing ability of the corresponding Gaussian kernel. Finally, the calculated Gaussian kernel is used to filter the low-signal data to suppress low-signal noise. However, this method uses all neighborhood data of the low signal. There may be low signals with excessively large noise level deviations in the neighborhood data, and their corresponding filtering weights are also relatively large. These very noisy low signals can still have a significant impact on the filtered output, resulting in unsatisfactory noise suppression. For Gaussian filter kernels with excessively large differences in smoothness, different spatial neighborhood ranges are required. Furthermore, calculating the corresponding spatial Gaussian kernel and performing corresponding spatial filtering for each low-signal data point involves a large amount of computation and is time-consuming.

[0005] The journal doi:10.1118 / 1.598410 also proposes a projection domain method for suppressing low-signal noise. This method proposes an adaptive mean filtering method for subtracting neighborhood data based on the noise attributes of the detector. The method determines the neighborhood range and the number of maximum and minimum signals to be subtracted within the neighborhood range based on the noise level of the current channel's low signal. Finally, it calculates the mean of the remaining neighborhood data after subtraction; this mean is the filtered value for the current channel's low signal. However, this method requires sorting the selected neighborhood data during the process of subtracting the maximum and minimum signals. This sorting calculation consumes significant computational resources, affecting the program's computation speed. Furthermore, although this method subtracts some of the maximum and minimum data within the neighborhood, there is still a possibility that low signals with excessively high noise levels may not be subtracted.

[0006] Therefore, the methods proposed in patent US8965144B2 and journal doi:10.1118 / 1.598410 select neighborhood data of different ranges based on signal levels and suppress noise by calculating the data within the neighborhood range. However, when there are low signals with particularly high noise levels in the neighborhood, the noise suppression effect is not significant.

[0007] Existing technologies mainly focus on two aspects: first, eliminating artifacts caused by low-signal noise in the reconstructed image; and second, filtering the projection data in the projection domain to suppress noise. However, the methods in patent US8965144B2 and journal doi:10.1118 / 1.598410 belong to the category of noise suppression methods in the projection domain, and neither can guarantee the elimination of the influence of abnormal data. Summary of the Invention

[0008] In order to overcome the above-mentioned technical defects, the present invention aims to provide a method, system and computer-readable storage medium for suppressing low signal noise in a CT scanning device that is faster and more reliable.

[0009] This invention discloses a method for suppressing low signal noise in a CT scanning device, comprising the following steps: passing the data acquired by the CT scanning device through a low signal threshold T... LS The process involves: identifying low-signal channels that require filtering; obtaining the range of a fixed spatial neighborhood for each low-signal channel; obtaining an upper and lower threshold for the low-signal channel to be processed; selecting signals to be processed within the fixed spatial neighborhood based on the upper and lower thresholds; assigning different weights to several signals to be processed according to their magnitudes for weighted calculation to obtain the final linear attenuation coefficient μ0 for the low-signal channel; and using the final linear attenuation coefficient μ0 to filter the signal channel to obtain suppressed low-signal noise.

[0010] Preferably, the fixed spatial neighborhood includes an angular neighborhood; the fixed spatial neighborhood is l×m; where l is the width of the fixed spatial neighborhood in the channel direction, and m is the width of the fixed spatial neighborhood in the angular direction.

[0011] Preferably, obtaining the upper and lower thresholds of the low-signal channel to be processed includes: the upper threshold is obtained by... The lower threshold is determined by... To determine; where I is the acquired signal strength and k is the amplification factor of the photon acquisition system.

[0012] Preferably, the upper threshold is μ+n·δ + The lower threshold is μ-n·δ - Where μ is the linear attenuation coefficient of the low signal channel, and n is the multiple of the upper and lower bound deviations applied.

[0013] Preferably, the step of assigning different weights to several signals to be processed according to their magnitude for weighted calculation to obtain the final linear attenuation coefficient μ0 of the low signal channel includes: the final linear attenuation coefficient μ0 of the low signal channel is determined by selecting the starting offset a of the N weights in the center of the one-dimensional low-pass filter kernel. Wherein, N is the number of the signals to be processed.

[0014] Preferably, the final linear attenuation coefficient Where, μ i For the i-th signal to be processed, which is the result of arranging the signals to be processed in ascending order, w a+i The weight is the i-th signal to be processed after arranging the signals to be processed in ascending order.

[0015] This invention also discloses a low-signal noise suppression system for a CT scanning device, comprising a signal processing module and a filtering module connected together; the signal processing module processes the data acquired by the CT scanning device through a low-signal threshold T. LS The process involves: identifying low-signal channels that require filtering; obtaining a fixed spatial neighborhood range for each low-signal channel using the signal processing module; acquiring an upper and lower threshold for each low-signal channel; selecting signals to be processed within the fixed spatial neighborhood based on the upper and lower thresholds; assigning different weights to several signals to be processed for weighted calculation to obtain the final linear attenuation coefficient μ0 for each low-signal channel; and using the final linear attenuation coefficient μ0 to filter the signal channel using the filtering module to obtain suppressed low-signal noise.

[0016] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the low signal-to-noise suppression method described above.

[0017] Compared with existing technologies, the above technical solution has the following advantages:

[0018] 1. Compared to patent US8965144B2, when processing each low signal channel, there is no need to calculate the neighborhood range, generate a spatial filter kernel, or use a spatial filter kernel, which reduces the computational complexity, saves computing resources, and reduces computation time.

[0019] 2. Patent US8965144B2 uses all neighborhood data of low signals, and journal doi:10.1118 / 1.598410 only sorts out the maxima and minima in the spatial neighborhood. Both of these methods inevitably use neighborhood data with larger errors to suppress low signals. However, this patent uses a physical model of noise and can effectively eliminate low signals with excessive noise levels in the neighborhood by filtering with the upper and lower thresholds, thus ensuring the effect of noise suppression.

[0020] 3. Patent US8965144B2 and journal doi:10.1118 / 1.598410 use spatial neighborhoods on the detector (channel neighborhood and detector row neighborhood), and both use adaptively changing spatial neighborhood ranges. However, this invention uses a fixed multidimensional spatial neighborhood that combines angular neighborhoods, which increases the amount of data sampled and improves the reliability of data filtering.

[0021] 4. This invention performs weighted calculations on the filtered data in the numerical space to further ensure the reliability of the data. Attached Figure Description

[0022] Figure 1 A flowchart of a method for suppressing low signal noise in a CT scanning device provided by the present invention;

[0023] Figure 2 A flowchart illustrating the process of calculating the final linear attenuation coefficient in the low signal noise suppression method for CT scanning equipment provided by this invention.

[0024] Figure 3 This is a schematic diagram illustrating the numerical sorting of N filtered signals to be processed in the low signal noise suppression method for the CT scanning device provided by the present invention.

[0025] Figure 4 The reconstructed image provided by the present invention is neither processed by the low signal noise suppression method nor by the low signal noise suppression method. Detailed Implementation

[0026] The advantages of the present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments.

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0028] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The singular forms “a,” “the,” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0029] It should be understood that although the terms first, second, third, etc., may be used in this disclosure to describe various information, such information should not be limited to these terms. These terms are used only to distinguish information of the same type from one another. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0030] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0031] In the description of this invention, unless otherwise specified and limited, it should be noted that the terms "installation", "connection" and "linking" should be interpreted broadly. For example, they can refer to mechanical or electrical connections, or internal connections between two components. They can be direct connections or indirect connections through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0032] In the following description, suffixes such as "module," "part," or "unit" used to denote elements are used only for the convenience of the description of the invention and have no specific meaning in themselves. Therefore, "module" and "part" can be used interchangeably.

[0033] See appendix Figure 1 This invention discloses a method for suppressing low signal noise in a CT scanning device, comprising the following steps:

[0034] S100, The data acquired by the CT scanning equipment is processed through a low signal threshold T. LS Filter out the low-signal channels that need filtering;

[0035] S200: Obtain the range of a fixed spatial neighborhood of the low-signal channel; obtain the upper and lower threshold values ​​of the low-signal channel to be processed;

[0036] S300. Select the signal to be processed within a fixed spatial neighborhood based on the upper and lower thresholds of the low signal channel.

[0037] S400. Assign different weights to several signals to be processed according to their magnitudes to perform weighted calculations and obtain the final linear attenuation coefficient μ0 of the low signal channel.

[0038] S500 uses the final linear attenuation coefficient μ0 to filter the signal channel, obtaining suppressed low signal noise.

[0039] The technical problem this invention aims to solve is that high-noise low-signal signals encountered during CT scan data acquisition can cause significant stripe artifacts in the reconstructed image. To address this issue, this invention proposes a low-signal suppression method based on neighborhood filtering within the projection domain. This method filters out low-signal noise within the projection domain image. Specifically, a filtering threshold is determined based on the signal noise level and a noise physical model. Based on this threshold, suitable data is selected from a fixed spatial neighborhood, and weighted calculations are performed in the numerical space. This eliminates the need to calculate the spatial filtering kernel every time; the application of the signal threshold effectively excludes low signals with excessively high noise levels within the neighborhood; and the use of weighted calculations in the numerical space improves the reliability of the final calculated signal value.

[0040] This invention requires filtering the low-signal channels in the acquired projection data. Therefore, it first needs to use the low-signal threshold (TLS) to locate the low-signal channels that need filtering in the acquired data. Once a certain channel signal I is determined to be low-signal data that needs processing, then... Figure 2 The final linear attenuation coefficient μ0 is calculated using the flowchart. The following is the specific method for calculating the final linear attenuation coefficient μ0.

[0041] The fixed spatial neighborhood of this invention includes an angular neighborhood. By combining the angular neighborhood, the relevance of the data is improved, and the amount of data sampled is increased, thereby increasing the amount of filtered data and improving the reliability of data calculation. Let the fixed spatial neighborhood be l×m, where l is the width of the fixed spatial neighborhood in the channel direction, and m is the width of the fixed spatial neighborhood in the angular direction. For example... Figure 2 The fixed spatial neighborhood is l×3, and the width of the angular direction in the neighborhood space is 3.

[0042] After determining the spatial neighborhood, for each low-signal channel, the upper and lower thresholds of the filtering signal corresponding to the current channel are calculated based on the noise physical model and the signal level of the current channel.

[0043] According to Beer-Lambert's Law, I = I₀e⁻¹ -∫μx·dx The linear attenuation coefficient μ of the low-signal channel can be calculated. Here, I0 represents the X-ray intensity before incident, I represents the emitted X-ray intensity, x represents the X-ray penetration length, and dx is the derivative of x.

[0044] In addition, for each channel that requires low-signal filtering, the upper and lower bound deviations δ within the linear attenuation coefficient space are calculated based on the channel's signal level. + δ - The upper and lower thresholds are determined based on the upper and lower bound deviations, and signals are filtered within the neighborhood.

[0045] According to the principle of Poisson noise, the lower the signal level, the lower the signal-to-noise ratio. Furthermore, based on the principle of Poisson noise, the relationship between the variance and intensity of the acquired signal is: δ 2 = k·l. Where δ is the standard deviation of the acquired signal, I is the acquired signal intensity, and k is the amplification factor of the photon acquisition system. Therefore, the standard deviation of the acquired signal can be calculated as: Therefore, by combining the above relationships, the upper and lower bound deviations δ can be determined. + δ - They are respectively

[0046] Furthermore, based on the upper and lower bound deviations δ + δ - The upper threshold is determined to be μ+n·δ + The lower threshold is μ-n·δ - Where μ is the linear attenuation coefficient of the low signal channel, and n is the multiple of the upper and lower bound deviations used to determine the numerical range of the screening signal.

[0047] Finally, weighted calculations in numerical space are applied to the filtered data to improve the reliability of the calculation results.

[0048] The final linear attenuation coefficient μ0 of the low-signal channel is determined by selecting the initial offset a of the N weights at the center in the one-dimensional low-pass filter kernel. Where N (1≤N≤l·3) is the number of signals to be processed.

[0049] See appendix Figure 3 The N filtered signals to be processed are numerically sorted, and the sorted linear attenuation coefficients can be represented as μ1, μ2, ..., μ N Then, a one-dimensional low-pass filter kernel (weight values) is established in the numerical direction, with a width of l×3. The preset one-dimensional low-pass filter kernel (weight values) composed of these l×3 weight elements can be represented as w1, w2, ..., w l×3 Then, the sorted N selected signals to be processed are weighted in numerical space with the N weights at the center of the one-dimensional low-pass filter kernel to obtain the final linear attenuation coefficient. Where, μ i Let w be the i-th signal to be processed after arranging several signals in ascending order. a+i The weight of the i-th signal to be processed after arranging several signals to be processed in ascending order.

[0050] Finally, the signal channel is filtered using the final linear attenuation coefficient μ0 to obtain suppressed low signal noise.

[0051] See appendix Figure 4 , Figure 4 (a) is the reconstructed image without low-signal suppression, which is clearly contaminated by low-signal noise, exhibiting numerous horizontal stripes; while Figure 4 (b) is a reconstructed image after using the low signal-to-noise suppression method based on neighborhood filtering of the present invention, which is compared with... Figure 4 (a) Compared to the previous version, the number and intensity of horizontal stripes were significantly reduced, image noise was significantly reduced, and image quality was significantly improved.

[0052] The present invention also discloses a low signal noise suppression system for a CT scanning device, comprising a signal processing module and a filtering module connected together.

[0053] The signal processing module processes the data acquired by the CT scan equipment through a low signal threshold T. LSThe low-signal channels that need to be filtered are selected; the fixed spatial neighborhood of the low-signal channels is obtained through the signal processing module; the upper and lower thresholds of the low-signal channels to be processed are obtained; the signals to be processed are selected within the fixed spatial neighborhood based on the upper and lower thresholds of the low-signal channels; several signals to be processed are assigned different weights according to their magnitude for weighted calculation to obtain the final linear attenuation coefficient μ0 of the low-signal channels.

[0054] The signal channel is filtered using the final linear attenuation coefficient μ0 by the filtering module to obtain suppressed low signal noise.

[0055] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, wherein the steps of a low signal-to-noise suppression method are implemented when the computer program is executed by a processor.

[0056] It should be noted that the embodiments of the present invention have better implementability and are not intended to limit the present invention in any way. Any person skilled in the art may use the above-disclosed technical content to change or modify it into equivalent effective embodiments. However, any modifications or equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of the technical solution of the present invention.

Claims

1. A method for suppressing low signal noise in a CT scanning device, characterized in that, Includes the following steps: The data acquired by the CT scanning device is processed through a low signal threshold T. LS Filter out the low-signal channels that need filtering; Obtain the range of a fixed spatial neighborhood of the low-signal channel; Obtain the upper and lower threshold values ​​of the low-signal channel to be processed; The signal to be processed is selected within the fixed spatial neighborhood based on the upper and lower thresholds of the low signal channel; The signals to be processed are assigned different weights according to their magnitudes for weighted calculation to obtain the final linear attenuation coefficient of the low signal channel. ; Using the final linear attenuation coefficient The signal channel is filtered to obtain suppressed low signal noise. The fixed spatial neighborhood includes the angular neighborhood; The fixed spatial neighborhood is ;in, The width of the channel in the fixed spatial neighborhood. The width of the fixed spatial neighborhood in the angular direction; The process involves assigning different weights to several signals to be processed according to their magnitudes for weighted calculation, thereby obtaining the final linear attenuation coefficient of the low-signal channel. include: The final linear attenuation coefficient of the low signal channel By selecting the center in the one-dimensional low-pass filter kernel Starting offset of each weight To determine, ;in, The number of the signals to be processed.

2. The low signal-to-noise suppression method according to claim 1, characterized in that, The acquisition of the upper and lower thresholds of the low signal channel to be processed includes: The upper threshold is passed through The lower threshold is determined by... To determine; in, For the collected signal strength, denoted as the amplification factor of the photon acquisition system.

3. The low signal-to-noise suppression method according to claim 2, characterized in that, The upper threshold is The lower threshold is ; in, This is the linear attenuation coefficient of the low-signal channel. This is the multiple of the upper and lower bound deviations applied.

4. The low signal-to-noise suppression method according to claim 1, characterized in that, Final linear decay coefficient ; in, The first of several signals to be processed, arranged in ascending order. The signal to be processed. The first of several signals to be processed, arranged in ascending order. The weights of the signals to be processed.

5. A low signal noise suppression system for a CT scanning device, characterized in that, This includes interconnected signal processing and filtering modules; The signal processing module processes the data acquired by the CT scanning device through a low signal threshold T. LS Filter out the low-signal channels that need filtering; The signal processing module obtains the range of a fixed spatial neighborhood of the low-signal channel; and obtains the upper and lower thresholds of the low-signal channel to be processed; selects the signal to be processed within the fixed spatial neighborhood based on the upper and lower thresholds of the low-signal channel; and assigns different weights to several signals to be processed according to their magnitude for weighted calculation to obtain the final linear attenuation coefficient of the low-signal channel. ; The final linear attenuation coefficient is used through the filtering module. The signal channel is filtered to obtain suppressed low signal noise. The fixed spatial neighborhood includes the angular neighborhood; The fixed spatial neighborhood is ;in, The width of the channel in the fixed spatial neighborhood. The width of the fixed spatial neighborhood in the angular direction; The process involves assigning different weights to several signals to be processed according to their magnitudes for weighted calculation, thereby obtaining the final linear attenuation coefficient of the low-signal channel. include: The final linear attenuation coefficient of the low signal channel By selecting the center in the one-dimensional low-pass filter kernel Starting offset of each weight To determine, ;in, The number of the signals to be processed.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the low signal noise suppression method according to any one of claims 1-4.

Images