Medical image processing method and computer readable storage medium
By using CT value projection technology and dynamic threshold correction in medical image processing, automatic positioning and accurate extraction of intervertebral disc parameters are solved, and the problems of low manual calibration accuracy and time-consuming in the existing technology are achieved, and efficient and accurate medical image processing is achieved.
Patent Information
- Application Number
- CN201810217520.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2018-03-16
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2038-11-26
AI Technical Summary
In the existing medical image processing technology, doctors need to manually calibrate the disc position and inclination angle, resulting in low accuracy and time-consuming and labor-consuming, and the identification and positioning accuracy of automatic detection technology needs to be improved.
Through projection technology based on CT value, a two-dimensional profile image is generated and projected along the center line of the imaging tissue. Combined with the difference in CT value of the vertebrae and intervertebral disc, the position of the intervertebral disc is automatically positioned, and the error is corrected by dynamic thresholds to accurately determine the central position of the intervertebral disc. , angle and thickness.
It improves the accuracy of identification and positioning of imaging tissues, reduces the error of manual operation, realizes efficient and accurate extraction of intervertebral disc parameters, and reduces the operating burden of doctors.
Smart Images

Figure CN110269633B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical imaging detection, and in particular, to a medical image processing method and a computer-readable storage medium. Background Art
[0002] In the process of medical image processing, it is often necessary to detect certain features of the target object (or imaging tissue) from medical images obtained by imaging devices such as computed tomography (CT) or magnetic resonance imaging (MR). For example, when diagnosing cervical spondylosis, doctors need to determine the position information of the intervertebral disc, such as the center position, inclination angle, thickness, etc., based on the medical images of the human cervical spine. In the prior art, it is sometimes necessary to manually operate the image on the operation interface to obtain the above position information. Fig.11 As shown in the figure, after the doctor completes the scout image scan, he needs to manually calibrate the position and inclination angle of each intervertebral disc on the scout image of the cervical spine, and then perform an axial scan or a spiral scan to extract the slice image near each intervertebral disc for lesion analysis. The examination performed in this way will cause different errors in the manually calibrated intervertebral disc parameters due to different experiences and habits of doctors. At the same time, the manual calibration method also causes doctors to repeat a lot of operations, which is time-consuming and labor-intensive. Although some automatic detection technologies for imaging tissues have been developed in the prior art, the accuracy of their recognition and positioning needs to be further improved. Summary of the invention
[0003] In view of the deficiencies of the prior art, the present invention aims to provide a medical image processing method so as to improve the accuracy and robustness of imaging tissue identification and positioning.
[0004] The present invention discloses a medical image processing method, comprising:
[0005] reconstructing a two-dimensional cross-sectional image of the imaging tissue based on the volume image of the imaging tissue;
[0006] Projecting the CT value of the imaging tissue along the normal direction of the center line of the imaging tissue in the two-dimensional cross-sectional image; and,
[0007] The imaging tissue is positioned according to the projection result of the CT value of the imaging tissue.
[0008] Another aspect of the present invention discloses a computer-readable storage medium, which is configured to store a computer program, wherein the computer program is configured to execute the above-mentioned medical image processing method. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.
[0010] Figure 1 A schematic diagram of a CT medical imaging system provided in an embodiment of the present invention is shown;
[0011] Figure 2 A flowchart of a medical image processing method provided by an embodiment of the present invention is shown;
[0012] Figure 3 A positioning image of the cervical spine acquired according to an exemplary embodiment of the present invention is shown.
[0013] Figure 4 A volumetric image of the cervical spine acquired according to an exemplary embodiment of the present invention is shown.
[0014] Figure 5 shows a sagittal two-dimensional cross-sectional image of the cervical spine acquired according to an exemplary embodiment of the present invention;
[0015] Figure 6 shows a binarized image acquired according to an exemplary embodiment of the present invention;
[0016] Figure 7 A schematic diagram showing a CT value projection along the normal direction of the center line of the cervical vertebral body according to an exemplary embodiment of the present invention is shown;
[0017] Figure 8 A schematic diagram showing a distribution sequence of an intervertebral disc acquired according to an exemplary embodiment of the present invention;
[0018] Fig. 9 A schematic diagram showing determination of the position center, angle and thickness of an intervertebral disc according to an exemplary embodiment of the present invention is shown;
[0019] Fig.10 A schematic diagram of calibrating the position of an intervertebral disc according to an exemplary embodiment of the present invention is shown;
[0020] Fig.11 shows a coronal slice image of an intervertebral disc acquired by image reconstruction along a calibration direction according to an exemplary embodiment of the present invention;
[0021] Fig.12 The flowchart of CT imaging and positioning of the cervical spine in the prior art is shown. DETAILED DESCRIPTION
[0022] The specific embodiments of the present invention will be described below. It should be noted that in the specific description of these embodiments, in order to provide a concise description, it is impossible for this specification to provide a detailed description of all the features of the actual embodiments. These embodiments are not intended to limit the scope of the claims of the present invention. On the contrary, these embodiments are only intended to summarize the possible forms of the present invention. In fact, the subject matter of the present invention can cover various forms similar to or different from the following embodiments.
[0023] The present invention provides a medical image processing method, which can be executed in a computer unit of a CT imaging system, for example, to accurately locate an imaging tissue, for example, a vertebral body of a human body, based on scanned images.
[0024] Figure 1 FIG. 1 is a schematic diagram of a CT imaging system in an embodiment of the present invention. Figure 1 The CT imaging system shown may include a scanning unit 11, which includes a scanning frame 111, a cylindrical scanning cavity 112 formed on the scanning frame 111 to accommodate a patient, and a patient table 113 for supporting the patient. A tube 114 and a detector 115 are relatively arranged on the scanning frame 111. During the rotation of the scanning frame 111, the X-rays emitted by the tube 114 penetrate human tissues and are received by the detector 115, and the X-rays received by the detector 115 are converted into digital image signals.
[0025] The CT imaging system further includes a data acquisition unit 12, which is used to acquire the digital image signal and transmit it as CT original image data to the image reconstruction unit 131 of the computer unit 13 for image reconstruction.
[0026] The image processing unit 132 in the computer unit can further process the reconstructed image to meet the needs of medical diagnosis. In addition, the reconstructed or processed image can be output to the display unit 14 for display.
[0027] The computer unit 13 further includes a control unit 133 for controlling the working parameters, states, etc. of the scanning unit 11 and the data acquisition unit 12 .
[0028] Figure 2 A flowchart of an embodiment of the medical image processing method provided by the present invention is as follows Figure 2 As shown, the method includes the following steps S21, S23 and S25.
[0029] In step S21, a two-dimensional cross-sectional image of the imaging tissue is reconstructed based on the volume image of the imaging tissue. The volume image of the imaging tissue may include a plurality of continuously acquired slices, which may be acquired based on CT spiral scanning, CT axial scanning or other scanning methods. Before performing CT spiral scanning or circumferential scanning, a low-dose scout scan may be performed to acquire a human body plain film. The scanning region of interest may be determined by the position of the imaging tissue in the plain film, and then the volume image of the imaging tissue may be acquired by the above-mentioned three-dimensional scanning method.
[0030] In this embodiment, a two-dimensional cross-sectional image of the imaging tissue may be generated based on the position of the imaging tissue in different slices. Taking a human vertebra as an example, the two-dimensional cross-sectional image may be a sagittal image of the human vertebra.
[0031] In step S23, the CT value of the imaging tissue is projected along the normal direction of the center line of the imaging part in the two-dimensional cross-sectional image.
[0032] Specifically, in step S21, the center line of the imaging tissue may be formed based on the center position of the imaging tissue in the volume image, and a two-dimensional cross-sectional image may be generated along the center line, so that the generated two-dimensional cross-sectional image contains information of the center line.
[0033] In step S25, the imaging tissue is positioned according to the projection result of the CT value of the imaging tissue.
[0034] The present invention takes the positioning of vertebrae as an example and specifically illustrates the implementation mode of the present invention in conjunction with the accompanying drawings. Those skilled in the art may also apply the positioning method disclosed in the present invention to image processing of other vertebrae (e.g., thoracic vertebrae, lumbar vertebrae), bones and interosseous tissues, or other imaging tissues.
[0035] The above-mentioned vertebral bodies include vertebrae and intervertebral discs. Specifically, in step S25, positioning the imaging tissue according to the projection result of the CT value of the imaging tissue includes: obtaining the position of the peak value in the projection result, and positioning the center position between the positions of two adjacent peak values to the position of the intervertebral disc. Since the CT values of the vertebrae and intervertebral disc tissues are obviously different, when reflected in the image, the vertebral tissue part is brighter and the intervertebral disc tissue part is darker. Therefore, in the projection result, the peak value of the projection value is regarded as the vertebral position, and the position between two adjacent peak values is regarded as the intervertebral disc. Such projection combined with the normal direction of the vertebral centerline can quickly and accurately locate the intervertebral disc.
[0036] In order to further correct the above intervertebral disc positioning results, the determined intervertebral disc position can also be adjusted by using the conventional features of the intervertebral disc and vertebrae (such as relatively evenly spaced arrangement). Specifically, in step S25, positioning the imaging tissue according to the projection result of the CT value of the imaging tissue also includes: comparing the spacing between adjacent intervertebral disc positions with a predetermined maximum threshold and a minimum threshold, and when the spacing between two adjacent intervertebral disc positions is less than the minimum threshold, one of the two adjacent intervertebral disc positions is eliminated; when the spacing between two adjacent intervertebral disc positions is greater than the maximum threshold, a new intervertebral disc position is added between the two adjacent intervertebral disc positions. In this way, the false detection or missed detection of the intervertebral disc caused by metal or lesions can be eliminated, making the detection result more robust.
[0037] The above-mentioned maximum threshold and minimum threshold may be dynamic thresholds, which may be dynamically determined according to the median of the distances between multiple adjacent intervertebral disc positions of the current patient, for example, by adding an appropriate value to the median as the maximum threshold and reducing an appropriate value to the median as the minimum threshold.
[0038] The dynamic threshold value can also be obtained based on the cloud, for example, the dynamic threshold value is determined based on the median value of the distance between the intervertebral disc positions of multiple patients recorded in the cloud, of course, the multiple patients can be patients automatically matched with the current patient based on patient information, or patients automatically matched with the intervertebral disc of the current patient based on the shape and distance of the intervertebral disc of the patient. The cloud can perform data transmission with the image processing unit 132 of the computer unit 13 through the cloud network.
[0039] Optionally, before step S23, the method may also include: performing binary segmentation on the CT value of the imaging tissue in the two-dimensional cross-sectional image, for example, assigning 1 to a CT value greater than a specific value and assigning 0 to a CT value less than or equal to the specific value. In this way, a more accurate positioning result can be obtained with a higher projection efficiency.
[0040] Optionally, in step S25, positioning the imaging tissue according to the projection result of the CT value of the imaging tissue further includes: determining the center position of the intervertebral disc, the inclination angle of the intervertebral disc, and the thickness of the intervertebral disc according to the position of the intervertebral disc and the image features of the vertebrae adjacent thereto. Specifically, the upper and lower edges and corner positions of the two vertebrae adjacent to each intervertebral disc can be obtained by using, for example, the continuity characteristics of the image grayscale gradient information and the vertebral body edge, and then accurately determining the three-dimensional parameter information of each intervertebral disc, including the center position of the intervertebral disc, the inclination angle of the intervertebral disc (for example, the inclination angle of the intervertebral disc relative to the horizontal line), and the thickness of the intervertebral disc.
[0041] The present invention is further described below by taking CT imaging and processing of a patient's cervical spine to locate the cervical intervertebral disc as an example:
[0042] Figure 3-9 The images are obtained at different stages of CT imaging and processing of the cervical spine in the embodiment of the present invention. First, the patient's neck is scanned (also called Scout scan, plain scan, etc.) by the scanning unit of the CT imaging system to obtain the following: Figure 3 Based on the positioning image, a region of interest can be determined manually or automatically by a computer unit. In this example, the region of interest includes the patient's cervical spine.
[0043] The scanning unit 12 performs a spiral scan on the patient based on the region of interest, and the image reconstruction unit 131 in the computer unit 13 performs image reconstruction based on the spiral scan data to obtain Figure 4 Volume image shown.
[0044] Based on the volume image, the image processing unit 132 in the computer unit 13 can identify the cervical vertebra in each slice of the volume image and determine the center position of the cervical vertebra, that is, the center position of the cervical vertebra in the volume image. Based on the identification result and the determination result of the center position, the following can be reconstructed: Figure 5 The sagittal two-dimensional cross-sectional image of the cervical spine is shown in FIG. From the sagittal two-dimensional cross-sectional image, it can be clearly seen that the cervical vertebral bodies are vertically distributed on the vertebral body centerline.
[0045] The image processing unit 132 further performs binary segmentation on the cervical vertebral body in the sagittal two-dimensional cross-sectional image to obtain Figure 6 The binary image shown in the figure shows that the multiple white irregular quadrilaterals in the center of the image are cervical vertebrae, and the black gaps between the white vertebrae are cervical intervertebral discs.
[0046] like Figure 7 As shown, the image processing unit selects a target area near the center line of the vertebra in the binarized image, and the target area can at least cover the vertebrae and the intervertebral disc, and projects the binarized CT value of the cervical vertebra along the normal direction of the vertebral center line (for example, as shown by the arrow) (i.e., the accumulated value of the CT value in the normal direction). Multiple peak positions are obtained. During this process, only the peak position of the projection result at the intersection of zero value and positive value can be retained, and the center between the two nearest adjacent peak positions is preliminarily located as the intervertebral disc position. Similarly, after the positioning of all the cervical intervertebral discs in the target area is completed, it can be estimated as follows: Figure 8 The original distribution sequence of intervertebral disc positions is shown in Figure 1. The “nearest neighbor” mentioned above means that Figure 7 As shown, the current peak has two adjacent peaks, one of which is farther away and the other is closer. In this embodiment, the closer adjacent peak is used as the nearest adjacent peak of the current peak.
[0047] As mentioned above, in order to make the intervertebral disc positioning result more accurate, after all the intervertebral discs in the target area are positioned, the positioning result can be further revised. Figure 8 As shown in the figure above, the image processing unit 132 uses the CT value projection peak value in the normal direction of the center curve of the intervertebral disc as the intervertebral disc width, calculates and counts the distance between each intervertebral disc and its adjacent intervertebral disc according to the preliminary estimated position of each intervertebral disc, and processes the original distribution sequence through the cervical intervertebral disc distribution statistical algorithm to obtain the following: Figure 8The figure below shows the distribution sequence of the corrected intervertebral discs. For example, the median of the distances between multiple adjacent intervertebral discs is used as the main distance, and dynamic thresholds, i.e., the maximum threshold and the minimum threshold, are set based on the main distance. If the initial calculated distance between adjacent cervical intervertebral discs is less than the minimum threshold, the adjacent cervical intervertebral disc position (e.g. Figure 8 If one of the interference value positions in the figure above is greater than the maximum threshold, an intervertebral disc position (such as Figure 8 The missing value position in the figure above), we can get Figure 8 The distribution sequence is shown in the figure below.
[0048] Furthermore, if Fig. 9 As shown, in the binary image, the image processing unit 132 can find the two left corner points (P1, P2), the two right corner points (P3, P4) and the center point of the line connecting the two left corner points (P1, P2) and the center point of the line connecting the two right corner points (P3, P4) for each intervertebral disc by using the image grayscale gradient information, and generate a straight line from the two center points. The straight line is determined to be the center line that bisects the two vertebrae. The center position (O) of the straight line relative to the center point between the two center points is the center position of the intervertebral disc. The inclination angle of the straight line is the inclination angle of the intervertebral disc. The value obtained by calculating the sum of the shortest distances from the center position (O) of the intervertebral disc to the two vertebrae is the thickness of the intervertebral disc.
[0049] After the position of the intervertebral disc is determined, the position of the intervertebral disc is marked on the binary image by the image processing unit, for example Fig.10 The white line segment shown indicates the position of the cervical intervertebral disc and is displayed on the display unit 14 .
[0050] Based on the marked intervertebral disc position, the slice image can be reconstructed along the marked direction to obtain Fig.11 The coronal slice image of the intervertebral disc is shown for doctors to analyze the lesion.
[0051] Those skilled in the art should understand that, in the above-mentioned stages, the image processing performed on the basis of the binarized image may also be replaced by: performing the image processing directly on the two-dimensional cross-sectional image that has not been binarized.
[0052] The embodiment of the present invention further discloses a computer-readable storage medium, which can be installed in the computer unit of the above-mentioned CT imaging system and is used to store a computer program, wherein the computer program is at least configured to execute the above-mentioned medical image processing method, including executing the following steps:
[0053] reconstructing a two-dimensional cross-sectional image of the imaging tissue based on the volume image of the imaging tissue;
[0054] Projecting the CT value of the imaging tissue along the normal direction of the center line of the imaging tissue in the two-dimensional cross-sectional image; and,
[0055] The imaging tissue is positioned according to the projection result of the CT value of the imaging tissue.
[0056] Furthermore, before projecting the CT value of the imaging tissue, the computer program is also configured to: perform binary segmentation of the CT value of the imaging tissue in the two-dimensional cross-sectional image.
[0057] Further, the computer program is also configured to: determine a center line of the imaged tissue in the two-dimensional cross-sectional image based on a position of a center of the imaged tissue in the volume image.
[0058] Furthermore, the imaging tissue includes a vertebral body, which includes vertebrae and intervertebral discs. The computer program is further configured to: obtain the positions of peaks in the projection results, and locate the center position between the positions of two adjacent peaks as the intervertebral disc position.
[0059] Furthermore, the computer program is also configured to: compare the distance between adjacent intervertebral disc positions with a maximum threshold and a minimum threshold; when the distance between two adjacent intervertebral disc positions is less than the minimum threshold, remove one of the two adjacent intervertebral disc positions; when the distance between two adjacent intervertebral disc positions is greater than the maximum threshold, add a new intervertebral disc position to the two adjacent intervertebral disc positions.
[0060] Furthermore, the computer program is also configured to dynamically determine the maximum threshold and the minimum threshold according to the median value of the distances between a plurality of adjacent intervertebral disc positions.
[0061] Further, the maximum threshold is determined based on the median value of the distances between multiple adjacent intervertebral disc positions of the vertebra of the current patient, or is determined based on the median value of the distances between multiple adjacent intervertebral disc positions of the vertebra of multiple patients obtained from the cloud.
[0062] Furthermore, the computer program is also configured to determine the position center, tilt angle and thickness of the intervertebral disc according to the position of the intervertebral disc and image features of the vertebrae adjacent thereto.
[0063] The embodiments of the present invention can generate a two-dimensional cross-sectional image of the imaging tissue based on the three-dimensional reconstructed image, and based on the projection, positioning, and correction of the position information of the imaging tissue obtained along the normal direction of the center line of the imaging tissue, realize high-precision and fully automatic extraction of clinical images, and provide accurate image information for doctors to quickly analyze lesions.
[0064] The preferred embodiments of the medical image processing device and method of the present invention are described above with reference to the accompanying drawings. However, the present invention is not limited to the above exemplary embodiments and drawings. Those skilled in the art can think of various alternative, modified and revised embodiments within the scope of the technical ideas disclosed in the present application, which also fall within the scope of protection claimed in the present application.
Claims
1. A medical image processing method, characterized in that: include: reconstructing a two-dimensional cross-sectional image of the imaging tissue based on the volume image of the imaging tissue; Projecting the CT value of the imaging tissue along the normal direction of the center line of the imaging tissue in the two-dimensional cross-sectional image; as well as, The imaging tissue is positioned according to the projection result of the CT value of the imaging tissue.
2. The medical image processing method according to claim 1, characterized in that: Before projecting the CT value of the imaging tissue, the method further includes: The imaging tissue in the two-dimensional cross-sectional image is subjected to binary segmentation of CT value.
3. The medical image processing method according to claim 1, characterized in that: Reconstructing a two-dimensional cross-sectional image of the imaging tissue based on the volume image of the imaging tissue includes: A centerline of the imaged tissue is determined in the two-dimensional cross-sectional image based on the position of the center of the imaged tissue in the volume image.
4. The medical image processing method according to claim 1, characterized in that: The imaging tissue includes a vertebral body, which includes vertebrae and intervertebral discs. Positioning the imaging tissue according to the projection result of the CT value of the imaging tissue includes: obtaining the position of the peak in the projection result, and positioning the center position between the positions of two adjacent peaks as the intervertebral disc position.
5. The medical image processing method according to claim 4, characterized in that: Positioning the imaging tissue according to the projection result of the CT value of the imaging tissue also includes: The distance between adjacent intervertebral disc positions is compared with the maximum threshold and the minimum threshold. When the distance between two adjacent intervertebral disc positions is less than the minimum threshold, one of the two adjacent intervertebral disc positions is eliminated; when the distance between two adjacent intervertebral disc positions is greater than the maximum threshold, a new intervertebral disc position is added to the two adjacent intervertebral disc positions.
6. The medical image processing method according to claim 5, characterized in that: Also includes: The maximum threshold and the minimum threshold are dynamically determined according to the median value of the distances between a plurality of adjacent intervertebral disc positions.
7. The medical image processing method according to claim 6, characterized in that: The maximum threshold is determined according to a median value of the distances between a plurality of adjacent intervertebral disc positions of the vertebral body of the current patient.
8. The medical image processing method according to claim 6, characterized in that: The maximum threshold is determined according to a median value between a plurality of adjacent intervertebral disc positions of vertebral bodies of a plurality of patients obtained from the cloud.
9. The medical image processing method according to claim 4, characterized in that: Positioning the imaging tissue according to the projection result of the CT value of the imaging tissue also includes: The position center, inclination angle and thickness of the intervertebral disc are determined based on the image features of the intervertebral disc and the adjacent vertebrae.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is configured to store a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 9.
Citation Information
Patent Citations
Positioning method and positioning device for cervical vertebral body axial lines and relevant tissues of MRI (magnetic resonance imaging) images
CN103300856A
Medical image processing device and medical image processing method
CN103402435A
Method for three-dimensional reconstruction of liver computed tomography (CT) image
CN103679810A
Two-dimensional intravascular photo-acoustic image modeling and simulation method
CN104867178A
Optical tomography method based on self-adaptive voxel subdivision
CN105629652A
