Medical image lesion segmentation method and system for radiology department

Abstract

The invention relates to a medical image lesion segmentation method and system for the radiology department, and the method comprises the steps: obtaining a medical image of the radiology department; performing denoising processing on the medical image to obtain a denoised medical image; carrying out edge detection processing on the medical image subjected to noise removal to obtain an initial edge detection result; judging whether the edge pixel points meet a preset condition or not; reserving the edge pixel points meeting a preset condition; and completing segmentation of the medical image by using the edge pixel points to obtain a target image. According to the invention, after edge detection processing is carried out on the de-noised medical image, the pixel points really corresponding to the lesion edge can be effectively filtered out by judging whether the edge pixel points meet the preset conditions, the conditions of false detection and missing detection are reduced, and the accuracy of medical image segmentation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a medical image lesion segmentation method and system for radiology. Background Art

[0002] With the rapid development of medical imaging technology, radiology plays an increasingly important role in disease diagnosis and treatment. In particular, the widespread application of advanced imaging technologies such as computed tomography (CT), X-ray imaging and ultrasound enables doctors to identify and analyze lesions in patients more intuitively and efficiently. However, lesion segmentation remains a major difficulty in medical image analysis.

[0003] Traditional lesion segmentation methods mainly rely on manual annotation and empirical interpretation. On the one hand, this method has the disadvantages of strong subjectivity, low efficiency, and long time consumption; on the other hand, due to the complexity and diversity of image data, traditional methods are often difficult to achieve accurate segmentation. Summary of the invention

[0004] To solve the above problems, an object of the embodiments of the present invention is to provide a medical image lesion segmentation method and system for the radiology department.

[0005] A medical image lesion segmentation method for radiology, comprising:

[0006] Step 1: Obtain medical images from the radiology department;

[0007] Step 2: performing denoising processing on the medical image to obtain a medical image from which noise is removed;

[0008] Step 3: performing edge detection processing on the medical image from which noise has been removed to obtain an initial edge detection result;

[0009] Step 4: Determine whether the edge pixel meets the preset conditions;

[0010] Step 5: retain edge pixels that meet preset conditions;

[0011] Step 6: Use the edge pixels to complete the segmentation of the medical image to obtain the target image.

[0012] Preferably, the step 2: performing denoising on the medical image to obtain a medical image from which noise is removed, comprises:

[0013] Step 2.1: performing N-layer wavelet decomposition on the medical image to obtain wavelet coefficients of each layer;

[0014] Step 2.2: construct a threshold based on the number of layers of wavelet decomposition;

[0015] Step 2.3: Use the threshold comparison method to process the wavelet coefficients of each layer of wavelet decomposition to obtain the wavelet coefficients after threshold processing;

[0016] Step 2.4: Perform inverse wavelet transform on the wavelet coefficients after threshold processing to obtain the medical image with noise removed.

[0017] Preferably, the step 2.2: constructing a threshold based on the number of layers of wavelet decomposition includes:

[0018] Step 2.2.1: Determine the threshold value according to the number of pixels on the medical image and the number of layers of wavelet decomposition; wherein the threshold value calculation formula is:

[0019]

[0020] Among them, λ represents the threshold, δ represents the noise standard deviation, D represents the length of the medical image, M represents the width of the medical image, and N represents the number of layers of wavelet decomposition.

[0021] Preferably, the step 2.3: using a threshold comparison method to process the wavelet coefficients of each layer of wavelet decomposition to obtain wavelet coefficients after threshold processing includes:

[0022] Step 2.3.1: When the wavelet coefficient is greater than or equal to the threshold, the wavelet coefficient is processed by the first processing formula to obtain the wavelet coefficient after threshold processing; wherein the first processing formula is:

[0023]

[0024] Step 2.3.2: When the wavelet coefficient is less than the threshold, the wavelet coefficient is processed by a second processing formula to obtain the wavelet coefficient after threshold processing; wherein the second processing formula is:

[0025]

[0026] In the formula, represents the wavelet coefficient after threshold processing, a represents the adjustment coefficient, W j,k It represents the kth wavelet coefficient at the jth decomposition level, and sign represents the sign function.

[0027] Preferably, the step 4: determining whether the edge pixel point meets a preset condition includes:

[0028] Step 4.1: using a first segmentation threshold and a second segmentation threshold to divide the pixels on the medical image after noise removal into multiple classes;

[0029] Step 4.2: Calculate the distribution probability and pixel mean of each class;

[0030] Step 4.3: Construct the objective function based on the distribution probability of the pixel points and the pixel mean;

[0031] Step 4.4: Calculate the maximum value of the objective function;

[0032] Step 4.5: taking the first segmentation threshold and the second segmentation threshold corresponding to the maximum value of the objective function as the first judgment threshold and the second judgment threshold;

[0033] Step 4.6: Pixel points whose values ​​are smaller than the first judgment threshold are regarded as non-edge points, and pixel points whose values ​​are larger than the second judgment threshold are regarded as edge pixel points satisfying preset conditions.

[0034] Preferably, in step 4.2, the calculation formula for the distribution probability of the pixel point is:

[0035]

[0036] Where S represents the first segmentation threshold, T represents the second segmentation threshold, n i represents the number of pixels with pixel i, K represents the total number of pixels on the medical image after noise removal, w0 represents the distribution probability of pixels on the first class, w1 represents the distribution probability of pixels on the second class, and w2 represents the distribution probability of pixels on the third class.

[0037] Preferably, in step 4.2, the calculation formula of the pixel mean is:

[0038]

[0039] Among them, u0 represents the mean of the pixels in the first class, u1 represents the mean of the pixels in the second class, and u2 represents the mean of the pixels in the third class.

[0040] Preferably, step 4.3: constructing an objective function according to the distribution probability of the pixel points and the pixel mean value includes:

[0041] Using the formula:

[0042]

[0043] Construct the objective function; where σ A represents the objective function, u T Represents the total pixel mean.

[0044] The present invention also provides a medical image lesion segmentation system for radiology, comprising:

[0045] A medical image acquisition module, used to acquire medical images of the radiology department;

[0046] A denoising module, used for performing denoising processing on the medical image to obtain a medical image with noise removed;

[0047] An edge detection module, used for performing edge detection processing on the medical image from which noise has been removed to obtain an initial edge detection result;

[0048] An edge pixel point judgment module is used to judge whether an edge pixel point meets a preset condition;

[0049] An edge pixel retention module is used to retain edge pixels that meet preset conditions;

[0050] The segmentation module is used to segment the medical image using the edge pixels to obtain a target image.

[0051] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps in the above-mentioned medical image lesion segmentation method for the radiology department are implemented.

[0052] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0053] The present invention relates to a medical image lesion segmentation method, system and computer-readable storage medium for radiology. Compared with the prior art, the present invention can effectively filter out the pixel points that truly correspond to the lesion edge by performing edge detection processing on the denoised medical image and judging whether the edge pixel points meet the preset conditions, thereby reducing the situations of false detection and missed detection and improving the accuracy of medical image segmentation.

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0056] Figure 1 A flow chart of a medical image lesion segmentation method for radiology provided by the present invention;

[0057] Figure 2 This is the medical image edge detection result diagram provided by the present invention. DETAILED DESCRIPTION

[0058] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0059] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0060] See also Figure 1 , a medical image lesion segmentation method for radiology, comprising:

[0061] Step 1: Obtain medical images from the radiology department;

[0062] Step 2: performing denoising processing on the medical image to obtain a medical image from which noise is removed;

[0063] Furthermore, step 2 includes:

[0064] Step 2.1: performing N-layer wavelet decomposition on the medical image to obtain wavelet coefficients of each layer;

[0065] Step 2.2: construct a threshold based on the number of layers of wavelet decomposition;

[0066] The wavelet threshold denoising process requires the estimation and judgment of the threshold. The threshold represents the boundary between the effective signal and the noise. The selection of the threshold is particularly important in image denoising. The traditional threshold is a fixed threshold (such as soft threshold and hard threshold). In the process of using a fixed threshold, the threshold remains unchanged and all wavelet coefficients are processed in the same way. However, the noise coefficient of each layer of decomposition has a certain randomness. When using a fixed threshold for processing, the noise of a certain layer may not be removed enough, and the effective information of some decomposition layers may be mistakenly removed as noise, which is not conducive to retaining the effective signal of the image, and cannot remove or suppress the noise of the image.

[0067] In step 2.2, the present invention determines a threshold value according to the number of pixels on the medical image and the number of layers of wavelet decomposition; wherein the threshold value calculation formula is:

[0068]

[0069] Among them, λ represents the threshold, δ represents the noise standard deviation, D represents the length of the medical image, M represents the width of the medical image, and N represents the number of layers of wavelet decomposition.

[0070] The present invention takes into account the characteristics of effective information of the image, and according to the fact that the noise coefficient decreases with the increase of the number of decomposition layers, the selected threshold value gradually decreases, which can well remove the image noise while retaining the effective signal of the image.

[0071] Step 2.3: Use the threshold comparison method to process the wavelet coefficients of each layer of wavelet decomposition to obtain the wavelet coefficients after threshold processing;

[0072] Furthermore, step 2.3 includes:

[0073] Step 2.3.1: When the wavelet coefficient is greater than or equal to the threshold, the wavelet coefficient is processed by the first processing formula to obtain the wavelet coefficient after threshold processing; wherein the first processing formula is:

[0074]

[0075] Step 2.3.2: When the wavelet coefficient is less than the threshold, the wavelet coefficient is processed by a second processing formula to obtain the wavelet coefficient after threshold processing; wherein the second processing formula is:

[0076]

[0077] In the formula, represents the wavelet coefficient after threshold processing, a represents the adjustment coefficient, W j,k represents the kth wavelet coefficient at the jth decomposition level, and sign represents the sign function;

[0078] When the adjustment coefficient tends to 0, the improved processing function can be converted into a hard threshold function, and when the adjustment coefficient a tends to 1, the improved processing function can be close to a soft threshold function. Therefore, when the improved processing function takes different values ​​of a, the processing function can be converted between soft and hard threshold functions, so the noise on the image can be removed well.

[0079] Step 2.4: Perform inverse wavelet transform on the wavelet coefficients after threshold processing to obtain the medical image with noise removed.

[0080] Step 3: performing edge detection processing on the medical image from which noise has been removed to obtain an initial edge detection result;

[0081] In the present invention, the Sobel operator can be used to perform edge detection processing on the medical image from which noise has been removed. Since the Sobel operator is based on gradient calculation and its detection direction is fixed, when edge detection is performed on the medical image, it may cause blurring of the edge boundary, resulting in unclear edges and affecting subsequent processing. Therefore, the present invention needs to judge the edge pixels.

[0082] Step 4: Determine whether the edge pixel meets the preset conditions;

[0083] Furthermore, step 4 includes:

[0084] Step 4.1: using a first segmentation threshold and a second segmentation threshold to divide the pixels on the medical image after noise removal into multiple classes;

[0085] Step 4.2: Calculate the distribution probability and pixel mean of each class;

[0086] In step 4.2, the calculation formula of the distribution probability of the pixel point is:

[0087]

[0088] Where S represents the first segmentation threshold, T represents the second segmentation threshold, n i represents the number of pixels with pixel i, K represents the total number of pixels on the medical image after noise removal, w0 represents the distribution probability of pixels on the first class, w1 represents the distribution probability of pixels on the second class, and w2 represents the distribution probability of pixels on the third class.

[0089] The calculation formula for pixel mean is:

[0090]

[0091] Among them, u0 represents the mean of the pixels in the first class, u1 represents the mean of the pixels in the second class, and u2 represents the mean of the pixels in the third class.

[0092] Step 4.3: Construct the objective function based on the distribution probability of the pixel points and the pixel mean;

[0093] In step 4.3, the present invention may adopt the formula:

[0094]

[0095] Construct the objective function; where σ A represents the objective function, u T Represents the total pixel mean.

[0096] Step 4.4: Calculate the maximum value of the objective function;

[0097] Step 4.5: taking the first segmentation threshold and the second segmentation threshold corresponding to the maximum value of the objective function as the first judgment threshold and the second judgment threshold;

[0098] Step 4.6: Pixel points whose values ​​are smaller than the first judgment threshold are regarded as non-edge points, and pixel points whose values ​​are larger than the second judgment threshold are regarded as edge pixel points satisfying preset conditions.

[0099] like Figure 2 As shown, the present invention considers the grayscale distribution of the entire image, calculates the objective function of each class under different thresholds, and finds the threshold that maximizes the objective function (maximum pixel difference). Therefore, the first segmentation threshold and the second segmentation threshold are determined, which can reflect the optimal segmentation threshold of the image. The optimal segmentation threshold is used to determine whether the edge pixel point is a real edge pixel point, which can effectively improve the edge detection accuracy of medical images.

[0100] Step 5: retain edge pixels that meet preset conditions;

[0101] Step 6: Use the edge pixels to complete the segmentation of the medical image to obtain the target image.

[0102] The present invention performs edge detection processing on the denoised medical image and determines whether the edge pixels meet preset conditions, thereby effectively filtering out the pixels that truly correspond to the edge of the lesion, reducing false detections and missed detections, and improving the accuracy of medical image segmentation.

[0103] The present invention also provides a medical image lesion segmentation system for radiology, comprising:

[0104] A medical image acquisition module, used to acquire medical images of the radiology department;

[0105] A denoising module, used for performing denoising processing on the medical image to obtain a medical image with noise removed;

[0106] An edge detection module, used for performing edge detection processing on the medical image from which noise has been removed to obtain an initial edge detection result;

[0107] An edge pixel point judgment module is used to judge whether an edge pixel point meets a preset condition;

[0108] An edge pixel retention module is used to retain edge pixels that meet preset conditions;

[0109] The segmentation module is used to segment the medical image using the edge pixels to obtain a target image.

[0110] Compared with the prior art, the beneficial effects of the medical image lesion segmentation system for radiology provided by the present invention are the same as the beneficial effects of the medical image lesion segmentation method for radiology described in the above technical solution, which will not be repeated here.

[0111] An embodiment of the present invention also provides an electronic device, including a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor. The transceiver, the memory, and the processor are respectively connected via the bus. When the computer program is executed by the processor, each process of the embodiment of the above-mentioned medical image lesion segmentation method for the radiology department is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0112] In addition, an embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the various processes of the embodiment of the above-mentioned medical image lesion segmentation method for the radiology department are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0113] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technical solution that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A medical image lesion segmentation method for radiology, characterized in that: include: Step 1: Obtain medical images from the radiology department; Step 2: performing denoising processing on the medical image to obtain a medical image from which noise is removed; Step 3: performing edge detection processing on the medical image from which noise has been removed to obtain an initial edge detection result; Step 4: Determine whether the edge pixel meets the preset conditions; Step 5: retain edge pixels that meet preset conditions; Step 6: Use the edge pixels to complete the segmentation of the medical image to obtain the target image.

2. A medical image lesion segmentation method for radiology according to claim 1, characterized in that: The step 2: performing denoising on the medical image to obtain a medical image from which noise has been removed, comprises: Step 2.1: performing N-layer wavelet decomposition on the medical image to obtain wavelet coefficients of each layer; Step 2.2: construct a threshold based on the number of layers of wavelet decomposition; Step 2.3: Use the threshold comparison method to process the wavelet coefficients of each layer of wavelet decomposition to obtain the wavelet coefficients after threshold processing; Step 2.4: Perform inverse wavelet transform on the wavelet coefficients after threshold processing to obtain the medical image with noise removed.

3. The medical image lesion segmentation method for radiology according to claim 2, characterized in that: The step 2.2: constructing a threshold based on the number of layers of wavelet decomposition, comprises: Step 2.2.1: Determine the threshold value according to the number of pixels on the medical image and the number of layers of wavelet decomposition; wherein the threshold value calculation formula is: Among them, λ represents the threshold, δ represents the noise standard deviation, D represents the length of the medical image, M represents the width of the medical image, and N represents the number of layers of wavelet decomposition.

4. The medical image lesion segmentation method for radiology according to claim 3, characterized in that: The step 2.3: using a threshold comparison method to process the wavelet coefficients of each layer of wavelet decomposition to obtain wavelet coefficients after threshold processing, includes: Step 2.3.1: When the wavelet coefficient is greater than or equal to the threshold, the wavelet coefficient is processed by the first processing formula to obtain the wavelet coefficient after threshold processing; wherein the first processing formula is: Step 2.3.2: When the wavelet coefficient is less than the threshold, the wavelet coefficient is processed by a second processing formula to obtain the wavelet coefficient after threshold processing; wherein the second processing formula is: In the formula, represents the wavelet coefficient after threshold processing, a represents the adjustment coefficient, W j,k It represents the kth wavelet coefficient at the jth decomposition level, and sign represents the sign function.

5. The medical image lesion segmentation method for radiology according to claim 1, characterized in that: The step 4: determining whether the edge pixel point meets the preset condition includes: Step 4.1: using a first segmentation threshold and a second segmentation threshold to divide the pixels on the medical image after noise removal into multiple classes; Step 4.2: Calculate the distribution probability and pixel mean of each class; Step 4.3: Construct the objective function based on the distribution probability of the pixel points and the pixel mean; Step 4.4: Calculate the maximum value of the objective function; Step 4.5: taking the first segmentation threshold and the second segmentation threshold corresponding to the maximum value of the objective function as the first judgment threshold and the second judgment threshold; Step 4.6: Pixel points whose values ​​are smaller than the first judgment threshold are regarded as non-edge points, and pixel points whose values ​​are larger than the second judgment threshold are regarded as edge pixel points satisfying preset conditions.

6. The medical image lesion segmentation method for radiology according to claim 5, characterized in that: In step 4.2, the calculation formula of the distribution probability of the pixel point is: Where S represents the first segmentation threshold, T represents the second segmentation threshold, n i represents the number of pixels with pixel i, K represents the total number of pixels on the medical image after noise removal, w0 represents the distribution probability of pixels on the first class, w1 represents the distribution probability of pixels on the second class, and w2 represents the distribution probability of pixels on the third class.

7. The medical image lesion segmentation method for radiology according to claim 6, characterized in that: In step 4.2, the calculation formula of the pixel mean is: Among them, u0 represents the mean of the pixels in the first class, u1 represents the mean of the pixels in the second class, and u2 represents the mean of the pixels in the third class.

8. The medical image lesion segmentation method for radiology according to claim 7, characterized in that: Step 4.3: Construct the objective function based on the distribution probability of the pixel points and the pixel mean, including: Using the formula: Construct the objective function; where σ A represents the objective function, u T Represents the total pixel mean.

9. A medical image lesion segmentation system for radiology, characterized in that: include: A medical image acquisition module is used to acquire medical images of the radiology department; A denoising module, used for performing denoising processing on the medical image to obtain a medical image with noise removed; An edge detection module, used for performing edge detection processing on the medical image from which noise has been removed to obtain an initial edge detection result; An edge pixel point judgment module is used to judge whether an edge pixel point meets a preset condition; An edge pixel retention module is used to retain edge pixels that meet preset conditions; The segmentation module is used to segment the medical image using the edge pixels to obtain a target image.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a medical image lesion segmentation method for radiology department are implemented as claimed in any one of claims 1 to 8.

Images