Body part recognition system for medical image

Abstract

The invention is suitable for the technical field of medical image body part recognition, and provides a medical image body part recognition system which comprises a scanning unit, the scanning unit obtains patient information according to an interactive interface, and the scanning unit comprises a scanning bed, a rotating device, an X-ray emitting device and a detector array. Scanning the whole body of the patient through the scanning unit to obtain scanning data, and transmitting the scanning data; the image processing unit is used for receiving the scanning data and the patient information transmitted by the scanning unit, processing the scanning data, processing the scanning data into a scanning image and transmitting the scanning image; the image enhancement unit receives the scanned image and the patient information transmitted by the image processing unit, the device solves the problem that accurate observation is difficult to achieve by only observing the whole image through body part segmentation, and the purpose of accurately observing each body part through body part segmentation is achieved.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and more specifically, to a body part recognition system for medical imaging. Background Art

[0002] CT machines are widely used in medical imaging to image various parts of the body to help doctors make diagnoses and plan treatments. CT machines are able to generate high-resolution cross-sectional images that clearly display the anatomical structures of different body parts. The identification and application of CT machines in various parts of the body provide clinicians with rich imaging information, which helps improve the accuracy of diagnosis and the effectiveness of treatment.

[0003] Usually, after a patient is scanned by a CT machine, a scanned image of the patient will be obtained. Generally speaking, if the lungs and thoracic spine are examined, a specific scan of the patient's upper body will be performed, and the scanned image includes the entire upper body of the patient. If a patient needs to have his thoracic spine examined in the orthopedics department and his lungs examined in the respiratory department, when examining the lungs, the CT image of the thoracic spine will affect the specific shape of the lung image, and when examining the thoracic spine, the CT image of the lungs will affect the specific shape of the thoracic spine.

[0004] By segmenting specific organs or lesion areas, doctors can more accurately locate and evaluate the scope and nature of lesions. For example, segmenting tumor areas in CT images helps determine the size, shape, and location of the tumor. In addition, by segmenting and analyzing image data of specific areas, patients' surgical risks, treatment effects, and prognosis can be evaluated, providing strong support for clinical decision-making. Based on the segmentation results, doctors can develop personalized treatment plans to improve the targeting and effectiveness of treatment. Part segmentation has many application values ​​in CT image analysis. It not only significantly improves the accuracy of diagnosis and treatment, but also promotes the advancement of medical imaging technology and promotes the scientific nature of clinical decision-making.

[0005] Therefore, it is necessary to design a body part recognition system for medical imaging to solve the above problems. Summary of the invention

[0006] In view of the deficiencies in the prior art, an object of the present invention is to provide a body part recognition system for medical imaging that automatically segments images of different body parts through a region segmentation method to obtain segmented images of different body parts, thereby facilitating doctors' examinations.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A body part recognition system for medical imaging, comprising:

[0009] A scanning unit, wherein the scanning unit obtains patient information according to an interactive interface, and the scanning unit includes a scanning bed, a rotating device, an X-ray emitting device, and a detector array. The scanning unit scans the entire body of the patient, obtains scanning data, and transmits the scanning data;

[0010] An image processing unit, which receives the scan data and patient information transmitted by the scanning unit, processes the scan data, processes the scan data into a scan image, and transmits the scan image;

[0011] An image enhancement unit, which receives the scanned image and patient information transmitted by the image processing unit, further processes the contrast and details of the scanned image, and outputs enhanced scanned image data;

[0012] An image segmentation unit, which receives the enhanced scanned image data and patient information transmitted by the image enhancement unit, segments the enhanced scanned image data according to an image segmentation method to obtain a segmented image, and performs post-processing on images of different body parts to output a segmented image;

[0013] The characterization unit receives the scanned image data and the patient information output by the image enhancement unit, receives the segmented image and the patient information transmitted by the image segmentation unit, and characterizes the scanned image data and the segmented image.

[0014] The present invention is further configured as follows: the scanning unit realizes a whole-body scan of the patient, including the following steps: the patient lies on a patient platform and is positioned and fixed as needed; a doctor or technician sets scanning parameters, such as scanning range, layer thickness, exposure time, etc.; the X-ray emitting device and detector array rotate around the patient and scan from multiple angles; the detector records the attenuation data of the X-rays.

[0015] The present invention is further configured as follows: the image enhancement unit generates a CT image by using a reconstruction algorithm, uses a filter to remove noise in the detection data, uses a filter to enhance the contrast and details of the image, and normalizes the detection data to a uniform scale for subsequent processing.

[0016] By adopting the above technical solutions, filters are used to remove noise from the detection data, such as Gaussian filters or median filters, to improve the clarity and signal-to-noise ratio of the image. At the same time, contrast enhancement filters are used to enhance the contrast and details of the image, making the key structures more prominent. Finally, the detection data is normalized to a uniform scale to ensure that the image has consistent brightness and contrast under different scanning conditions, which is convenient for subsequent image processing and analysis.

[0017] The present invention is further configured as follows: the image segmentation unit adopts a region segmentation method, automatically selects a center point through an edge detection algorithm, and implements region expansion through an extended region method according to a similarity criterion to obtain segmentation and a segmented image.

[0018] The present invention is further configured as follows: the edge detection algorithm uses the Canny edge detection algorithm to detect edges in the image and selects edge points as center points.

[0019] The present invention is further configured as follows: the similarity criterion is defined based on the gray value of the scanned image data, and the difference between the gray value of the pixel and the gray value of the center point is less than a set gray value threshold;

[0020] Among them, the gray value threshold is set according to the required precision.

[0021] The present invention is further configured as follows: the method for extending the area comprises the following steps:

[0022] S1. Set eight groups of neighborhoods, which are respectively the top, bottom, left, right, upper left, upper right, lower left, and lower right of the current pixel;

[0023] S2, starting from the center point, check whether the neighboring pixels of the current pixel meet the similarity criterion, if so, add them to the current area; according to the similarity criterion, if there is a neighboring pixel that meets the similarity criterion, execute S1; if all neighboring pixels do not meet the similarity criterion, execute S3;

[0024] S3. Stop expansion.

[0025] The present invention is further configured as follows: the image segmentation unit performs post-processing on the segmented image, comprising the following steps:

[0026] A1. Set the discard threshold;

[0027] A2, remove small areas whose area is smaller than the discard threshold;

[0028] A3. Use morphological operations (such as dilation and erosion) to smooth the boundaries of the region.

[0029] By adopting the above technical solution, the image segmentation unit traverses all areas in the segmented image and removes those small areas whose area is smaller than the discard threshold. These small areas may be noise or irrelevant background areas. Removing them helps to improve the purity and accuracy of the segmentation results. Finally, morphological operations (such as expansion and erosion) are used to smooth the boundaries of the remaining segmented areas. The expansion operation can expand the boundaries of the region outward to fill possible gaps; while the erosion operation can shrink the boundaries of the region inward to remove possible protrusions. Through these morphological operations, the boundaries of the regions can be effectively smoothed to make them more continuous and natural, thereby improving the overall quality of the segmented image.

[0030] The present invention is further configured as follows: the representation unit includes a total display component and a partial display component, and the scanned image data is represented by the total display component, and the segmented images are represented separately according to the doctor's needs by the partial display component.

[0031] The present invention is further configured as follows: the total display component and the partial display component are arranged in parallel, wherein the total display component represents patient information and scanned image data; and the partial display component displays body part names and segmented images.

[0032] By adopting the above technical solution, through the coordinated work of the overall display component and the partial display components, the characterization unit can comprehensively and meticulously characterize the scanned image data, meet the diverse needs of doctors in diagnosis and treatment planning, and improve the accuracy and efficiency of clinical decision-making.

[0033] In summary, the present application includes at least one of the following beneficial technical effects:

[0034] 1. Use filters to remove noise from the detection data, such as Gaussian filters or median filters, to improve the clarity and signal-to-noise ratio of the image. At the same time, use contrast enhancement filters to enhance the contrast and details of the image to make the key structures more prominent. Finally, normalize the detection data to a uniform scale to ensure that the image has consistent brightness and contrast under different scanning conditions, which is convenient for subsequent image processing and analysis.

[0035] 2. The image segmentation unit traverses all regions in the segmented image and removes those small regions whose area is smaller than the discard threshold. These small regions may be noise or irrelevant background regions. Removing them helps improve the purity and accuracy of the segmentation results. Finally, morphological operations (such as dilation and erosion) are used to smooth the boundaries of the remaining segmented regions. The dilation operation can expand the boundaries of the region outward to fill possible gaps; while the erosion operation can shrink the boundaries of the region inward to remove possible protrusions. Through these morphological operations, the boundaries of the regions can be effectively smoothed to make them more continuous and natural, thereby improving the overall quality of the segmented image.

[0036] 3. Through the collaborative work of the total display component and the partial display components, the characterization unit can comprehensively and meticulously characterize the scanned image data, thereby meeting the diverse needs of doctors in diagnosis and treatment planning, and significantly improving the accuracy and efficiency of clinical decision-making. First, the total display component provides a global perspective, allowing doctors to quickly understand the patient's overall condition. The total display component usually displays images at different levels in a multi-window format, such as cross-sections, coronal planes, and sagittal planes, as well as three-dimensional reconstructed images. This global display method helps doctors comprehensively evaluate the patient's anatomical structure and lesions to ensure that no important information is missed. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Schematic diagram of the modules of the body part recognition system for medical imaging of the present invention.

[0038] Figure 2 It is a schematic diagram of region division of the image segmentation method in the present invention. DETAILED DESCRIPTION

[0039] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0040] It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meanings as commonly understood by ordinary technicians in the technical field to which this application belongs.

[0041] See also Figure 1-Figure 2 , the present invention provides the following technical solutions:

[0042] For example, see Figure 1-Figure 2 ,

[0043] A body part recognition system for medical imaging includes: a scanning unit, an image processing unit, an image enhancement unit, an image segmentation unit and a representation unit; a patient is scanned by the scanning unit; the image processing unit processes the scan data to obtain a scanned image, and the image enhancement unit performs enhancement processing on the scanned image to increase its visibility; the image segmentation unit can perform segmentation according to different body parts; and the representation unit can represent the scanned image and the segmented image.

[0044] A scanning unit, which obtains patient information according to an interactive interface. The scanning unit includes a scanning bed, a rotating device, an X-ray emitting device, and a detector array. The scanning unit is used to scan the entire body of the patient, obtain scanning data, and transmit the scanning data.

[0045] An image processing unit, the image processing unit receives the scan data and patient information transmitted by the scanning unit, processes the scan data, processes the scan data into a scan image, and transmits the scan image;

[0046] An image enhancement unit, which receives the scanned image and patient information transmitted by the image processing unit, further processes the contrast and details of the scanned image, and outputs enhanced scanned image data;

[0047] An image segmentation unit, which receives the enhanced scanned image data and patient information transmitted by the image enhancement unit, segments the enhanced scanned image data according to an image segmentation method to obtain a segmented image, performs post-processing on images of different body parts, and outputs a segmented image;

[0048] The characterization unit receives the scanned image data and the patient information output by the image enhancement unit, receives the segmented image and the patient information transmitted by the image segmentation unit, and characterizes the scanned image data and the segmented image.

[0049] The scanning unit scans the patient's entire body. First, the patient is guided to the CT scanning room and lies on a dedicated patient platform, which usually has horizontal and vertical movement functions to place the patient in the best scanning position. Depending on the specific needs of the scan, the doctor or technician will use positioning markers (such as laser lines) to ensure that the patient is in the correct position, and use fixation devices (such as head frames) when necessary to keep the patient stable. Subsequently, the doctor or technician sets the scanning parameters, including scanning range, layer thickness, exposure time, etc., to ensure that high-quality image data is obtained. The X-ray emitting device and detector array rotate around the patient and scan from multiple angles. The detector records the attenuation data of the X-rays, which are then used to reconstruct high-resolution cross-sectional images, thereby achieving detailed imaging of the patient's entire body.

[0050] The scanning unit in this embodiment adopts the existing technology. Any scanning unit that can be directly or improved by those skilled in the art can be applied to the system of this embodiment. The specific structure of the scanning unit is not absolutely limited here.

[0051] Specifically, the image enhancement unit generates CT images by using a reconstruction algorithm. First, the unit receives the raw attenuation data from the detector and applies an advanced reconstruction algorithm to generate a high-resolution cross-sectional image. Subsequently, a filter is used to remove noise from the detection data, such as a Gaussian filter or a median filter, to improve the clarity and signal-to-noise ratio of the image. At the same time, a contrast enhancement filter is used to enhance the contrast and details of the image to make the key structures more prominent. Finally, the detection data is normalized to a uniform scale to ensure that the image has consistent brightness and contrast under different scanning conditions, which is convenient for subsequent image processing and analysis.

[0052] The image segmentation unit uses a region segmentation method. First, the center points are automatically selected through the edge detection algorithm. These center points are usually located at the boundary or inside the region of interest in the image. Subsequently, the region is expanded by the extended region method based on similarity criteria, such as gray value, texture or color. Specifically, starting from the initial center point, the algorithm gradually expands outward, incorporating adjacent pixels into the same region until the preset similarity conditions are met or the boundary is reached. This process ensures the continuity and consistency of the segmented region, thereby obtaining an accurate segmented image.

[0053] The similarity criterion is defined based on the gray value of the scanned image data, and the difference between the gray value of the pixel and the gray value of the center point is less than the set gray value threshold;

[0054] Among them, the gray value threshold is set according to the required precision.

[0055] It should be further explained that the edge detection algorithm uses the Canny edge detection algorithm to detect edges in the image and selects edge points as center points.

[0056] The method for extending the area includes the following steps:

[0057] S1, set eight groups of neighborhoods;

[0058] First, eight groups of neighborhoods are set for the current pixel, which are located in the eight directions of the current pixel: up, down, left, right, upper left, upper right, lower left, and lower right. These neighborhood pixels will serve as the basis for the expansion area to ensure the integrity and continuity of the segmented area.

[0059] S2, starting from the center point, check whether the neighboring pixels of the current pixel meet the similarity criterion;

[0060] Starting from the initial center point, the algorithm checks the neighboring pixels of the current pixel one by one to see if they meet the preset similarity criteria. This algorithm uses grayscale value comparison. If a neighboring pixel meets the similarity criteria, it will be added to the current area, and the neighboring pixels of the neighboring pixel will continue to be checked to achieve area expansion.

[0061] S3, stop expanding;

[0062] When all neighborhood pixels do not meet the similarity criteria, the expansion process stops. At this point, the segmented region has reached the boundary or no longer meets the similarity condition, ensuring that the boundary of the segmented region is clear and accurate. This step ensures the integrity and consistency of the segmentation results, providing a reliable basis for subsequent image analysis and processing.

[0063] Through the above steps, the extended region method can effectively realize image segmentation and ensure the continuity and consistency of the segmented region, thereby obtaining a high-quality segmented image.

[0064] In this embodiment, the image segmentation unit performs post-processing on the segmented image, including the following steps:

[0065] A1. Set the discard threshold;

[0066] First, a rejection threshold is set, which is used to determine which small areas should be removed. The size of the rejection threshold depends on the resolution of the image and the expected size of the segmented area, and is usually determined by experiments or experience.

[0067] A2, remove small areas whose area is smaller than the discard threshold;

[0068] Next, the image segmentation unit traverses all regions in the segmented image and removes those small regions whose area is smaller than the discard threshold. These small regions may be noise or irrelevant background regions, and removing them helps improve the purity and accuracy of the segmentation results.

[0069] A3, use morphological operations to smooth the boundaries of the region;

[0070] Finally, morphological operations (such as dilation and erosion) are used to smooth the boundaries of the remaining segmented regions. The dilation operation can expand the boundaries of the region outward to fill any gaps, while the erosion operation can shrink the boundaries of the region inward to remove any protrusions. Through these morphological operations, the boundaries of the regions can be effectively smoothed to make them more continuous and natural, thereby improving the overall quality of the segmented image.

[0071] In the actual diagnosis and treatment process, after obtaining the scanned image and the segmented image, they need to be represented by media facilities to ensure that the doctor has a clear understanding. Specifically, the representation unit includes a total display component and a partial display component, and the total display component and the partial display component are arranged in parallel, wherein the total display component represents the patient information and the scanned image data; the partial display component displays the name of the body part and the segmented image. Its main function is to comprehensively and carefully represent the scanned image data through these components. First, the total display component is responsible for the overall representation of the scanned image data, providing a global perspective so that the doctor can quickly understand the overall condition of the patient. The total display component usually displays images at different levels in the form of multiple windows, such as cross-sections, coronal planes and sagittal planes, as well as three-dimensional reconstructed images, to help doctors comprehensively evaluate the patient's anatomical structure and lesions;

[0072] Secondly, the partial display component characterizes the segmented images separately according to the specific needs of the doctor. The partial display component allows the doctor to select specific areas or structures for detailed viewing, such as tumors, blood vessels, bones, etc. Through the partial display component, the doctor can zoom in, rotate and adjust the contrast and brightness of the image to observe the details more clearly. In addition, the partial display component can also provide dynamic image display, such as angiography or heart movement, to help doctors diagnose and evaluate lesions more accurately.

[0073] Through the coordinated work of the total display component and the partial display components, the characterization unit can comprehensively and meticulously characterize the scanned image data, meet the diverse needs of doctors in diagnosis and treatment planning, and improve the accuracy and efficiency of clinical decision-making.

[0074] Obviously, the above-described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

Claims

1. A body part recognition system for medical imaging, characterized in that: include: A scanning unit, wherein the scanning unit obtains patient information according to an interactive interface, and the scanning unit includes a scanning bed, a rotating device, an X-ray emitting device, and a detector array. The scanning unit scans the entire body of the patient, obtains scanning data, and transmits the scanning data; An image processing unit, which receives the scan data and patient information transmitted by the scanning unit, processes the scan data, processes the scan data into a scan image, and transmits the scan image; An image enhancement unit, which receives the scanned image and patient information transmitted by the image processing unit, further processes the contrast and details of the scanned image, and outputs enhanced scanned image data; An image segmentation unit, which receives the enhanced scanned image data and patient information transmitted by the image enhancement unit, segments the enhanced scanned image data according to an image segmentation method to obtain a segmented image, and performs post-processing on images of different body parts to output a segmented image; The characterization unit receives the scanned image data and the patient information output by the image enhancement unit, receives the segmented image and the patient information transmitted by the image segmentation unit, and characterizes the scanned image data and the segmented image.

2. A body part recognition system for medical imaging according to claim 1, characterized in that: The scanning unit realizes a whole-body scan of the patient, including the following steps: the patient lies on the patient platform and is positioned and fixed as needed; the doctor or technician sets the scanning parameters, such as scanning range, layer thickness, exposure time, etc.; the X-ray emitting device and the detector array rotate around the patient and scan from multiple angles; the detector records the attenuation data of the X-rays.

3. A body part recognition system for medical imaging according to claim 1, characterized in that: The image enhancement unit generates a CT image by using a reconstruction algorithm, removes noise in the detection data by using a filter, enhances the contrast and details of the image by using a filter, and normalizes the detection data to a uniform scale for subsequent processing.

4. A body part recognition system for medical imaging according to claim 1, characterized in that: The image segmentation unit adopts a region segmentation method, automatically selects a center point through an edge detection algorithm, and implements region expansion through an expansion region method according to a similarity criterion to obtain segmentation and a segmented image.

5. A body part recognition system for medical imaging according to claim 4, characterized in that: The edge detection algorithm uses the Canny edge detection algorithm to detect edges in an image and selects edge points as center points.

6. A body part recognition system for medical imaging according to claim 4, characterized in that: The similarity criterion is defined based on the gray value of the scanned image data, and the difference between the gray value of the pixel and the gray value of the center point is less than a set gray value threshold; Among them, the gray value threshold is set according to the required precision.

7. A body part recognition system for medical imaging according to claim 4, characterized in that: The extended area method comprises the following steps: S1. Set eight groups of neighborhoods, which are respectively the top, bottom, left, right, upper left, upper right, lower left, and lower right of the current pixel; S2, starting from the center point, check whether the neighboring pixels of the current pixel meet the similarity criterion, if so, add them to the current area; according to the similarity criterion, if there is a neighboring pixel that meets the similarity criterion, execute S1; if all neighboring pixels do not meet the similarity criterion, execute S3; S3. Stop expansion.

8. A body part recognition system for medical imaging according to claim 1, characterized in that: The image segmentation unit performs post-processing on the segmented image, comprising the following steps: A1. Set the discard threshold; A2, remove small areas whose area is smaller than the discard threshold; A3. Use morphological operations (such as dilation and erosion) to smooth the boundaries of the region.

9. The body part recognition system for medical imaging according to claim 1, characterized in that: The representation unit includes a total display component and a partial display component. The scanned image data is represented by the total display component, and the segmented images are represented separately according to the doctor's needs by the partial display component.

10. A body part recognition system for medical imaging according to claim 9, characterized in that: The total display component and the partial display component are arranged in parallel, wherein the total display component represents patient information and scanned image data; and the partial display component displays body part names and segmented images.

Images