A method and apparatus for generating a three-dimensional CT image

Abstract

The present specification discloses a three-dimensional CT image generation method and device, which can generate a compensation image before generating a three-dimensional CT image according to a raw image set collected by a computer tomography device, so as to avoid the problems of bar-shaped artifacts in the reconstructed three-dimensional CT image due to the small number of two-dimensional projection images in the raw image set, the low dose caused by the small number of two-dimensional images, and the low quality of the three-dimensional CT image, thereby improving the signal-to-noise ratio of the reconstructed three-dimensional CT image.

Description

Technical Field

[0001] This specification relates to the field of medical imaging, and in particular to a method and apparatus for generating three-dimensional CT images. Background Technology

[0002] Cone-beam computed tomography (CBCT) is a medical imaging technique widely used in image-guided surgery. It uses the attenuation patterns of X-rays in different tissues of the human body to create images. Compared with traditional sector-beam CT, cone-beam CT has advantages such as high X-ray utilization and fast scanning speed.

[0003] However, since X-ray radiation can affect the patient's health, although the 3D CT images generated by high-dose cone-beam CT have better image clarity, they may cause pathological changes in the patient's body. On the other hand, 3D CT images generated by low-dose cone-beam CT often have problems such as low contrast and stripe artifacts.

[0004] Therefore, how to improve the signal-to-noise ratio of 3D CT images generated from low-dose cone-beam CT is an urgent problem to be solved. Summary of the Invention

[0005] This specification provides a method and apparatus for generating three-dimensional CT images, in order to partially solve the aforementioned problems existing in the prior art.

[0006] The following technical solution is adopted in this specification:

[0007] This specification provides a method for generating three-dimensional CT images, including:

[0008] Acquire two-dimensional projection images from different acquisition angles using a computed tomography (CT) scanner and construct an original image set;

[0009] For each frame of two-dimensional projection image contained in the original image set, determine the two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image of that frame, and use it as the associated image corresponding to the two-dimensional projection image of that frame. Based on the two-dimensional projection image of that frame and the associated image, generate a two-dimensional projection image whose acquisition angle is between the two-dimensional projection image of that frame and the associated image, and use it as the first compensation image.

[0010] A three-dimensional CT image is generated based on the first compensated image and the original image set.

[0011] Optionally, based on the frame's two-dimensional projection image and the associated image, a two-dimensional projection image with an acquisition angle between the frame's two-dimensional projection image and the associated image is generated, specifically including:

[0012] Determine the angle difference between the acquisition angle corresponding to the two-dimensional projection image of the frame and the acquisition angle corresponding to the associated image;

[0013] Determine whether the angle difference exceeds a preset threshold;

[0014] If so, based on the two-dimensional projection image of the frame and the associated image, generate a two-dimensional projection image with an acquisition angle between the two-dimensional projection image of the frame and the associated image.

[0015] Optionally, based on the frame's two-dimensional projection image and the associated image, a two-dimensional projection image with an acquisition angle between the frame's two-dimensional projection image and the associated image is generated as a first compensation image, specifically including:

[0016] The two-dimensional projection image of the frame and the associated image are input into a pre-trained image conversion model to determine the amount of image change between the two-dimensional projection image of the frame and the associated image through the image conversion model.

[0017] The first compensation image is generated based on the image change, the acquisition angle corresponding to the two-dimensional projection image of the frame, and the acquisition angle corresponding to the associated image.

[0018] Optionally, training the image conversion model specifically includes:

[0019] Obtain the first sample projection image and the second sample projection image;

[0020] The first sample projection image and the second sample projection image are input into the image conversion model to obtain the image change between the first sample projection image and the second sample projection image, which is used as the image change to be optimized.

[0021] Based on the change amount of the image to be optimized, the first sample projection image is transformed to obtain the transformed projection image;

[0022] The image conversion model is trained with the optimization objective of minimizing the deviation between the converted projection image and the second sample projection image.

[0023] Optionally, before generating a three-dimensional CT image based on the first compensated image and the original image set, the method further includes:

[0024] Based on each frame of the two-dimensional projection image in the original image set, an initial three-dimensional CT image is constructed;

[0025] Using the initial three-dimensional CT image, a two-dimensional projection image with the same acquisition angle as the first compensation image is generated as the second compensation image;

[0026] Generating a 3D CT image based on the first compensated image and the original image set specifically includes:

[0027] A three-dimensional CT image is generated based on the first compensated image, the second compensated image, and the original image set.

[0028] Optionally, a three-dimensional CT image is generated based on the first compensated image, the second compensated image, and the original image set, specifically including:

[0029] Based on the number of two-dimensional projected images contained in the original image set, the weights corresponding to the first compensation image and the second compensation image are determined. The weights corresponding to the first compensation image are negatively correlated with the number of images, and the weights corresponding to the second compensation image are positively correlated with the number of images.

[0030] Based on the weights corresponding to the first compensation image and the second compensation image, the first compensation image and the second compensation image are fused to generate a fused image.

[0031] A three-dimensional CT image is generated based on the fused image and the original image set.

[0032] Optionally, the method further includes:

[0033] The three-dimensional CT image is input into a pre-trained optimization model to optimize the three-dimensional CT image, thereby obtaining an optimized three-dimensional CT image. The dose of medical radiation used by the CT device when acquiring each frame of two-dimensional projection image contained in the original image set is less than the dose of medical radiation corresponding to the optimized three-dimensional CT image.

[0034] Optionally, training the optimized model specifically includes:

[0035] A first sample 3D CT image and a second sample 3D CT image of the same area are acquired, wherein the dose of medical radiation corresponding to the first sample 3D CT image is less than the dose of medical radiation corresponding to the second sample 3D CT image.

[0036] The first sample 3D CT image is input into the optimization model to obtain the optimized first sample 3D CT image output by the optimization model.

[0037] The optimization model is trained with the goal of minimizing the deviation between the optimized first sample 3D CT image and the second sample 3D CT image.

[0038] This specification provides a device for generating three-dimensional CT images, including:

[0039] The acquisition module is used to acquire various two-dimensional projection images acquired by a computed tomography (CT) device at different acquisition angles and to construct an original image set;

[0040] The first generation module is used to determine, for each frame of two-dimensional projection image contained in the original image set, a two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image of that frame, as the associated image corresponding to that frame of two-dimensional projection image, and generate a two-dimensional projection image whose acquisition angle is between that frame of two-dimensional projection image and the associated image based on the two-dimensional projection image of that frame and the associated image, as the first compensation image.

[0041] The second generation module is used to generate a three-dimensional CT image based on the first compensated image and the original image set.

[0042] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for generating three-dimensional CT images.

[0043] This specification provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for generating three-dimensional CT images.

[0044] The above-mentioned technical solutions adopted in this specification can achieve the following beneficial effects:

[0045] In the method for generating three-dimensional CT images provided in this specification, firstly, two-dimensional projection images acquired by a computed tomography (CT) device at different acquisition angles are obtained, and an original image set is constructed. For each frame of two-dimensional projection images contained in the original image set, two-dimensional projection images whose acquisition angles are within a preset neighborhood range of the corresponding acquisition angle of the frame of two-dimensional projection images are determined as associated images corresponding to the frame of two-dimensional projection images. Based on the frame of two-dimensional projection images and the associated images, a two-dimensional projection image whose acquisition angle is between the frame of two-dimensional projection images and the associated images is generated as a first compensation image. Based on the first compensation image and the original image set, a three-dimensional CT image is generated.

[0046] As can be seen from the above method, this scheme can generate a compensation image before generating a three-dimensional CT image based on the original image set acquired by the computed tomography CT equipment. This can avoid the problem of bar artifacts in the generated three-dimensional CT image caused by the small number of two-dimensional projection images in the original image set, thereby improving the signal-to-noise ratio of the generated three-dimensional CT image. Attached Figure Description

[0047] The accompanying drawings, which are included to provide a further understanding of this specification and form part of this specification, illustrate exemplary embodiments and are used to explain this specification, but do not constitute an undue limitation thereof. In the drawings:

[0048] Figure 1 This is a flowchart illustrating a method for generating a three-dimensional CT image provided in this specification.

[0049] Figure 2 This is a schematic diagram of a three-dimensional CT image generation process provided in this specification;

[0050] Figure 3 This is a schematic diagram of a three-dimensional CT image generation device provided in this specification;

[0051] Figure 4 The one provided in this specification corresponds to Figure 1 A schematic diagram of an electronic device. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0053] The technical solutions provided in the various embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0054] Figure 1 This is a flowchart illustrating a method for generating a three-dimensional CT image provided in this specification, including the following steps:

[0055] S101: Acquire the two-dimensional projection images acquired by the computed tomography (CT) device at different acquisition angles and construct the original image set.

[0056] Currently, 3D CT images are a commonly used type of medical image in the field of medical imaging. Doctors can understand the lesion status of the affected parts of the body based on the patient's 3D CT images, and then perform surgical treatment based on their understanding of the affected parts. In order to ensure that doctors can have a detailed understanding of the patient's physical condition and improve the success rate of surgery, the signal-to-noise ratio of 3D CT images is particularly important.

[0057] Based on this, in this specification, the terminal device can acquire various two-dimensional projection images acquired by a computed tomography (CT) device at different acquisition angles and construct an original image set. Then, based on the two-dimensional projection images in the original image set, a compensation image can be generated. Finally, based on the generated compensation image and the two-dimensional projection images in the original image set, a three-dimensional CT image can be generated. The two-dimensional projection image is the attenuation image of X-rays acquired by the CT device after passing through the human body. In other words, it is the two-dimensional projection result obtained by scanning a cross-section of the diseased area from a certain angle with X-rays.

[0058] Furthermore, in order to improve the signal-to-noise ratio of the compensated images generated from each two-dimensional projection image, after acquiring the two-dimensional projection image, the terminal device can perform image denoising on each two-dimensional projection image in the original image set using a preset denoising method. The preset denoising method can be, for example, convolutional filtering, image denoising algorithm using sparse three-dimensional transform domain collaborative filtering (Block Matching 3D, BM3D), or neural network denoising using three-dimensional block matching neural network (BlockMatching 3D, DnCNN).

[0059] In this specification, the execution subject used to implement the method for generating three-dimensional CT images can refer to a specified device such as a server, or a terminal device such as a desktop computer or a laptop computer. For ease of description, the following description will only use a terminal device as the execution subject to illustrate the method for generating three-dimensional CT images provided in this specification.

[0060] S102: For each frame of two-dimensional projection image contained in the original image set, determine the two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image of that frame, and use it as the associated image corresponding to the two-dimensional projection image of that frame. Based on the two-dimensional projection image of that frame and the associated image, generate a two-dimensional projection image whose acquisition angle is between the two-dimensional projection image of that frame and the associated image, and use it as the first compensation image.

[0061] S103: Generate a three-dimensional CT image based on the first compensated image and the original image set.

[0062] After acquiring the original image set, the terminal device can determine the two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image in each frame of the original image set, and use it as the associated image corresponding to the two-dimensional projection image in that frame. Based on the two-dimensional projection image in that frame and the associated image, a two-dimensional projection image whose acquisition angle is between the two-dimensional projection image in that frame and the associated image is generated as the first compensation image. Then, a three-dimensional CT image can be generated based on the first compensation image and the original image set. The preset neighborhood range of the two-dimensional projection image in that frame can take many forms, such as: the acquisition angle corresponding to the two-dimensional projection image in that frame is within a preset angle range, the number of scan frames corresponding to the two-dimensional projection image in that frame is within a preset number of frames, or the scan time corresponding to the two-dimensional projection image in that frame is within a preset time range.

[0063] Specifically, the terminal device can determine the angle difference between the acquisition angle corresponding to each frame of two-dimensional projection image contained in the original image set and the acquisition angle corresponding to the associated image of the frame of two-dimensional projection image. It can then determine whether the angle difference between the acquisition angle of the frame of two-dimensional projection image and the associated image of the frame of two-dimensional projection image exceeds a preset threshold. If so, it can generate a two-dimensional projection image with an acquisition angle between the two-dimensional projection image and the associated image of the frame of two-dimensional projection image based on the two-dimensional projection image and the associated image of the frame of two-dimensional projection image.

[0064] For example: Suppose there are two adjacent frames of two-dimensional projection images with acquisition angles of 84° and 86° respectively, and the preset threshold is 1°. At this time, it can be determined that the angle difference between the acquisition angles of these two frames of two-dimensional projection images is 2°. Then, the terminal device can generate a two-dimensional projection image with an acquisition angle between the two-dimensional projection image and the associated image based on the two-dimensional projection image and the associated image. That is, a two-dimensional projection image with an acquisition angle of 85° is used as the first compensation image.

[0065] In addition, in practical applications, the angle difference between the acquisition angles of the two-dimensional projection image and its associated image may be large. In this case, it is necessary to generate multiple two-dimensional projections with acquisition angles between the two-dimensional projection image and its associated image.

[0066] Specifically, the terminal device can determine the target number of two-dimensional projection images whose acquisition angles fall between the acquisition angles of the two-dimensional projection image and its associated image, based on the acquisition angle corresponding to the frame of the two-dimensional projection image, the angle difference between the acquisition angles of the frame of the two-dimensional projection image, and a preset threshold. Then, based on the frame of the two-dimensional projection image and its associated image, the terminal device can generate the target number of two-dimensional projection images whose acquisition angles fall between the acquisition angles of the frame of the two-dimensional projection image and its associated image, as the first compensation image.

[0067] For example: Suppose there are two adjacent frames of two-dimensional projection images with acquisition angles of 84° and 87° respectively, and the preset threshold is 1°. In this case, it can be determined that the angle difference between the acquisition angles of these two frames of two-dimensional projection images is 3°. Then, based on the preset threshold and the above angle difference, the target number of two-dimensional projection images to be generated is 2, that is, two frames of two-dimensional projection images with acquisition angles of 85° and 86° need to be generated. Then, based on the two-dimensional projection images and the associated images, two frames of two-dimensional projection images with acquisition angles located between the two-dimensional projection images and the associated images are generated as the first compensation images.

[0068] In the above content, the method by which the terminal device generates a two-dimensional projection image whose acquisition angle is between the acquisition angle corresponding to the two-dimensional projection image and the acquisition angle corresponding to ...

[0069] For example: Suppose a terminal device needs to generate a first compensation image with a sampling angle of 86° based on two frames of 2D projection images with sampling angles of 84° and 90° respectively. In this case, the terminal device can determine the image change amount required to convert the 2D projection image with a sampling angle of 84° to one with a sampling angle of 90°. Then, based on the 6° angle difference between the sampling angles of these two 2D projection images, and the 2° angle difference between the sampling angle of the first compensation image and the sampling angle of the 84° 2D projection image, the deformation coefficient required to generate the first compensation image is determined to be 1 / 3 (based on the two angles mentioned above). The ratio of the difference 6° and 2° is used to determine the image change in converting a 2D projection image with an 84° acquisition angle to a 90° acquisition angle. This parameter can be modified to one-third of the original value. For example, if the image change in converting a 2D projection image with an 84° acquisition angle to a 90° acquisition angle includes shifting the horizontal coordinate of each pixel in the 84° acquisition angle 2D projection image to the right by 6 units and increasing the grayscale value by 12, then when generating the compensation image, it can be modified to shift the horizontal coordinate of each pixel in the 84° acquisition angle 2 units to the right and increase the grayscale value by 4.

[0070] In the above content, the method by which the terminal device determines the image change amount that can be converted into the associated image of the two-dimensional projection image of the frame can be that the terminal device inputs the two-dimensional projection image of the frame and the associated image of the frame into a pre-trained image conversion model, so as to output the image change amount that transforms the two-dimensional projection image of the frame into the associated image of the frame through the preset image conversion model.

[0071] The training method for the image conversion model can be as follows: the terminal device acquires a first sample projection image and a second sample projection image, inputs the first sample projection image and the second sample projection image into the image conversion model, obtains the image change between the first sample projection image and the second sample projection image as the image change to be optimized, converts the first sample projection image according to the image change to be optimized, and trains the image conversion model with the optimization objective of minimizing the deviation between the converted projection image and the second sample projection image. The first sample projection image and the second sample projection image are two two-dimensional projection images of the same part taken from different acquisition angles.

[0072] It should be noted that, in order to make the trained image conversion model more robust, the first and second sample projection images acquired from historically acquired two-dimensional projection images can be selected from X-ray projection images of different parts of the body as much as possible (the X-ray projection images of different parts of the body mentioned here can be, for example, two frames of two-dimensional projection images of the patient's cranium at different acquisition angles in the first acquisition, and two frames of two-dimensional projection images of the ankle at different acquisition angles in the second acquisition, in addition to the cranium). This allows the image conversion model to perform image conversion on two-dimensional projection images of various diseased parts of various patients and obtain the corresponding image changes.

[0073] In the above content, minimizing the deviation between the transformed projection image and the second sample projection image is the optimization objective. The method for training the image transformation model can be to determine the image loss based on the transformed projection image and the second sample projection image, and train the image transformation model with the goal of minimizing the image loss. Here, the smaller the image loss, the smaller the deviation between the transformed projection image and the second sample projection image.

[0074] In the above content, image loss may include at least one of grayscale loss and feature loss, wherein grayscale loss is used to characterize the degree of difference in grayscale information between the transformed projection image and the second sample projection image, and feature loss is used to characterize the degree of difference in edge features, color features, and brightness features between the transformed projection image and the second sample projection image.

[0075] In practical applications, in order to further optimize the acquired first compensated image, the terminal device can also construct an initial three-dimensional CT image based on each two-dimensional projection image in the original image set, and then generate a second compensated image based on the initial three-dimensional CT image.

[0076] Furthermore, the terminal device can determine the weights corresponding to the first compensation image and the second compensation image based on the number of two-dimensional projection images contained in the original image set. The weight of the first compensation image is negatively correlated with the number of images, and the weight of the second compensation image is positively correlated with the number of images. Based on the weights of the first and second compensation images, the first and second compensation images are fused to generate a fused image. Based on the fused image and the original image set, a three-dimensional CT image is generated.

[0077] It should be noted that since the second compensation image is obtained from the initial 3D CT image, and the more 2D projection images contained in the original image set, the higher the confidence level of the generated initial 3D CT image, the higher the confidence level of the second compensation image obtained from the initial 3D CT image, and therefore, the greater the weight determined for the second compensation image.

[0078] In the above content, based on the weights of the first compensation image and the second compensation image, and the first compensation image and the second compensation image, there are many ways to generate a fused image. For example, a grayscale fusion algorithm can be used to fuse the grayscale information of each pixel contained in the two compensation images to obtain a fused image.

[0079] In the above content, there are many ways for the terminal device to construct the initial three-dimensional CT image based on the two-dimensional projection images contained in the original image set. For example, the initial three-dimensional CT image can be generated by performing an inverse Radon transform on all the two-dimensional projection images in the original image set. Other methods will not be listed here.

[0080] In practical applications, due to the exposure parameters and dose configured in low-dose CT acquisition equipment, the contrast of image regions related to the soft tissue of the patient's diseased area in the acquired two-dimensional projection images is often low. Consequently, the contrast of image regions related to soft tissue in the generated three-dimensional CT images is also low. Therefore, in order to further optimize the three-dimensional CT images generated by the above method, the terminal device can also input the generated three-dimensional CT images into a pre-trained optimization model. The optimization model can then optimize the three-dimensional CT images to obtain optimized three-dimensional CT images. In this case, the dose of medical radiation used by the CT equipment when acquiring each frame of two-dimensional projection images contained in the original image set is less than the dose of medical radiation corresponding to the optimized three-dimensional CT images.

[0081] In the above content, the dose of medical radiation corresponding to the optimized 3D CT image refers to the dose of medical radiation required when the CT equipment acquires each 2D projection image needed to construct the optimized 3D CT image.

[0082] In the above content, the training method of the optimization model can be as follows: acquire a first sample three-dimensional CT image and a second sample three-dimensional CT image for the same region (i.e., the patient's body region), wherein the dose of medical radiation corresponding to the first sample three-dimensional CT image is less than the dose of medical radiation corresponding to the second sample three-dimensional CT image, input the first sample three-dimensional CT image into the optimization model, and obtain the optimized first sample three-dimensional CT image output by the optimization model, with the optimization objective being to minimize the deviation between the optimized first sample three-dimensional CT image and the second sample three-dimensional CT image, and train the optimization model.

[0083] To facilitate understanding, this manual also provides a flowchart illustrating the process of generating three-dimensional CT images, such as... Figure 2 As shown.

[0084] Figure 2This is a schematic diagram of a three-dimensional CT image generation process provided in this specification.

[0085] Combination Figure 2 It can be seen that the terminal device can acquire various two-dimensional projection images of the patient's diseased area through CT equipment, and generate a first compensation image based on the acquired two-dimensional projection images.

[0086] Furthermore, the terminal device can construct an initial three-dimensional CT image based on the acquired two-dimensional projection images, and obtain a second compensation image based on the constructed initial three-dimensional CT image. Then, the weights of the first compensation image and the second compensation image can be determined, and the first compensation image and the second compensation image can be fused to obtain a fused image.

[0087] Furthermore, the terminal device can generate a 3D CT image by reconstructing the fused image and the 2D projection images contained in the original image set, and optimize the reconstructed 3D CT image by using a preset optimization model to obtain an optimized 3D CT image.

[0088] It should be noted that all actions involving the acquisition of signals, information, or data in this manual are performed in accordance with the relevant data protection laws and regulations of the country where the device is located, and with the authorization granted by the owner of the relevant device.

[0089] As can be seen from the above, the terminal device can generate first and second compensation images based on the two-dimensional projection images acquired by the CT image acquisition device using low-dose X-rays. Then, based on the first and second compensation images, it can generate fused images. Thus, based on the generated fused images and the original image set, a three-dimensional CT image can be generated. The generated three-dimensional CT image is then optimized again to reduce noise and enhance the contrast of soft tissue regions contained in the generated three-dimensional CT image, thereby improving the signal-to-noise ratio of the generated three-dimensional CT image.

[0090] The above describes one or more methods for generating three-dimensional CT images according to this specification. Based on the same idea, this specification also provides corresponding three-dimensional CT image generation devices, such as... Figure 3 As shown.

[0091] Figure 3 A schematic diagram of a three-dimensional CT image generation device provided in this specification includes:

[0092] The acquisition module 301 is used to acquire various two-dimensional projection images acquired by the computed tomography CT device at different acquisition angles and to construct the original image set;

[0093] The first generation module 302 is used to determine, for each frame of two-dimensional projection image contained in the original image set, a two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image of that frame, as the associated image corresponding to that frame of two-dimensional projection image, and generate a two-dimensional projection image whose acquisition angle is between that frame of two-dimensional projection image and the associated image, as the first compensation image, based on the two-dimensional projection image of that frame and the associated image.

[0094] The second generation module 303 is used to generate a three-dimensional CT image based on the first compensated image and the original image set.

[0095] Optionally, the first generation module 302 is specifically used to: determine the angle difference between the acquisition angle corresponding to the frame of the two-dimensional projection image and the acquisition angle corresponding to the associated image; determine whether the angle difference exceeds a preset threshold; if so, generate a two-dimensional projection image with an acquisition angle between the frame of the two-dimensional projection image and the associated image based on the two-dimensional projection image and the associated image.

[0096] Optionally, the first generation module 302 is specifically used to input the frame of two-dimensional projection image and the associated image into a pre-trained image conversion model, so as to determine the amount of image change between the frame of two-dimensional projection image and the associated image through the image conversion model; and to generate the first compensation image based on the amount of image change, the acquisition angle corresponding to the frame of two-dimensional projection image and the acquisition angle corresponding to the associated image.

[0097] Optionally, the device further includes:

[0098] The first training module 304 is used to acquire a first sample projection image and a second sample projection image; input the first sample projection image and the second sample projection image into the image conversion model to obtain the image change between the first sample projection image and the second sample projection image, which is used as the image change to be optimized; convert the first sample projection image according to the image change to be optimized to obtain a converted projection image; and train the image conversion model with minimizing the deviation between the converted projection image and the second sample projection image as the optimization objective.

[0099] Optionally, the first generation module 302 is specifically used to construct an initial three-dimensional CT image based on each frame of two-dimensional projection images in the original image set; and to generate a two-dimensional projection image with the same acquisition angle as the first compensation image using the initial three-dimensional CT image as the second compensation image.

[0100] The second generation module 303 is specifically used to generate a three-dimensional CT image based on the first compensated image, the second compensated image, and the original image set.

[0101] Optionally, the first generation module 302 is specifically configured to: determine the weights corresponding to the first compensation image and the second compensation image based on the number of two-dimensional projection images contained in the original image set, wherein the weights corresponding to the first compensation image are negatively correlated with the number and the weights corresponding to the second compensation image are positively correlated with the number; fuse the first compensation image and the second compensation image based on the weights corresponding to the first compensation image and the second compensation image to generate a fused image; and generate a three-dimensional CT image based on the fused image and the original image set.

[0102] Optionally, the second generation module 303 is specifically used to input the three-dimensional CT image into a pre-trained optimization model, so as to optimize the three-dimensional CT image through the optimization model to obtain an optimized three-dimensional CT image, wherein the dose of medical radiation used by the CT device when acquiring each frame of two-dimensional projection image contained in the original image set is less than the dose of medical radiation corresponding to the optimized three-dimensional CT image.

[0103] Optionally, the device further includes:

[0104] The second training module 305 is used to acquire a first sample three-dimensional CT image and a second sample three-dimensional CT image for the same region, wherein the dose of medical radiation corresponding to the first sample three-dimensional CT image is less than the dose of medical radiation corresponding to the second sample three-dimensional CT image; input the first sample three-dimensional CT image into the optimization model to obtain the optimized first sample three-dimensional CT image output by the optimization model; and train the optimization model with the goal of minimizing the deviation between the optimized first sample three-dimensional CT image and the second sample three-dimensional CT image.

[0105] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 A method for generating three-dimensional CT images is provided.

[0106] This instruction manual also provides Figure 4 One of the corresponding Figure 1 A schematic diagram of the structure of an electronic device. (e.g.) Figure 4At the hardware level, the electronic device includes a processor, an internal bus, a network interface, memory, and non-volatile storage, and may also include other hardware required for the business operations. The processor reads the corresponding computer program from the non-volatile storage into memory and then runs it to achieve the above-mentioned functions. Figure 1 The method for generating three-dimensional CT images is described above. Of course, besides software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution entity of the following processing flow is not limited to individual logic units, but can also be hardware or logic devices.

[0107] In the 1990s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many methodological improvements today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that a methodological improvement cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program and "integrate" a digital system onto a PLD themselves, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0108] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0109] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0110] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware.

[0111] Those skilled in the art will understand that embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0112] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0113] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0114] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0115] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0116] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0117] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0118] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0119] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0120] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0121] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0122] The above description is merely an embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification.

Claims

1. A method for generating three-dimensional CT images, characterized in that, include: Acquire two-dimensional projection images from different acquisition angles using a computed tomography (CT) scanner and construct an original image set; For each frame of two-dimensional projection image contained in the original image set, determine the two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image of that frame, and use it as the associated image corresponding to the two-dimensional projection image of that frame. Based on the two-dimensional projection image of that frame and the associated image, generate a two-dimensional projection image whose acquisition angle is between the two-dimensional projection image of that frame and the associated image, and use it as the first compensation image. Based on each frame of the two-dimensional projection image in the original image set, an initial three-dimensional CT image is constructed; Using the initial three-dimensional CT image, a two-dimensional projection image with the same acquisition angle as the first compensation image is generated as the second compensation image; A three-dimensional CT image is generated based on the first compensated image, the second compensated image, and the original image set.

2. The method as described in claim 1, characterized in that, Based on the frame's two-dimensional projection image and the associated image, a two-dimensional projection image with an acquisition angle between the frame's two-dimensional projection image and the associated image is generated, specifically including: Determine the angle difference between the acquisition angle corresponding to the two-dimensional projection image of the frame and the acquisition angle corresponding to the associated image; Determine whether the angle difference exceeds a preset threshold; If so, based on the two-dimensional projection image and the associated image, generate a two-dimensional projection image with an acquisition angle between the two-dimensional projection image of the frame and the associated image.

3. The method as described in claim 1, characterized in that, Based on the frame's two-dimensional projection image and the associated image, a two-dimensional projection image with an acquisition angle between the frame's two-dimensional projection image and the associated image is generated as the first compensation image, specifically including: The two-dimensional projection image of the frame and the associated image are input into a pre-trained image conversion model to determine the amount of image change between the two-dimensional projection image of the frame and the associated image through the image conversion model. The first compensation image is generated based on the image change, the acquisition angle corresponding to the two-dimensional projection image of the frame, and the acquisition angle corresponding to the associated image.

4. The method as described in claim 3, characterized in that, Training the image conversion model specifically includes: Obtain the first sample projection image and the second sample projection image; The first sample projection image and the second sample projection image are input into the image conversion model to obtain the image change between the first sample projection image and the second sample projection image, which is used as the image change to be optimized. Based on the change amount of the image to be optimized, the first sample projection image is transformed to obtain the transformed projection image; The image conversion model is trained with the optimization objective of minimizing the deviation between the converted projection image and the second sample projection image.

5. The method as described in claim 1, characterized in that, Generating a 3D CT image based on the first compensated image, the second compensated image, and the original image set, specifically includes: Based on the number of two-dimensional projected images contained in the original image set, the weights corresponding to the first compensation image and the second compensation image are determined. The weights corresponding to the first compensation image are negatively correlated with the number of images, and the weights corresponding to the second compensation image are positively correlated with the number of images. Based on the weights corresponding to the first compensation image and the second compensation image, the first compensation image and the second compensation image are fused to generate a fused image. A three-dimensional CT image is generated based on the fused image and the original image set.

6. The method as described in claim 1, characterized in that, The method further includes: The three-dimensional CT image is input into a pre-trained optimization model to optimize the three-dimensional CT image, thereby obtaining an optimized three-dimensional CT image. The dose of medical radiation used by the CT device when acquiring each frame of two-dimensional projection image contained in the original image set is less than the dose of medical radiation corresponding to the optimized three-dimensional CT image.

7. The method as described in claim 6, characterized in that, Training the optimized model specifically includes: A first sample 3D CT image and a second sample 3D CT image of the same area are acquired, wherein the dose of medical radiation corresponding to the first sample 3D CT image is less than the dose of medical radiation corresponding to the second sample 3D CT image. The first sample 3D CT image is input into the optimization model to obtain the optimized first sample 3D CT image output by the optimization model. The optimization model is trained with the goal of minimizing the deviation between the optimized first sample 3D CT image and the second sample 3D CT image.

8. A device for generating three-dimensional CT images, characterized in that, include: The acquisition module is used to acquire various two-dimensional projection images acquired by a computed tomography (CT) device at different acquisition angles and to construct an original image set; The first generation module is configured to, for each frame of two-dimensional projection image contained in the original image set, determine a two-dimensional projection image whose acquisition angle is within a preset neighborhood range of the two-dimensional projection image of that frame, as the associated image corresponding to that frame of two-dimensional projection image; and generate a two-dimensional projection image whose acquisition angle is between that frame of two-dimensional projection image and the associated image, as a first compensation image, based on the two-dimensional projection image of that frame and the associated image; construct an initial three-dimensional CT image based on each frame of two-dimensional projection image in the original image set; and generate a two-dimensional projection image with the same acquisition angle as the first compensation image, as a second compensation image, using the initial three-dimensional CT image. The second generation module is used to generate a three-dimensional CT image based on the first compensated image, the second compensated image, and the original image set.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1 to 7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 7.

Images