Medical image processing method and device, computer device and storage medium

Abstract

The application relates to a medical image processing method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring deep point cloud data corresponding to an MR coil, and acquiring a CT coil attenuation map corresponding to the MR coil; the deep point cloud data is determined according to coil depth data corresponding to the MR coil; the CT coil attenuation map is determined according to coil CT scanning data corresponding to the MR coil; the CT coil attenuation map is adjusted according to the deep point cloud data to determine a target attenuation map; and PET data corresponding to a target scanning object wearing the MR coil is corrected by using the target attenuation map to obtain a reconstructed PET image. The method can improve the accuracy of the reconstructed PET image.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a medical image processing method, apparatus, computer equipment, storage medium, and computer program product. Background Technology

[0002] In recent years, with the continuous development of medical technology, medical imaging systems have become increasingly mature. Common medical imaging systems include single-modal imaging systems or multi-modal imaging systems, such as positron emission tomography (PET) systems, computed tomography (CT) systems, and magnetic resonance imaging (MR) systems, or multi-modal imaging systems such as PET / CT and PET / MR.

[0003] During PET scanning, the scattering and signal attenuation of 511 keV photons are physical effects that reduce the quantitative accuracy of PET images. Related techniques often use pre-generated hardware attenuation maps for attenuation correction of PET data. However, this method is only suitable for rigid hardware components with a fixed position relative to the bed. In PET / MR scanning, the MR coil is an easily overlooked source of attenuation. To improve PET image quality, it is necessary to correct the attenuation of the MR coil. However, since the MR coil does not have a fixed position or shape, this method cannot accurately correct the attenuation of the MR coil, resulting in low accuracy of the reconstructed PET image.

[0004] Therefore, the relevant technologies suffer from low accuracy of reconstructed PET images during PET / MR scanning imaging. Summary of the Invention

[0005] Therefore, it is necessary to provide a medical image processing method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can improve the accuracy of reconstructed PET images in response to the above-mentioned technical problems.

[0006] Firstly, this application provides a medical image processing method. The method includes:

[0007] The depth point cloud data corresponding to the MR coil and the CT coil attenuation map corresponding to the MR coil are obtained; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil.

[0008] Based on the depth point cloud data, the CT coil attenuation map is adjusted to determine the target attenuation map;

[0009] The PET data corresponding to the target scanning object wearing the MR coil is attenuated by the target attenuation map to obtain the reconstructed PET image.

[0010] In one embodiment, adjusting the CT coil attenuation map based on the depth point cloud data to determine the target attenuation map includes:

[0011] Based on the depth point cloud data, the attenuation map of the CT coil is adjusted to obtain the adjusted attenuation map of the MR coil.

[0012] The target attenuation map is determined based on the adjusted coil attenuation map and the object body attenuation map corresponding to the target scanning object.

[0013] In one embodiment, adjusting the CT coil attenuation map based on the depth point cloud data to obtain the adjusted coil attenuation map corresponding to the MR coil includes:

[0014] Acquire the CT point cloud data corresponding to the MR coil; the CT point cloud data is determined based on the coil CT scan data corresponding to the MR coil.

[0015] The CT point cloud data and the depth point cloud data are registered to obtain the transformation parameters between the depth point cloud data and the CT point cloud data;

[0016] Based on the transformation parameters, the attenuation map of the CT coil is adjusted to obtain the adjusted attenuation map of the MR coil.

[0017] In one embodiment, the registration of the CT point cloud data and the depth point cloud data to obtain transformation parameters between the depth point cloud data and the CT point cloud data includes:

[0018] The CT point cloud data and the depth point cloud data are registered to obtain registered CT point cloud data; the distance between the centroid of the registered CT point cloud data and the centroid of the depth point cloud data is less than a first preset distance threshold.

[0019] The registered CT point cloud data and the depth point cloud data are iteratively registered to obtain the transformation parameters.

[0020] In one embodiment, the iterative registration of the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters includes:

[0021] The registered CT point cloud data and the depth point cloud data are registered to obtain optimized registered CT point cloud data; the average distance between corresponding points between the optimized registered CT point cloud data and the depth point cloud data is less than a second preset distance threshold.

[0022] The transformation parameters are determined based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data.

[0023] In one embodiment, registering the registered CT point cloud data with the depth point cloud data to obtain optimized registered CT point cloud data includes:

[0024] In the depth point cloud data, the points corresponding to each point in the registered CT point cloud data are determined to obtain the target corresponding point set between the registered CT point cloud data and the depth point cloud data; the distance between each pair of corresponding points in the target corresponding set satisfies a preset distance condition.

[0025] Based on the target point set, determine the current rotation parameters and the current translation parameters;

[0026] The registered CT point cloud data is adjusted according to the current rotation parameters and the current translation parameters to obtain the adjusted CT point cloud data.

[0027] If the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is less than the second preset distance threshold, the adjusted CT point cloud data is used as the optimized and registered CT point cloud data.

[0028] In one embodiment, determining the points corresponding to each point in the registered CT point cloud data within the depth point cloud data, to obtain the target corresponding point set between the registered CT point cloud data and the depth point cloud data, includes:

[0029] The registered CT point cloud data is downsampled in layers to obtain downsampled CT point cloud data; the distribution ratio of points between each layer of point cloud data in the downsampled CT point cloud data is the same as the distribution ratio of points between each layer of point cloud data in the registered CT point cloud data.

[0030] In the depth point cloud data, the points corresponding to each point in the downsampled CT point cloud data are determined to obtain the target corresponding point set.

[0031] In one embodiment, the registered CT point cloud data includes first CT point cloud data and second CT point cloud data; the first CT point cloud data and the second CT point cloud data are point cloud data on both sides of the central axis of the MR coil in the registered CT point cloud data; the optimized registered CT point cloud data includes first target CT point cloud data and second target CT point cloud data; the step of registering the registered CT point cloud data with the depth point cloud data to obtain the optimized registered CT point cloud data includes:

[0032] By using the iterative nearest point algorithm, the first CT point cloud data and the depth point cloud data are registered to obtain the first target CT point cloud data;

[0033] Furthermore, by using an iterative nearest-point algorithm, the second CT point cloud data and the depth point cloud data are registered to obtain the second target CT point cloud data;

[0034] The step of determining the transformation parameters based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data includes:

[0035] The transformation parameters are determined based on the first rotation parameters and the first translation parameters corresponding to the first target CT point cloud data, and the second rotation parameters and the second translation parameters corresponding to the second target CT point cloud data.

[0036] Secondly, this application also provides a medical image processing apparatus. The apparatus includes:

[0037] The acquisition module is used to acquire depth point cloud data corresponding to the MR coil and to acquire CT coil attenuation map corresponding to the MR coil; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil.

[0038] The adjustment module is used to adjust the CT coil attenuation map based on the depth point cloud data to determine the target attenuation map;

[0039] The correction module is used to perform attenuation correction on the PET data corresponding to the target scanning object wearing the MR coil using the target attenuation map, so as to obtain the reconstructed PET image.

[0040] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0041] The depth point cloud data corresponding to the MR coil and the CT coil attenuation map corresponding to the MR coil are obtained; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil.

[0042] Based on the depth point cloud data, the CT coil attenuation map is adjusted to determine the target attenuation map;

[0043] The PET data corresponding to the target scanning object wearing the MR coil is attenuated by the target attenuation map to obtain the reconstructed PET image.

[0044] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0045] The depth point cloud data corresponding to the MR coil and the CT coil attenuation map corresponding to the MR coil are obtained; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil.

[0046] Based on the depth point cloud data, the CT coil attenuation map is adjusted to determine the target attenuation map;

[0047] The PET data corresponding to the target scanning object wearing the MR coil is attenuated by the target attenuation map to obtain the reconstructed PET image.

[0048] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0049] The depth point cloud data corresponding to the MR coil and the CT coil attenuation map corresponding to the MR coil are obtained; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil.

[0050] Based on the depth point cloud data, the CT coil attenuation map is adjusted to determine the target attenuation map;

[0051] The PET data corresponding to the target scanning object wearing the MR coil is attenuated by the target attenuation map to obtain the reconstructed PET image.

[0052] The aforementioned medical image processing method, apparatus, computer equipment, storage medium, and computer program product acquire depth point cloud data corresponding to the MR coil and CT coil attenuation map corresponding to the MR coil. The depth point cloud data is determined based on the coil depth data corresponding to the MR coil. The CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil. The CT coil attenuation map is adjusted based on the depth point cloud data to determine the target attenuation map. The PET data corresponding to the target scanning object wearing the MR coil is attenuated using the target attenuation map to obtain the reconstructed PET image.

[0053] Thus, by using the depth point cloud data corresponding to the MR coil, three-dimensional surface information characterizing the shape and position of the MR coil can be obtained. Therefore, by adjusting the attenuation map of the CT coil corresponding to the MR coil using the depth point cloud data, the resulting target attenuation map can be adapted to the shape and position of the MR coil in PET / MR scanning imaging of a target scanning object wearing an MR coil. This allows for accurate correction of attenuation caused by MR coils with non-fixed positions and shapes during the attenuation correction process of the PET data corresponding to the target scanning object based on the target attenuation map. This results in a more accurate reconstructed PET image based on the corrected PET data. This solves the problem of poor correction accuracy caused by using a fixed attenuation map in PET / MR scanning imaging due to the non-fixed position and shape of the MR coil during the scanning task. Consequently, it improves the quantitative accuracy of PET in PET / MR scanning imaging of target scanning objects wearing MR coils, making the reconstructed PET image more accurate. Attached Figure Description

[0054] Figure 1 This is a flowchart illustrating a medical image processing method in one embodiment;

[0055] Figure 2 This is a flowchart illustrating the steps for obtaining transformation parameters in one embodiment;

[0056] Figure 3 This is a flowchart illustrating the steps for obtaining optimized and registered CT point cloud data in one embodiment.

[0057] Figure 4 This is a flowchart illustrating a medical image processing method in another embodiment;

[0058] Figure 5 This is a flowchart of an attenuation correction method for a target scanning object wearing an MR coil, as described in one embodiment.

[0059] Figure 6 This is a structural block diagram of a medical image processing device in one embodiment;

[0060] Figure 7 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0062] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. 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.

[0063] This application provides a medical image processing method that can be applied to computer devices equipped with an integrated medical imaging system (including a single-modal imaging system and a multimodal hybrid imaging system). The integrated medical imaging system may include PET / MR systems, PET / CT systems, CT systems, etc., and is not limited thereto. The computer device can be implemented as part or all of a computer device through software, hardware, or a combination of both. In the following method embodiments, the execution subject is always described using a computer device as an example.

[0064] In one embodiment, such as Figure 1 As shown, a medical image processing method is provided, including the following steps:

[0065] Step S110: Obtain the depth point cloud data corresponding to the MR coil, and obtain the CT coil attenuation map corresponding to the MR coil.

[0066] The depth point cloud data is determined based on the coil depth data corresponding to the MR coil.

[0067] The coil depth data corresponding to the MR coil is determined based on the image acquired by the depth data acquisition device (e.g., a 3D depth camera) when the PET / MR system scans the target object wearing the MR coil; wherein the image includes the MR coil.

[0068] The target object to be scanned can be a living organism (such as a human or an animal).

[0069] Among them, the MR coil can be a human body coil or an animal coil.

[0070] Among them, the depth point cloud data is dense point cloud data used to characterize the surface information of the MR coil, which is obtained based on the coil depth data.

[0071] Among them, the CT coil attenuation map can be the template attenuation map corresponding to the MR coil.

[0072] The CT coil attenuation map was determined based on the CT scan data of the corresponding MR coil.

[0073] In practice, when the PET / MR system scans a target object wearing an MR coil, the depth data acquisition device captures an image of the target object, obtaining an image containing the MR coil. Then, the computer can segment the MR coil from this image to obtain the coil depth data, and convert the coil depth data into three-dimensional point cloud data to obtain dense point cloud data characterizing the surface information of the MR coil, which serves as the depth point cloud data corresponding to the MR coil.

[0074] Simultaneously, the computer equipment can also acquire coil CT scan data obtained by scanning the MR coil using the CT module in the PET / CT system or a standalone CT system, perform image reconstruction on the coil CT scan data to obtain the coil CT image corresponding to the MR coil, and generate a coil template attenuation map corresponding to the MR coil based on the coil CT image, which serves as the CT coil attenuation map.

[0075] Step S120: Adjust the CT coil attenuation map based on the depth point cloud data to determine the target attenuation map.

[0076] In practice, the computer equipment can adjust the CT coil attenuation map corresponding to the MR coil based on the depth point cloud data corresponding to the MR coil, so as to obtain the target attenuation map corresponding to the target scanning object wearing the MR coil.

[0077] Step S130: Attenuation correction is performed on the PET data corresponding to the target scanning object wearing the MR coil using the target attenuation map to obtain the reconstructed PET image.

[0078] The PET data was obtained by scanning a target object wearing an MR coil using the PET module in a PET / MR system.

[0079] In practice, when the PET / MR system scans a target object wearing an MR coil, the depth data acquisition device takes a picture of the target object to obtain an image containing the MR coil. During this process, the PET module in the PET / MR system can scan the target object wearing an MR coil so that the computer device can obtain the PET data corresponding to the target object wearing an MR coil.

[0080] Computer equipment can perform attenuation correction on the above PET data using the target attenuation map, and obtain the reconstructed PET image corresponding to the target scanning object wearing an MR coil using the corrected PET data.

[0081] In the aforementioned medical image processing method, depth point cloud data corresponding to the MR coil and CT coil attenuation map corresponding to the MR coil are acquired. The depth point cloud data is determined based on the coil depth data corresponding to the MR coil. The CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil. The CT coil attenuation map is adjusted based on the depth point cloud data to determine the target attenuation map. The PET data corresponding to the target scanning object wearing the MR coil is attenuated using the target attenuation map to obtain the reconstructed PET image.

[0082] Thus, by using the depth point cloud data corresponding to the MR coil, three-dimensional surface information characterizing the shape and position of the MR coil can be obtained. Therefore, by adjusting the attenuation map of the CT coil corresponding to the MR coil using the depth point cloud data, the resulting target attenuation map can be adapted to the shape and position of the MR coil in PET / MR scanning imaging of a target scanning object wearing an MR coil. This allows for accurate correction of attenuation caused by MR coils with non-fixed positions and shapes during the attenuation correction process of the PET data corresponding to the target scanning object based on the target attenuation map. This results in a more accurate reconstructed PET image based on the corrected PET data. This solves the problem of poor correction accuracy caused by using a fixed attenuation map in PET / MR scanning imaging due to the non-fixed position and shape of the MR coil during the scanning task. Consequently, it improves the quantitative accuracy of PET in PET / MR scanning imaging of target scanning objects wearing MR coils, making the reconstructed PET image more accurate.

[0083] In one embodiment, adjusting the CT coil attenuation map based on depth point cloud data to determine the target attenuation map includes: adjusting the CT coil attenuation map based on depth point cloud data to obtain the adjusted coil attenuation map corresponding to the MR coil; and determining the target attenuation map based on the adjusted coil attenuation map and the object body attenuation map corresponding to the target scan object.

[0084] Among them, the object body attenuation map is the attenuation map corresponding to the target scan object body.

[0085] Among them, the object body attenuation map is determined based on the object body MR image corresponding to the target scan object.

[0086] Among them, the MR image of the object itself is the image obtained by performing magnetic resonance scanning on the target object wearing an MR coil using the MR module in the PET / MR system.

[0087] In practice, during the process of adjusting the CT coil attenuation map based on the depth point cloud data and determining the target attenuation map, the computer equipment can adjust the CT coil attenuation map corresponding to the MR coil based on the depth point cloud data corresponding to the MR coil, thereby obtaining the adjusted coil attenuation map corresponding to the MR coil.

[0088] Furthermore, while the PET / MR system scans a target object wearing an MR coil, the depth data acquisition device captures an image of the target object containing the MR coil. Simultaneously, the MR module within the PET / MR system performs a magnetic resonance scan on the target object, allowing the computer to acquire the corresponding MR image of the object itself. In this process, the PET module and MR module in the PET / MR system scan the target object wearing the MR coil synchronously.

[0089] In this way, the computer device can determine the object attenuation map corresponding to the target scanned object based on the object's MR image. Then, the computer device can obtain the target attenuation map corresponding to the target scanned object wearing the MR coil based on the adjusted coil attenuation map and the object's attenuation map. Specifically, the computer device can add the adjusted coil attenuation map and the object's attenuation map to obtain the target attenuation map.

[0090] The technical solution of this embodiment adjusts the CT coil attenuation map based on depth point cloud data to obtain an adjusted coil attenuation map corresponding to the MR coil. Based on the adjusted coil attenuation map and the object body attenuation map corresponding to the target scanned object, a target attenuation map is determined. Thus, by adjusting the CT coil attenuation map corresponding to the MR coil using the depth point cloud data, the resulting adjusted coil attenuation map can be adapted to the shape and position of the MR coil in PET / MR scanning imaging of a target scanned object equipped with an MR coil, enabling highly accurate attenuation correction of the PET data corresponding to the target scanned object. Therefore, the target attenuation map obtained from the adjusted coil attenuation map and the object body attenuation map can not only correct attenuation caused by the target scanned object itself but also accurately correct attenuation caused by MR coils with non-fixed positions and shapes, resulting in a more accurate reconstructed PET image.

[0091] In one embodiment, adjusting the CT coil attenuation map based on depth point cloud data to obtain the adjusted coil attenuation map corresponding to the MR coil includes: acquiring CT point cloud data corresponding to the MR coil; registering the CT point cloud data and depth point cloud data to obtain transformation parameters between the depth point cloud data and the CT point cloud data; and adjusting the CT coil attenuation map based on the transformation parameters to obtain the adjusted coil attenuation map corresponding to the MR coil.

[0092] The CT point cloud data is determined based on the coil CT scan data corresponding to the MR coil.

[0093] Among them, CT point cloud data is dense point cloud data obtained from coil CT scan data to characterize the surface information of MR coils.

[0094] In practice, during the process of adjusting the CT coil attenuation map to obtain the adjusted coil attenuation map corresponding to the MR coil, the computer device can acquire the CT point cloud data corresponding to the MR coil. Specifically, after the computer device performs image reconstruction on the coil CT scan data to obtain the coil CT image corresponding to the MR coil, it can obtain dense point cloud data representing the surface information of the MR coil based on the coil CT image, which serves as the CT point cloud data corresponding to the MR coil.

[0095] Then, the computer equipment can register the CT point cloud data with the depth point cloud data to obtain the transformation parameters between the two. Specifically, the computer equipment can use the depth point cloud data as the target point cloud data, register the depth point cloud data with the CT point cloud data, and determine the transformation parameters between the depth point cloud data and the CT point cloud data when the computer equipment determines that the overlap between the transformed CT point cloud data and the depth point cloud data meets the preset conditions.

[0096] In this way, the computer equipment can adjust the CT coil attenuation map by transforming the parameters between the depth point cloud data and the CT point cloud data, and obtain the adjusted coil attenuation map corresponding to the MR coil.

[0097] The technical solution of this embodiment involves acquiring CT point cloud data corresponding to the MR coil; the CT point cloud data is determined based on the coil CT scan data corresponding to the MR coil; the CT point cloud data and the depth point cloud data are registered to obtain the transformation parameters between the depth point cloud data and the CT point cloud data; and the CT coil attenuation map is adjusted according to the transformation parameters to obtain the adjusted coil attenuation map corresponding to the MR coil.

[0098] Thus, since the depth point cloud data corresponding to the MR coil can acquire three-dimensional surface information characterizing the shape and position of the MR coil, by registering the depth point cloud data corresponding to the MR coil and the CT point cloud data, the transformation parameters between the two are obtained. These transformation parameters include the transformation parameters corresponding to the shape and position of the MR coil. Therefore, the adjusted coil attenuation map obtained by adjusting the CT coil attenuation map corresponding to the MR coil through these transformation parameters can be adapted to the shape and position of the MR coil in PET / MR scanning imaging of a target scanning object wearing an MR coil, so as to perform high-accuracy attenuation correction on the PET data corresponding to the target scanning object.

[0099] In one embodiment, such as Figure 2 As shown, CT point cloud data and depth point cloud data are registered to obtain transformation parameters between the depth point cloud data and CT point cloud data, including:

[0100] Step S210: Register the CT point cloud data and the depth point cloud data to obtain the registered CT point cloud data.

[0101] Among them, the distance between the centroid of the registered CT point cloud data and the centroid of the depth point cloud data is less than the first preset distance threshold.

[0102] In practice, the computer device can register CT point cloud data and depth point cloud data to obtain registered CT point cloud data. Specifically, the computer device can use depth point cloud data as the target point cloud data and CT point cloud data as the source point cloud data, performing rigid registration to roughly align the centroids of the CT and depth point cloud data. When the distance between the centroids of the transformed CT and depth point cloud data is less than a first preset distance threshold, the computer device can use the transformed CT point cloud data as the registered CT point cloud data. At this point, the depth point cloud data and the registered CT point cloud data are aligned in terms of translation.

[0103] Step S220: Iteratively register the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters.

[0104] The transformation parameters are used to characterize the pose of the MR coil in the depth point cloud data, relative to the change in the pose of the MR coil in the CT point cloud data.

[0105] In practice, since the PET / MR scan corresponding to the depth point cloud data and the CT scan corresponding to the CT point cloud data belong to different scanning sessions, the pose parameters such as the curvature, position, and shape of the MR coil in the PET / MR scan imaging and the MR coil in the CT scan imaging are different. Therefore, after rigid registration (i.e., pre-registration) of the CT point cloud data and the depth point cloud data, the computer equipment can also perform iterative registration between the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters between the depth point cloud data and the CT point cloud data. Specifically, the computer equipment can use the depth point cloud data as the target point cloud data and the registered CT point cloud data as the source point cloud data, and perform iterative fine registration between the registered CT point cloud data and the depth point cloud data to match the pose parameters of the MR coil in the CT scan imaging with the pose parameters of the MR coil in the PET / MR scan imaging, ultimately obtaining the transformation parameters between the depth point cloud data and the CT point cloud data.

[0106] The technical solution of this embodiment involves registering CT point cloud data and depth point cloud data to obtain registered CT point cloud data. The distance between the centroid of the registered CT point cloud data and the centroid of the depth point cloud data is less than a first preset distance threshold. Iterative registration is then performed on the registered CT point cloud data and depth point cloud data to obtain transformation parameters. Since the pose parameters corresponding to the MR coils in the CT point cloud data differ from those in the depth point cloud data, by performing more refined iterative registration on the registered CT point cloud data and depth point cloud data, the pose parameters of the MR coils in the CT point cloud data can be more accurately matched with those in the depth point cloud data, thus allowing for a more precise determination of the transformation parameters between the CT point cloud data and the depth point cloud data.

[0107] In one embodiment, iterative registration is performed on the registered CT point cloud data and depth point cloud data to obtain transformation parameters, including: registering the registered CT point cloud data and depth point cloud data to obtain optimized registered CT point cloud data; the average distance between corresponding points between the optimized registered CT point cloud data and depth point cloud data is less than a second preset distance threshold; and determining the transformation parameters based on the rotation parameters and translation parameters corresponding to the optimized registered CT point cloud data.

[0108] Among them, the average distance between corresponding points in the optimized registered CT point cloud data and the depth point cloud data is the root mean square error between the optimized registered CT point cloud data and the depth point cloud data.

[0109] The transformation parameters may include depth point cloud data relative to CT point cloud data, bending angle change data of MR coil wings, translation change data, rotation change data, etc.

[0110] In the specific implementation, during the process of iteratively registering the registered CT point cloud data and the depth point cloud data to obtain transformation parameters, the computer device can use the iterative nearest point algorithm to iteratively register the registered CT point cloud data and the depth point cloud data. When the average distance between corresponding points between the registered CT point cloud data and the depth point cloud data after transformation is less than or equal to a second preset distance threshold, the registered CT point cloud data after transformation is used as the optimized registered CT point cloud data.

[0111] Thus, by iterating the nearest point algorithm, the registration accuracy between the registered CT point cloud data and the depth point cloud data can be improved through continuous iteration, resulting in a higher degree of matching between the finally optimized registered CT point cloud data and the depth point cloud data.

[0112] Therefore, the computer equipment can determine the transformation parameters between the depth point cloud data and the CT point cloud data based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data. In practical applications, during the registration process between the registered CT point cloud data and the depth point cloud data, the computer equipment can use the depth point cloud data as the target point cloud data and the registered CT point cloud data as the source point cloud data.

[0113] The technical solution of this embodiment involves registering the registered CT point cloud data with the depth point cloud data to obtain optimized registered CT point cloud data. The average distance between corresponding points in the optimized registered CT point cloud data and the depth point cloud data is less than a second preset distance threshold. Transformation parameters are determined based on the rotation and translation parameters corresponding to the optimized registered CT point cloud data. Thus, by determining the transformation parameters using the rotation and translation parameters corresponding to the optimized registered CT point cloud data, and given that the average distance between corresponding points in the optimized registered CT point cloud data and the depth point cloud data is less than the second preset distance threshold, the transformation parameters between the depth point cloud data and the CT point cloud data can be accurately determined based on the optimized registered CT point cloud data, which has a higher matching degree with the depth point cloud data.

[0114] In one embodiment, such as Figure 3As shown, the registered CT point cloud data and depth point cloud data are registered to obtain optimized registered CT point cloud data, including:

[0115] Step S310: In the depth point cloud data, determine the points corresponding to each point in the registered CT point cloud data, and obtain the target corresponding point set between the registered CT point cloud data and the depth point cloud data.

[0116] Among them, the distance between each pair of points in the target corresponding point set satisfies the preset distance condition.

[0117] In practice, during the process of registering the registered CT point cloud data and the depth point cloud data to obtain the optimized registered CT point cloud data, the computer device can first determine the points in the depth point cloud data that correspond to each point in the registered CT point cloud data, thereby obtaining the target corresponding point set between the registered CT point cloud data and the depth point cloud data. The distance (e.g., Euclidean distance) between each pair of corresponding points in the target corresponding point set satisfies the preset distance condition.

[0118] In practical applications, if a certain point to be matched in the depth point cloud data has multiple corresponding points in the registered CT point cloud data, the corresponding point with the shortest Euclidean distance to the point to be matched is retained and added to the target corresponding point set together with the point to be matched in the depth point cloud data, forming a one-to-one corresponding point pair in the target corresponding point set.

[0119] Step S320: Determine the current rotation parameters and current translation parameters based on the target point set.

[0120] In practice, the computer device can determine the rotation matrix and translation vector that minimize the error function value between the registered CT point cloud data and the depth point cloud data based on the target corresponding point set, and obtain the current rotation parameters and the current translation parameters.

[0121] Step S330: Adjust the registered CT point cloud data according to the current rotation parameters and the current translation parameters to obtain the adjusted CT point cloud data.

[0122] Step S340: If the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is less than or equal to the second preset distance threshold, the adjusted CT point cloud data is used as the optimized and registered CT point cloud data.

[0123] In a specific implementation, if the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is greater than the second preset distance threshold, the computer device can use the adjusted CT point cloud data as the registered CT point cloud data and return to the step of determining the points corresponding to each point in the registered CT point cloud data in the depth point cloud data, until the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is less than or equal to the second preset distance threshold, or the number of iterations meets the preset number of iterations threshold.

[0124] Furthermore, during the iterative registration process, if the average distance of the corresponding points in the current iteration is greater than the average distance of the corresponding points in the previous iteration, the rotation parameters corresponding to the current iteration are discarded, and a smaller rotation angle or a change in the rotation direction is applied to obtain new rotation parameters. The registered CT point cloud data is then adjusted according to the new rotation parameters to obtain the adjusted CT point cloud data.

[0125] Furthermore, during the iterative registration process, if the average distance of the corresponding points continues to decrease, the rotation direction in the current rotation parameters is maintained, and the registered CT point cloud data is adjusted until the average distance of the corresponding points increases. Then, it is determined that the average distance of the corresponding points converged to the minimum value in the previous iterative registration process before the average distance of the corresponding points increased, and the iterative registration ends.

[0126] The technical solution of this embodiment involves determining the points corresponding to each point in the registered CT point cloud data from the depth point cloud data, thereby obtaining a target corresponding point set between the registered CT point cloud data and the depth point cloud data; the distance between each pair of corresponding points in the target corresponding point set satisfies a preset distance condition; the current rotation parameters and the current translation parameters are determined based on the target corresponding point set; the registered CT point cloud data is adjusted based on the current rotation parameters and the current translation parameters to obtain adjusted CT point cloud data; if the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is less than a second preset distance threshold, the adjusted CT point cloud data is used as the optimized registered CT point cloud data.

[0127] Thus, by using the target corresponding point set between the registered CT point cloud data and the depth point cloud data, the current rotation parameters and current translation parameters between the two are obtained, and the registered CT point cloud data is adjusted to obtain the adjusted CT point cloud data. The average distance between the corresponding points of the adjusted CT point cloud data and the depth point cloud data is used to measure the matching degree between the adjusted CT point cloud data and the depth point cloud data. Therefore, if the average distance between the corresponding points is less than a second preset distance threshold, the adjusted CT point cloud data is used as the optimized registered CT point cloud data, so as to effectively improve the matching degree between the optimized registered CT point cloud data and the depth point cloud data.

[0128] In one embodiment, determining the points corresponding to each point in the registered CT point cloud data within the depth point cloud data to obtain the target corresponding point set between the registered CT point cloud data and the depth point cloud data includes: performing layered downsampling on the registered CT point cloud data to obtain downsampled CT point cloud data; ensuring that the distribution ratio of points between each layer of point cloud data in the downsampled CT point cloud data is the same as the distribution ratio of points between each layer of point cloud data in the registered CT point cloud data; and determining the points corresponding to each point in the downsampled CT point cloud data within the depth point cloud data to obtain the target corresponding point set.

[0129] In the specific implementation, the computer device determines the points corresponding to each point in the registered CT point cloud data in the depth point cloud data, and obtains the target corresponding point set between the registered CT point cloud data and the depth point cloud data. During this process, the computer device can perform layered downsampling on the registered CT point cloud data. According to the distribution ratio of points between each layer of point cloud data in the registered CT point cloud data, points are randomly extracted from each layer of point cloud data to obtain downsampled CT point cloud data. This ensures that the distribution ratio of points in each layer of point cloud data in the downsampled CT point cloud data is the same as the distribution ratio of points in each layer of point cloud data in the registered CT point cloud data.

[0130] Then, the computer device can determine the points in the depth point cloud data that correspond one-to-one with each point in the downsampled CT point cloud data, and obtain each set of one-to-one corresponding point pairs between the depth point cloud data and the downsampled CT point cloud data. These sets of one-to-one corresponding point pairs are used as the target corresponding point set between the registered CT point cloud data and the depth point cloud data.

[0131] The technical solution of this embodiment obtains downsampled CT point cloud data by performing layered downsampling on the registered CT point cloud data; the distribution ratio of points between each layer of point cloud data in the downsampled CT point cloud data is the same as the distribution ratio of points between each layer of point cloud data in the registered CT point cloud data; in the depth point cloud data, the points corresponding to each point in the downsampled CT point cloud data are determined to obtain the target corresponding point set.

[0132] Thus, the distribution ratio of points between different layers of point cloud data in the downsampled CT point cloud data is the same as that in the registered CT point cloud data. However, the amount of data in the downsampled CT point cloud data is less than that in the registered CT point cloud data. By determining the points in the depth point cloud data that correspond to each point in the downsampled CT point cloud data, we can not only accurately characterize the target corresponding point set between the registered CT point cloud data and the depth point cloud data, but also calculate rotation and translation parameters based on the reduced target corresponding point set during the registration process, thereby improving registration efficiency.

[0133] In one embodiment, the registered CT point cloud data includes first CT point cloud data and second CT point cloud data; the first CT point cloud data and second CT point cloud data are point cloud data on both sides of the central axis of the MR coil in the registered CT point cloud data; the optimized registered CT point cloud data includes first target CT point cloud data and second target CT point cloud data.

[0134] Among them, the MR coil can be a dual-sided, multi-channel surface coil.

[0135] The registered CT point cloud data and depth point cloud data are registered to obtain optimized registered CT point cloud data, including: registering the first CT point cloud data and depth point cloud data using an iterative nearest point algorithm to obtain the first target CT point cloud data; and registering the second CT point cloud data and depth point cloud data using an iterative nearest point algorithm to obtain the second target CT point cloud data.

[0136] Based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data, the transformation parameters are determined, including: based on the first rotation and first translation parameters corresponding to the first target CT point cloud data, and the second rotation and second translation parameters corresponding to the second target CT point cloud data, the transformation parameters are determined.

[0137] In specific implementation, during the registration of the registered CT point cloud data and depth point cloud data to obtain optimized registered CT point cloud data, the computer equipment can divide the registered CT point cloud data into two sides along the central axis of the MR coil, obtaining first CT point cloud data and second CT point cloud data. These are respectively used as the first CT point cloud data corresponding to the first MR coil wing and the second CT point cloud data corresponding to the second MR coil wing. The first and second MR coil wings are MR coil wings obtained by dividing the MR coil into two sides along the central axis of the MR coil.

[0138] Then, the computer device can use the iterative nearest point algorithm to register the first CT point cloud data and the second CT point cloud data with the depth point cloud data respectively, to obtain the corresponding first target CT point cloud data and the corresponding second target CT point cloud data.

[0139] Specifically, the computer device can, on one side of the central axis of the MR coil in the registered CT point cloud data, use depth point cloud data as target point cloud data and first CT point cloud data as source point cloud data. Through an iterative nearest-point algorithm, the first CT point cloud data and depth point cloud data are registered, so that the first CT point cloud data is iteratively translated and rotated around the central axis of the MR coil until the average distance between corresponding points of the transformed first CT point cloud data and depth point cloud data is less than or equal to a second preset distance threshold. Then, the transformed first CT point cloud data is used as the first target CT point cloud data in the optimized registered CT point cloud data.

[0140] Similarly, the computer equipment can also use the depth point cloud data as the target point cloud data and the second CT point cloud data as the source point cloud data on the other side of the central axis of the MR coil in the registered CT point cloud data. By using the iterative nearest point algorithm, the second CT point cloud data and the depth point cloud data are registered, and finally the second target CT point cloud data in the optimized registered CT point cloud data is obtained.

[0141] Thus, during the process of determining transformation parameters based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data, the computer device can obtain the bending angle change data, translation change data, and rotation change data of the first MR coil wing based on the first translation and first rotation parameters corresponding to the first target CT point cloud data, which serve as the first transformation parameters between the CT point cloud data and the depth point cloud data corresponding to the first MR coil wing; based on the second translation and second rotation parameters corresponding to the second target CT point cloud data, the computer device can obtain the bending angle change data, translation data, and rotation change data of the second MR coil wing, which serve as the second transformation parameters between the CT point cloud data and the depth point cloud data corresponding to the second MR coil wing; thus, the computer device can obtain the transformation parameters between the depth point cloud data and the CT point cloud data based on the first and second transformation parameters.

[0142] In this embodiment, the registered CT point cloud data includes first CT point cloud data and second CT point cloud data. The first and second CT point cloud data are point cloud data on both sides of the central axis of the MR coil in the registered CT point cloud data. The optimized registered CT point cloud data includes first target CT point cloud data and second target CT point cloud data. The first CT point cloud data and the depth point cloud data are registered using an iterative nearest-point algorithm to obtain the first target CT point cloud data. The second CT point cloud data and the depth point cloud data are registered using an iterative nearest-point algorithm to obtain the second target CT point cloud data. The transformation parameters between the CT point cloud data and the depth point cloud data are determined based on the first rotation parameter and the first translation parameter corresponding to the first target CT point cloud data, and the second rotation parameter and the second translation parameter corresponding to the second target CT point cloud data.

[0143] Thus, by performing the same registration method on the point cloud data on both sides of the central axis of the MR coil in the depth point cloud data and the CT point cloud data respectively, the first rotation parameter and the first translation parameter, as well as the second rotation parameter and the second translation parameter corresponding to both sides of the central axis of the MR coil, are obtained. Based on the first rotation parameter and the first translation parameter, as well as the second rotation parameter and the second translation parameter, the complete transformation parameters corresponding to the MR coil between the CT point cloud data and the depth point cloud data can be obtained. The complete transformation parameters can be used to adjust the CT coil attenuation map corresponding to the MR coil, so that the adjusted coil attenuation map can accurately correct the attenuation caused by the non-fixed position and shape of the MR coil in the PET / MR scan imaging of the target scan object wearing the MR coil.

[0144] In another embodiment, such as Figure 4 As shown, a medical image processing method is provided. Taking the application of this method to the aforementioned computer device as an example, the method includes the following steps:

[0145] Step S410: Obtain the depth point cloud data corresponding to the MR coil, and obtain the CT coil attenuation map corresponding to the MR coil.

[0146] The MR coil includes a rigid head coil and a flexible body coil. This type of MR coil is connected to the medical imaging system via highly flexible cables, making it difficult to repeatedly place the coil in the same position within the field of view.

[0147] In the process of scanning a target object wearing an MR coil using a PET / MR system, the depth data acquisition device captures an image of the target object to obtain an image containing the MR coil. This image may include an RGB image and the corresponding infrared image of the RGB image.

[0148] Thus, the computer device segments the MR coil in the image to obtain the coil depth data corresponding to the MR coil, and converts the coil depth data into three-dimensional point cloud data to obtain dense point cloud data characterizing the surface information of the MR coil. In the process of obtaining the depth point cloud data corresponding to the MR coil, since the image acquired by the depth data acquisition device can provide the surface information of the imaged object, the computer device can obtain the segmentation mask image corresponding to the MR coil by segmenting the infrared image by determining the region of interest corresponding to the MR coil. Then, the computer device can extract the depth data corresponding to each pixel in the segmentation mask image to obtain the coil depth data corresponding to the MR coil.

[0149] In addition, during the process of computer equipment reconstructing the coil CT scan data to obtain the coil CT image corresponding to the MR coil, the computer equipment can perform image reconstruction on the coil CT scan data to obtain the initial CT image. Since the attenuation value corresponding to the MR coil is relatively fixed, the computer equipment can set a preset attenuation threshold according to the attenuation value corresponding to the MR coil, so as to extract the surface information of the MR coil in the initial CT image according to the preset attenuation threshold and obtain the coil CT image corresponding to the MR coil.

[0150] Step S420: Obtain the CT point cloud data corresponding to the MR coil.

[0151] Step S430: Register the CT point cloud data and the depth point cloud data to obtain the registered CT point cloud data.

[0152] Step S440: Iteratively register the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters between the depth point cloud data and the CT point cloud data.

[0153] Furthermore, to verify registration accuracy, the MR coil can be placed atop a cylindrical phantom on the bed used in PET / CT imaging. The phantom itself is embedded in a polystyrene support. The support provides a custom bending angle for the MR coil to simulate different MR coil placements. In practical applications, the MR coil placement can be divided into three cases: a 0° flat angle, a 20° angle, and a 30° angle. For MR coils corresponding to different bending angles, depth point cloud data and corresponding CT point cloud data are acquired respectively. These predefined datasets are used as real data. To simulate actual conditions, the CT point cloud data for the three different scenarios are registered with the corresponding depth point cloud data with different bending angles. The root mean square error between the registered CT point cloud data and the corresponding real CT point cloud data is used to evaluate the registration accuracy.

[0154] Step S450: Adjust the CT coil attenuation diagram according to the transformation parameters to obtain the adjusted coil attenuation diagram corresponding to the MR coil.

[0155] In the process of generating the coil template attenuation map corresponding to the MR coil based on the coil CT image, the computer equipment can apply bilinear transformation to transform the attenuation value represented by the coil CT image into an attenuation coefficient of 511keV, thus obtaining the CT coil attenuation map.

[0156] Before applying a bilinear transform to convert the attenuation value represented by the coil CT image into an attenuation coefficient of 511 keV, the computer device can smooth the coil CT image using a Gaussian smoothing filter.

[0157] Step S460: Determine the target attenuation map based on the adjusted coil attenuation map and the object body attenuation map corresponding to the target scanning object.

[0158] The target attenuation map is a spatially interpolated attenuation map, and its image size matches the image size corresponding to the PET data.

[0159] Step S470: Attenuation correction is performed on the PET data corresponding to the target scanning object wearing the MR coil using the target attenuation map to obtain the reconstructed PET image.

[0160] In the process of attenuation correction of PET data corresponding to the target scanning object wearing an MR coil by the computer equipment through the target attenuation map, and obtaining the reconstructed PET image corresponding to the target scanning object wearing an MR coil through the corrected PET data, in addition to attenuation correction of PET data, the computer equipment can also perform dead time correction, scattering correction, random correction and normalization correction on PET data to obtain corrected PET data. The corrected PET data is then reconstructed using the ordinary Poisson ordered subset expectation maximization algorithm (OSEM) to obtain the reconstructed PET image.

[0161] It should be noted that the specific limitations of the above steps can be found in the specific limitations of a medical image processing method described above.

[0162] In another embodiment, such as Figure 5 The diagram shows a flowchart of an attenuation correction method for a target scanning object wearing an MR coil. In practical applications, the MR coil can be a monkey coil, and the target scanning object can be a monkey. The following explanation uses the monkey coil as the MR coil and the monkey as the target scanning object. Figure 5As shown, when a PET / MR system scans a monkey wearing a monkey coil, a 3D depth camera can capture images of the monkey, obtaining images containing the monkey coil. Then, a computer device can segment the monkey coil from this image to obtain the coil depth data, and convert this coil depth data into 3D point cloud data, obtaining dense point cloud data characterizing the surface information of the monkey coil, which serves as the depth point cloud data corresponding to the monkey coil.

[0163] In addition, the computer equipment can acquire coil CT scan data obtained from scanning the monkey coil using the CT module of a PET / CT system or a standalone CT system. It can then perform image reconstruction on the coil CT scan data to obtain the coil CT image corresponding to the monkey coil. Based on the coil CT image, it can generate a coil template attenuation map corresponding to the monkey coil, serving as the CT coil attenuation map. Simultaneously, the computer equipment can also obtain dense point cloud data characterizing the surface information of the monkey coil from the coil CT image, serving as the CT point cloud data corresponding to the monkey coil.

[0164] Then, the computer device can register the CT point cloud data and the depth point cloud data to obtain the registered CT point cloud data. Then, the computer device can use the iterative nearest point algorithm to iteratively register the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters between the depth point cloud data and the CT point cloud data. Then, the computer device can adjust the CT coil attenuation map according to the transformation parameters to obtain the adjusted coil attenuation map corresponding to the monkey coil.

[0165] Furthermore, during the scanning of a monkey wearing a monkey coil by the PET / MR system, while the 3D depth camera captures images of the monkey to obtain images containing the monkey coil, the MR module in the PET / MR system can perform magnetic resonance scanning on the monkey wearing the monkey coil so that the computer equipment can obtain the corresponding MR image of the monkey. Additionally, the PET module in the PET / MR system can scan the monkey wearing the monkey coil so that the computer equipment can obtain the corresponding PET data of the monkey wearing the monkey coil.

[0166] In this way, the computer device can determine the attenuation map of the monkey corresponding to the object body based on the MR image of the object body. Then, the computer device can add the adjusted coil attenuation map and the object body attenuation map to obtain the target attenuation map corresponding to the monkey wearing the monkey coil. The PET data mentioned above is corrected by the target attenuation map to obtain the reconstructed PET image.

[0167] It should be noted that the specific limitations of the steps in the above method can be found in the specific limitations of a medical image processing method described above, and will not be repeated here.

[0168] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0169] Based on the same inventive concept, this application also provides a medical image processing apparatus for implementing the aforementioned medical image processing method. The solution provided by this apparatus is similar to the implementation described in the above method; therefore, the specific limitations in the one or more medical image processing apparatus embodiments provided below can be found in the above-described limitations of the medical image processing method, and will not be repeated here.

[0170] In one embodiment, such as Figure 6 As shown, a medical image processing device is provided, including: an acquisition module 610, an adjustment module 620, and a correction module 630, wherein:

[0171] The acquisition module 610 is used to acquire depth point cloud data corresponding to the MR coil and to acquire CT coil attenuation map corresponding to the MR coil; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil.

[0172] The adjustment module 620 is used to adjust the CT coil attenuation map according to the depth point cloud data to determine the target attenuation map.

[0173] The correction module 630 is used to perform attenuation correction on the PET data corresponding to the target scanning object wearing the MR coil through the target attenuation map to obtain the reconstructed PET image.

[0174] In one embodiment, the adjustment module 620 is specifically used to adjust the CT coil attenuation map according to the depth point cloud data to obtain the adjusted coil attenuation map corresponding to the MR coil; and to determine the target attenuation map according to the adjusted coil attenuation map and the object body attenuation map corresponding to the target scan object.

[0175] In one embodiment, the adjustment module 620 is specifically used to acquire CT point cloud data corresponding to the MR coil; the CT point cloud data is determined based on the coil CT scan data corresponding to the MR coil; the CT point cloud data and the depth point cloud data are registered to obtain transformation parameters between the depth point cloud data and the CT point cloud data; and the CT coil attenuation map is adjusted according to the transformation parameters to obtain the adjusted coil attenuation map corresponding to the MR coil.

[0176] In one embodiment, the adjustment module 620 is specifically used to register the CT point cloud data and the depth point cloud data to obtain registered CT point cloud data; the distance between the centroid of the registered CT point cloud data and the centroid of the depth point cloud data is less than a first preset distance threshold; and to perform iterative registration of the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters.

[0177] In one embodiment, the adjustment module 620 is specifically used to register the registered CT point cloud data with the depth point cloud data to obtain optimized registered CT point cloud data; the average distance between corresponding points between the optimized registered CT point cloud data and the depth point cloud data is less than a second preset distance threshold; and the transformation parameters are determined according to the rotation parameters and translation parameters corresponding to the optimized registered CT point cloud data.

[0178] In one embodiment, the adjustment module 620 is specifically configured to: determine, in the depth point cloud data, points corresponding to each point in the registered CT point cloud data, to obtain a target corresponding point set between the registered CT point cloud data and the depth point cloud data; the distance between each pair of corresponding points in the target corresponding point set satisfies a preset distance condition; determine the current rotation parameter and the current translation parameter based on the target corresponding point set; adjust the registered CT point cloud data based on the current rotation parameter and the current translation parameter to obtain adjusted CT point cloud data; and, if the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is less than a second preset distance threshold, use the adjusted CT point cloud data as the optimized registered CT point cloud data.

[0179] In one embodiment, the adjustment module 620 is specifically used to perform layered downsampling on the registered CT point cloud data to obtain downsampled CT point cloud data; the distribution ratio of points between each layer of point cloud data in the downsampled CT point cloud data is the same as the distribution ratio of points between each layer of point cloud data in the registered CT point cloud data; in the depth point cloud data, the points corresponding to each point in the downsampled CT point cloud data are determined to obtain the target corresponding point set.

[0180] In one embodiment, the registered CT point cloud data includes first CT point cloud data and second CT point cloud data; the first CT point cloud data and the second CT point cloud data are point cloud data on both sides of the central axis of the MR coil in the registered CT point cloud data; the optimized registered CT point cloud data includes first target CT point cloud data and second target CT point cloud data; the adjustment module 620 is specifically used to register the first CT point cloud data and the depth point cloud data through an iterative nearest-point algorithm to obtain the first target CT point cloud data; and to register the second CT point cloud data and the depth point cloud data through an iterative nearest-point algorithm to obtain the second target CT point cloud data; and to determine the transformation parameters according to the first rotation parameter and the first translation parameter corresponding to the first target CT point cloud data, and the second rotation parameter and the second translation parameter corresponding to the second target CT point cloud data.

[0181] The various modules in the aforementioned medical image processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0182] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 7As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a medical image processing method. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0183] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0184] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0185] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0186] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0187] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0188] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0189] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0190] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A medical image processing method, characterized in that, The method includes: The depth point cloud data corresponding to the MR coil and the CT coil attenuation map corresponding to the MR coil are obtained; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil. The process of adjusting the CT coil attenuation map based on the depth point cloud data to determine a target attenuation map includes: adjusting the CT coil attenuation map based on the depth point cloud data to obtain an adjusted coil attenuation map corresponding to the MR coil; and determining the target attenuation map based on the adjusted coil attenuation map and the object body attenuation map corresponding to the target scan object. Specifically, this includes: acquiring the CT point cloud data corresponding to the MR coil; registering the CT point cloud data with the depth point cloud data to obtain transformation parameters between the depth point cloud data and the CT point cloud data; and adjusting the CT coil attenuation map based on the transformation parameters to obtain an adjusted coil attenuation map corresponding to the MR coil. The PET data corresponding to the target scanning object wearing the MR coil is attenuated by the target attenuation map to obtain the reconstructed PET image.

2. The method according to claim 1, characterized in that, The registration of the CT point cloud data and the depth point cloud data to obtain the transformation parameters between the depth point cloud data and the CT point cloud data includes: The CT point cloud data and the depth point cloud data are registered to obtain registered CT point cloud data; the distance between the centroid of the registered CT point cloud data and the centroid of the depth point cloud data is less than a first preset distance threshold. The registered CT point cloud data and the depth point cloud data are iteratively registered to obtain the transformation parameters.

3. The method according to claim 2, characterized in that, The iterative registration of the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters includes: The registered CT point cloud data and the depth point cloud data are registered to obtain optimized registered CT point cloud data; the average distance between corresponding points between the optimized registered CT point cloud data and the depth point cloud data is less than a second preset distance threshold. The transformation parameters are determined based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data.

4. The method according to claim 3, characterized in that, The process of registering the registered CT point cloud data with the depth point cloud data to obtain optimized registered CT point cloud data includes: In the depth point cloud data, the points corresponding to each point in the registered CT point cloud data are determined to obtain the target corresponding point set between the registered CT point cloud data and the depth point cloud data; the distance between each pair of corresponding points in the target corresponding set satisfies a preset distance condition. Based on the target point set, determine the current rotation parameters and the current translation parameters; The registered CT point cloud data is adjusted according to the current rotation parameters and the current translation parameters to obtain the adjusted CT point cloud data. If the average distance between corresponding points in the adjusted CT point cloud data and the depth point cloud data is less than the second preset distance threshold, the adjusted CT point cloud data is used as the optimized and registered CT point cloud data.

5. The method according to claim 4, characterized in that, The step of determining the points corresponding to each point in the registered CT point cloud data within the depth point cloud data, and obtaining the target corresponding point set between the registered CT point cloud data and the depth point cloud data, includes: The registered CT point cloud data is downsampled in layers to obtain downsampled CT point cloud data; the distribution ratio of points between each layer of point cloud data in the downsampled CT point cloud data is the same as the distribution ratio of points between each layer of point cloud data in the registered CT point cloud data. In the depth point cloud data, the points corresponding to each point in the downsampled CT point cloud data are determined to obtain the target corresponding point set.

6. The method according to claim 3, characterized in that, The registered CT point cloud data includes first CT point cloud data and second CT point cloud data; the first CT point cloud data and the second CT point cloud data are point cloud data on both sides of the central axis of the MR coil in the registered CT point cloud data; the optimized registered CT point cloud data includes first target CT point cloud data and second target CT point cloud data; the step of registering the registered CT point cloud data with the depth point cloud data to obtain optimized registered CT point cloud data includes: By using the iterative nearest point algorithm, the first CT point cloud data and the depth point cloud data are registered to obtain the first target CT point cloud data; Furthermore, by using an iterative nearest-point algorithm, the second CT point cloud data and the depth point cloud data are registered to obtain the second target CT point cloud data; The step of determining the transformation parameters based on the rotation and translation parameters corresponding to the optimized and registered CT point cloud data includes: The transformation parameters are determined based on the first rotation parameters and the first translation parameters corresponding to the first target CT point cloud data, and the second rotation parameters and the second translation parameters corresponding to the second target CT point cloud data.

7. A medical image processing device, characterized in that, The device includes: The acquisition module is used to acquire depth point cloud data corresponding to the MR coil and to acquire CT coil attenuation map corresponding to the MR coil; the depth point cloud data is determined based on the coil depth data corresponding to the MR coil; the CT coil attenuation map is determined based on the coil CT scan data corresponding to the MR coil. The adjustment module is used to adjust the CT coil attenuation map based on the depth point cloud data to determine the target attenuation map; The adjustment module is specifically used to adjust the CT coil attenuation map according to the depth point cloud data to obtain the adjusted coil attenuation map corresponding to the MR coil; and to determine the target attenuation map according to the adjusted coil attenuation map and the object body attenuation map corresponding to the target scan object. The adjustment module is specifically used to acquire CT point cloud data corresponding to the MR coil; register the CT point cloud data with the depth point cloud data to obtain transformation parameters between the depth point cloud data and the CT point cloud data; and adjust the CT coil attenuation map according to the transformation parameters to obtain the adjusted coil attenuation map corresponding to the MR coil. The correction module is used to perform attenuation correction on the PET data corresponding to the target scanning object wearing the MR coil using the target attenuation map, so as to obtain the reconstructed PET image.

8. The apparatus according to claim 7, characterized in that, The adjustment module is specifically used to register the CT point cloud data and the depth point cloud data to obtain registered CT point cloud data; the distance between the centroid of the registered CT point cloud data and the centroid of the depth point cloud data is less than a first preset distance threshold; and to perform iterative registration of the registered CT point cloud data and the depth point cloud data to obtain the transformation parameters.

9. The apparatus according to claim 8, characterized in that, The adjustment module is specifically used to register the registered CT point cloud data with the depth point cloud data to obtain optimized registered CT point cloud data; the average distance between corresponding points between the optimized registered CT point cloud data and the depth point cloud data is less than a second preset distance threshold; and the transformation parameters are determined according to the rotation parameters and translation parameters corresponding to the optimized registered CT point cloud data.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

Images