Scanning Method, Device, Storage Medium, and CT Equipment

Abstract

The present application discloses a scanning method, apparatus, storage medium, and CT device. The scanning method includes: obtaining plain film data of a target object; inputting the plain film data into a target model to obtain predicted sectional information of the target object; and based on the predicted sectional information, controlling the CT device to scan a part to be scanned of the target object. The scanning method provided by the embodiments of the present application utilizes big data to obtain the accurate body shape of the target object and performs precise adjustment of the scanning dose and bed height, reducing the harm suffered by the target object while ensuring the imaging quality.

Description

Technical Field

[0001] The present application relates to the technical field of medical devices, and particularly to a scanning method, device, storage medium, and CT device. Background Art

[0002] Medical imaging refers to the technology and processing process of obtaining internal tissue images of the human body or a part of the human body in a non-invasive manner for medical treatment or medical research. Due to its non-invasive characteristics, it has become an important means of auxiliary diagnosis. Especially for computed tomography (CT) devices, they play a very important role in diagnosing whether there are lesions in the human central nervous system, abdominal organs, bone joints, etc.

[0003] During the scanning process, the higher the scanning dose of the CT device, the higher the quality of the generated image. Considering that X-rays and the like are all somewhat harmful to the human body, the greater the scanning dose, the greater the harm, and the attenuation of different parts of the human body is different. For example, the attenuation of the shoulder is large, and the attenuation of the lung is small. If X-rays are emitted with the same emission energy, it will result in different energies of the X-rays reaching the receiving end, causing differences in image quality, or an increase in the radiation dose received by the patient's part. For example, using the scanning dose of the shoulder to scan the lung will result in an increase in the radiation dose of the lung.

[0004] In the related art, the thickness measurement values of different positions of the patient are measured using the image depth information of the patient, and then the exposure parameters required by the X-ray generator are found in the E1 standard range table according to the thickness measurement values and X-ray scanning is performed. However, the estimation of the body shape based on the depth information is not accurate enough, which affects the selection of scanning, and it is necessary to add depth information acquisition equipment, increasing the working cost of the equipment. Summary of the Invention

[0005] In view of this, the present application provides a scanning method, device, storage medium, and CT device, which use big data to obtain the accurate body shape of the target object and perform precise adjustment of the scanning dose, reducing the harm received by the target object while ensuring the imaging quality.

[0006] According to one aspect of the present application, a scanning method is provided, including:

[0007] Obtain the plain film data of the target object;

[0008] Input the plain film data into the target model to obtain the estimated cross-sectional information of the target object;

[0009] Based on the estimated cross-sectional information, control the CT device to scan the part to be scanned of the target object.

[0010] Optionally, based on the estimated section information, control the CT device to scan the part to be scanned of the target object, including:

[0011] Based on the estimated section information, adjust the scanning dose of the radiation source of the CT device; and / or,

[0012] When the estimated section information includes an estimated section image, based on the long and short axis parameters associated with the estimated section image, adjust the bed height of the scanning bed of the CT device.

[0013] Optionally, based on the estimated section information, adjust the scanning dose of the radiation source of the CT device, including:

[0014] Based on the estimated section information, obtain the current modulation curve of the part to be scanned of the target object, and adjust the scanning dose based on the current modulation curve.

[0015] Optionally, based on the estimated section information, adjust the scanning dose of the radiation source of the CT device, including:

[0016] When the estimated section information includes an estimated section image, based on the long and short axis parameters associated with the estimated section image, determine the dose ratio between different parts in the part to be scanned of the target object, and adjust the scanning dose based on the dose ratio.

[0017] Optionally, when the estimated section information includes an estimated section image, based on the estimated section information, obtain the dose modulation curve of the part to be scanned of the target object, including:

[0018] Perform projection processing on the estimated section image of the part to be scanned of the target object according to multiple radiation positions, and determine the estimated attenuation domain information of the part to be scanned of the target object at multiple radiation positions;

[0019] According to the estimated attenuation domain information, configure the current modulation curve of the part to be scanned of the target object.

[0020] Optionally, before performing projection processing on the estimated section image of the part to be scanned of the target object according to multiple radiation positions, the scanning method further includes:

[0021] Based on the plain film data, obtain the part to be scanned of the target object; or,

[0022] When the estimated section information includes an estimated section image, perform part classification processing on the estimated section image according to the long and short axis parameters associated with the estimated section image, so as to obtain the part to be scanned of the target object.

[0023] Optionally, according to the estimated attenuation domain information, configure the current modulation curve of the part to be scanned of the target object, including:

[0024] Calculate the target current value of the part to be scanned of the target object at each wire placement position according to the estimated attenuation domain information;

[0025] Determine the current modulation curve of the part to be scanned of the target object according to the target current value of the part to be scanned of the target object at each wire placement position;

[0026] Among them, when calculating the target current value of the part to be scanned of the target object according to the attenuation domain information, the following formula is used:

[0027]

[0028] In the formula, mAs ACS represents the target current value, μ water represents the attenuation coefficient of water, D ref represents the preset equivalent water phantom diameter, μ i D i represents the attenuation domain information, mAs ref represents the preset reference current value, and adjCoef represents the preset exponential adjustment parameter.

[0029] Optionally, before configuring the current modulation curve of the part to be scanned of the target object according to the estimated attenuation domain information, the scanning method further includes:

[0030] When the CT scan mode is spiral scan, calculate the error amount according to the actual attenuation domain information and the estimated attenuation domain information of the already wire-placed positions in the estimated cross-sectional image;

[0031] Correct the estimated attenuation domain information of the non-wire-placed positions adjacent to the already wire-placed positions in the estimated cross-sectional image according to the error amount;

[0032] Among them, when correcting the estimated attenuation domain information of the non-wire-placed positions adjacent to the already wire-placed positions in the estimated cross-sectional image according to the error amount, the following formula is used:

[0033] PreAtt(w + 1) = PreAtt(w + 1)' × (RealAtt(w) / PreAtt(w)'),

[0034] In the formula, PreAtt(w + 1) represents the estimated attenuation domain information of the non-wire-placed position adjacent to the wth already wire-placed position after correction, PreAtt(w + 1)' represents the estimated attenuation domain information of the non-wire-placed position adjacent to the wth already wire-placed position before correction, PreAtt(w)' represents the estimated attenuation domain information of the wth already wire-placed position, and RealAtt(w) represents the actual attenuation domain information of the wth already wire-placed position.

[0035] Optionally, the scanning method further includes:

[0036] Based on the estimated cross-sectional image, obtain the major and minor axis parameters associated with the estimated cross-sectional image; or,

[0037] When inputting the plain film data into the target model, simultaneously obtain the major and minor axis parameters associated with the estimated cross-sectional image.

[0038] Optionally, the scanning method further includes:

[0039] Obtain multiple plain film data and multiple groups of cross-sectional information of the sample object, where the cross-sectional information of the sample object includes tomographic cross-sectional information or helical cross-sectional information;

[0040] Perform an association process on each row in the plain film image corresponding to the plain film data of each sample object and each cross-sectional information in each group of cross-sectional information of the sample object according to the bed code value;

[0041] Divide the multiple plain film data of the sample object into sample plain film data and test plain film data according to a preset ratio, and determine the cross-sectional information associated with the sample plain film data as sample cross-sectional information, and determine the cross-sectional information associated with the test plain film data as test cross-sectional information;

[0042] Train the neural network model according to the sample plain film data and the sample cross-sectional information;

[0043] Evaluate the trained neural network model according to the test plain film data and the test cross-sectional information to obtain evaluation metrics;

[0044] When the evaluation metrics meet the convergence conditions of the loss function, confirm the trained neural network model as the target model.

[0045] Optionally, the scanning method further includes:

[0046] Determine the major and minor axis parameters of the effective region according to the number of pixels of the sample object in the effective region of the cross-sectional image of the cross-sectional information of the sample object;

[0047] Perform an association process on the major and minor axis parameters of the effective region and the cross-sectional image.

[0048] Optionally, evaluating the trained neural network model according to the test plain film data and the test cross-sectional information includes:

[0049] Input the test plain film data into the trained neural network model to obtain the cross-sectional information to be tested;

[0050] Calculate the evaluation metrics according to the first number of pixels of the scanned object in the effective region of the cross-sectional image to be tested in the cross-sectional information to be tested and the second number of pixels of the scanned object in the effective region of the test cross-sectional image in the test cross-sectional information.

[0051] According to another aspect of the present application, a scanning device is provided, including:

[0052] An acquisition module, configured to acquire plain film data of a target object;

[0053] A data processing module, configured to input the plain film data into a target model to obtain predicted sectional information of the target object;

[0054] A control module, configured to control a CT device to scan a to-be-scanned part of the target object based on the predicted sectional information.

[0055] Optionally, the control module is specifically configured to adjust the scanning dose of the radiation source of the CT device based on the predicted sectional information; and / or, in the case where the predicted sectional information includes a predicted sectional image, adjust the bed height of the scanning bed of the CT device based on the major and minor axis parameters associated with the predicted sectional image.

[0056] Optionally, the control module is specifically configured to obtain a current modulation curve of the to-be-scanned part of the target object based on the predicted sectional information, and adjust the scanning dose based on the current modulation curve.

[0057] Optionally, the scanning device further includes: a first determination module, configured to determine the dose ratio between different parts in the to-be-scanned part of the target object based on the major and minor axis parameters associated with the predicted sectional image in the case where the predicted sectional information includes a predicted sectional image; the control module is specifically configured to adjust the scanning dose based on the dose ratio.

[0058] Optionally, in the case where the predicted sectional information includes a predicted sectional image, the scanning device further includes: a second determination module, configured to perform projection processing on the predicted sectional image of the to-be-scanned part of the target object according to a plurality of radiation positions to determine the predicted attenuation domain information of the to-be-scanned part of the target object at the plurality of radiation positions; and configure the current modulation curve of the to-be-scanned part of the target object according to the predicted attenuation domain information.

[0059] Optionally, the scanning device further includes: a third determination module, configured to obtain the to-be-scanned part of the target object based on the plain film data; or, in the case where the predicted sectional information includes a predicted sectional image, perform part classification processing on the predicted sectional image according to the major and minor axis parameters associated with the predicted sectional image to obtain the to-be-scanned part of the target object.

[0060] Optionally, the second determination module is specifically configured to calculate the target current value of the to-be-scanned part of the target object at each radiation position according to the predicted attenuation domain information; determine the current modulation curve of the to-be-scanned part of the target object according to the target current value of the to-be-scanned part of the target object at each radiation position; calculate the target current value of the to-be-scanned part of the target object at each radiation position according to the attenuation domain information, using the following formula:

[0061]

[0062] Wherein, mAs ACS represents the target current value, μ water represents the attenuation coefficient of water, D ref represents the preset equivalent water phantom diameter, μ i D i represents the attenuation domain information, mAs ref represents the preset reference current value, and adjCoef represents the preset exponential adjustment parameter.

[0063] Optionally, the scanning device further includes: a correction module, configured to calculate an error amount according to the actual attenuation domain information and the estimated attenuation domain information of the already laid position of the estimated sectional image when the CT scanning mode is spiral scanning; and correct the estimated attenuation domain information of the non-laid position adjacent to the already laid position in the estimated sectional image according to the error amount;

[0064] Correcting the estimated attenuation domain information of the non-laid position adjacent to the already laid position in the estimated sectional image by using the following formula:

[0065] PreAtt(w + 1) = PreAtt(w + 1)' × (RealAtt(w) / PreAtt(w)'),

[0066] Wherein, PreAtt(w + 1) represents the estimated attenuation domain information of the non-laid position adjacent to the w-th already laid position after correction, PreAtt(w + 1)' represents the estimated attenuation domain information of the non-laid position adjacent to the w-th already laid position before correction, PreAtt(w)' represents the estimated attenuation domain information of the w-th already laid position, and RealAtt(w) represents the actual attenuation domain information of the w-th already laid position.

[0067] Optionally, the scanning device further includes: an identification module, configured to obtain the long and short axis parameters associated with the estimated sectional image based on the estimated sectional image.

[0068] Optionally, the data processing module is further configured to obtain the long and short axis parameters associated with the estimated sectional image while inputting the plain film data into the target model.

[0069] Optionally, the acquisition module is further configured to acquire a plurality of plain film data and multiple groups of sectional information of a sample object, where the sectional information of the sample object includes tomographic sectional information or helical sectional information; the scanning device further includes: an association module, configured to perform association processing on each row in the plain film image corresponding to the plain film data of each sample object and each sectional information in each group of sectional information of the sample object according to the bed code value; the acquisition module is further configured to divide the plain film data of multiple sample objects into sample plain film data and test plain film data according to a preset ratio, and determine the sectional information associated with the sample plain film data as sample sectional information, and determine the sectional information associated with the test plain film data as test sectional information; the scanning device further includes: a training module, configured to train a neural network model according to the sample plain film data and the sample sectional information; evaluate the trained neural network model according to the test plain film data and the test sectional information to obtain an evaluation index; and confirm the trained neural network model as a target model when the evaluation index meets the convergence condition of the loss function.

[0070] Optionally, the scanning device further includes: a fourth determination module, configured to determine the long and short axis parameters of the effective region according to the number of pixels of the sample object in the effective region of the sectional image of the sectional information of the sample object; the association module is further configured to perform association processing on the long and short axis parameters of the effective region and the sectional image.

[0071] Optionally, the training module is specifically configured to input the test plain film data into the trained neural network model to obtain the to-be-tested sectional information; calculate the evaluation index according to the first number of pixels of the scanned object in the effective region of the to-be-tested sectional image in the to-be-tested sectional information and the second number of pixels of the scanned object in the effective region of the test sectional image in the test sectional information.

[0072] According to another aspect of the present application, there is provided a readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the above scanning method are implemented.

[0073] According to another aspect of the present application, there is provided a CT device, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, and when the processor executes the program, the steps of the above scanning method are implemented.

[0074] By the above technical solution, a large number of sample flat-plate data of sample objects are used as input, and the sample cross-section information associated with the sample flat-plate data is used as a label to train a deep learning network, so as to obtain a target model based on deep learning. The target model is used to match the flat-plate data of the target object with the possible estimated cross-section information of multiple different parts of the target object. Through the estimated cross-section information, the body shape of the target object can be estimated, and combined with the body shape and the parameters related to the estimated cross-section information, the scanning of the part to be scanned of the target object is completed. On the one hand, the estimated cross-section information can be obtained by using big data, and then the body shape of the target object can be analyzed, which improves the accuracy of body shape estimation, and there is no need to add additional detection equipment, which is beneficial to reducing equipment costs. On the other hand, through the body shape, different parts of the target object can be accurately analyzed, so that during the scanning process, appropriate scanning parameters can be set for different parts to be scanned of the same target object, which not only meets the clinical needs of the reconstructed image, but also reduces the damage of X-rays to the target object.

[0075] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] The drawings described herein are used to provide a further understanding of this application, and constitute a part of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation of this application. In the drawings:

[0077] Figure 1 One of the flow schematic diagrams of the scanning method provided by the embodiment of this application is shown;

[0078] Figure 2 Another flow schematic diagram of the scanning method provided by the embodiment of this application is shown;

[0079] Figure 3 Another flow schematic diagram of the scanning method provided by the embodiment of this application is shown;

[0080] Figure 4 Another flow schematic diagram of the scanning method provided by the embodiment of this application is shown;

[0081] Figure 5 The structural block diagram of the scanning device provided by the embodiment of this application is shown;

[0082] Figure 6 The schematic diagram of the current modulation curve provided by the embodiment of this application is shown;

[0083] Figure 7Shows the schematic diagram of CT device scanning provided by the embodiments of the present application;

[0084] Figure 8 Shows the schematic diagram of the scanning scene of the CT device provided by the embodiments of the present application. Detailed implementation manners

[0085] In the following, the present application will be described in detail with reference to the accompanying drawings and in conjunction with embodiments. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other.

[0086] The embodiments of the present application will be described in detail below. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary only for explaining the present application and should not be construed as limiting the present application.

[0087] Those skilled in the art of the present technology can understand that, unless specifically stated otherwise, the singular forms "a", "an", "the" and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of the present application means the presence of features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their groups. It should be understood that when we say that an element is "connected" or "joined" to another element, it can be directly connected or joined to other elements, or there may also be intermediate elements. In addition, the "connection" or "joining" used herein may include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any unit and all combinations of one or more related listed items.

[0088] Now, exemplary embodiments according to the present application will be described in more detail with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many different forms and should not be construed as being limited only to the embodiments set forth herein. It should be understood that these embodiments are provided so that the disclosure of the present application is thorough and complete, and the concepts of these exemplary embodiments are fully conveyed to those of ordinary skill in the art.

[0089] In this embodiment, a scanning method is provided, as Figure 1 shown, the method includes:

[0090] Step 101, obtaining the plain film data of the target object;

[0091] In actual application scenarios, the target object can be a human body, or other living or inanimate objects such as animals with medical imaging requirements. Plain film data is obtained by using the penetration, fluorescence, and photosensitivity of X-rays. The X-rays penetrate the target object, causing the film behind the target object to be exposed to light. Plain film images can be formed using the plain film data. Specifically, the higher the tissue density within the target object, the less the X-rays penetrate, the less the exposure, and the darker the image formed. Conversely, a brighter image is formed. According to the viewing habit, the black and white of the photo is reversed to form a plain film image.

[0092] It can be understood that when using a CT device for scanning, the user needs to position the target object differently according to the clinical manifestations and individual characteristics of the target object. After the positioning of the target object is determined, plain film data is usually scanned first. This plain film data has a positioning function. According to this plain film data, the scanning position and range can be determined. At least part of the target object corresponds to the plain film data, so that the scanning part for tomographic scanning or spiral scanning of the target object can be located through the plain film data. The plain film data can correspond to different parts of the target object, that is, one plain film data corresponds to one part. Of course, one plain film data can also correspond to multiple different parts of the scanned object at the same time. Taking the scanned object as a human body as an example, the different parts can be the head, chest, abdomen, limbs, etc.

[0093] Step 102: Input the plain film data into the target model to obtain the estimated cross-sectional information of the target object;

[0094] Among them, the estimated cross-sectional information includes cross-sectional images and / or cross-sectional data. For different scanning methods of CT devices, the estimated cross-sectional information can be tomographic cross-sectional information or spiral cross-sectional information.

[0095] In this embodiment, the plain film data of the target object is input into the target model to match the possible estimated cross-sectional information of multiple different parts of the target object for the plain film data of the target object through the target model. It not only realizes the determination of the estimated cross-sectional information using big data, facilitates the analysis of the body type of the target object, reduces the detection workload of the user, but also improves the accuracy of the body type estimation, which is beneficial to improving the work efficiency. In addition, there is no need to add additional detection equipment, which is beneficial to reducing the cost of CT devices.

[0096] Furthermore, as Figure 2 shown, as a refinement and extension of the specific implementation manner of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, before step 102, the scanning method further includes:

[0097] Step 201: Obtain multiple plain film data and multiple groups of cross-sectional information of the sample object;

[0098] Among them, the cross-sectional information of the sample object includes tomographic cross-sectional information or spiral cross-sectional information. The cross-sectional information includes cross-sectional data and cross-sectional images. The tomographic cross-sectional information is the information obtained by scanning the object to be scanned with the radiation source of the CT device rotating. The spiral cross-sectional information is the information obtained by scanning the object to be scanned while the radiation source of the CT device scans and the scanning bed also moves simultaneously.

[0099] It can be understood that each part of the sample object corresponds to a set of cross-sectional information, and each set of cross-sectional information includes at least one cross-sectional information.

[0100] In this embodiment, a large amount of clinical scanned plain film data and corresponding tomographic / spiral imaging information are collected. The parts scanned by the plain film data and the cross-sectional information should cover all possible parts of the object to be scanned, so as to provide sufficient data support for subsequent training and testing of the model.

[0101] Step 202, perform an association process on each row of the plain film image corresponding to the plain film data of each sample object and each cross-sectional information in each set of cross-sectional information of the sample object according to the bed code value;

[0102] In this embodiment, each row on the plain film image corresponding to the plain film data corresponds to a bed code value, and each cross-sectional information in each set of tomographic / spiral cross-sectional information also corresponds to a bed code value. There is a corresponding relationship between the bed code value of each row of the plain film image and the bed code value of each cross-sectional information in each set of tomographic / spiral cross-sectional information. For different models of CT devices, this corresponding relationship can be the same or different. Use this corresponding relationship to align the bed codes of each row of the plain film image corresponding to the plain film data and each cross-sectional information of each set of tomographic / spiral cross-sectional information, so as to associate the plain film data corresponding to each row of the plain film image and each cross-sectional information one by one. At this time, the plain film data corresponding to each row of the plain film image can be associated with a set of cross-sectional information, that is, at least one cross-sectional information, so as to use the cross-sectional information as the label of the plain film data to train the target model.

[0103] Specifically, for example, from the minimum value of the starting bed code of the plain film image and the tomographic / spiral image to the maximum value of the ending bed code of the plain film and the tomographic / spiral image, at a certain interval, such as the interval between each row of the plain film image (1 mm or 2 mm), a tomographic / spiral image can be located.

[0104] Step 203, divide the multiple plain film data of the sample object into sample plain film data and test plain film data according to a preset ratio, and determine the cross-sectional information associated with the sample plain film data as sample cross-sectional information, and determine the cross-sectional information associated with the test plain film data as test cross-sectional information;

[0105] In this embodiment, the associated multiple flat film data and multiple sets of cross-sectional information are divided into a training set and a test set according to a preset ratio. The training set is the sample flat film data and the sample cross-sectional information, and the test set is the test flat film data and the test cross-sectional information. The sample set is used to train the model, and the test set is used to test the trained model. Thus, sufficient data support is provided for subsequent training and testing of the target model, enabling the output result of the finally obtained target model to be more consistent with the actual cross-sectional information, which helps to improve the accuracy of scanning.

[0106] Among them, the preset ratio can be reasonably set according to the training and testing requirements of the target model. For example, the ratio of the training set to the test set is 9:1 or 8:2.

[0107] Step 204: Train the neural network model according to the sample flat film data and the sample cross-sectional information;

[0108] Specifically, the neural network (Neural Networks, NN) model can be a convolutional neural network (Convolutional Neural Networks, CNN) model, a deep residual shrinkage network (Deep Residual Shrinkage Network, DRSN) model, a fully connected neural network (Fully Connected Neural Network, FCNN) model, a recurrent neural network (Recurrent Neural Network, RNN) model, or a long short-term memory network (Long Short-Term Memory, LSTM) model. The embodiments of the present application do not make specific limitations.

[0109] Step 205: Evaluate the trained neural network model according to the test flat film data and the test cross-sectional information to obtain an evaluation index;

[0110] Specifically, to calculate the evaluation index of the test flat film data, the following formula is used:

[0111] R(f(z), a) = |f(z) - a|,

[0112] where z represents the test flat film data, f represents the trained neural network model, a represents the number of valid pixels of the cross-sectional image in the test cross-sectional information, and R represents the evaluation index of a single test flat film data.

[0113] It can be understood that if the number of test flat film data in the test set is m, that is, the test flat film data corresponds to each flat film image, and m is a positive integer greater than 1, the average value of the evaluation indexes of the m test flat film data is used as the evaluation index R of the test set mean , as follows:

[0114]

[0115] Specifically, according to the test plain film data and the test section information, the trained neural network model is evaluated, including:

[0116] Step 205-1: Input the test plain film data into the trained neural network model to obtain the to-be-tested section information;

[0117] Step 205-2: Calculate the evaluation index according to the first pixel number of the effective area where the scanned object of the to-be-tested section image in the to-be-tested section information and the second pixel number of the effective area where the scanned object of the test section image in the test section information are located.

[0118] In this embodiment, the sample plain film data and its labeled sample section information are respectively used as input data and labels to train the neural network model to obtain the trained neural network model. Then, the test plain film data is input into the trained neural network model, and the training result is evaluated according to the absolute value of the difference between the first pixel number of the to-be-tested section image output by the trained neural network model and the second pixel number of the test section image associated with the test plain film data. So as to calculate the loss function according to the evaluation index to determine the accuracy of the trained neural network model.

[0119] Step 206: When the evaluation index meets the convergence condition of the loss function, the trained neural network model is confirmed as the target model.

[0120] In this embodiment, when the loss function converges, it indicates that the section image matched by the trained neural network model is almost the same as the test section image. At this time, it can be determined that the training of the neural network model is completed, and the neural network model is output as the target model. So as to match the possible estimated section information of multiple different parts of the target object for the plain film data of the target object through the target model, and then realize obtaining the estimated section information through big data to analyze the body shape of the target object, improve the accuracy of body shape estimation, and at the same time, there is no need to add additional reminder detection equipment, which is beneficial to reducing the equipment cost.

[0121] In actual application scenarios, the loss function can adopt the Mean Square Error (MSE) function, the Mean Absolute Error (MAE) function, the Root Mean Square Error (RMSE) function, the Mean Squared Log Error function, or the Mean Relative Error (MRE) function. Taking the use of the Mean Absolute Error (MAE) as the loss function for neural network model training as an example, the loss function is as follows:

[0122]

[0123] Among them, L represents the loss function, MAE represents the Mean Absolute Error function, and n represents the number of training samples. When the MAE function is reduced to less than 1%, it is determined that the training accuracy of the model meets the standard, and the target model is output.

[0124] In some possible embodiments, after step 202, that is, after associating each row in the radiograph image corresponding to the radiograph data of each sample object and each cross-sectional information in the cross-sectional information of each group of sample objects according to the bed code value, the scanning method further includes:

[0125] Step 207: Determine the long and short axis parameters of the effective region according to the number of pixels of the sample object in the effective region in the cross-sectional image of the cross-sectional information of the sample object;

[0126] Among them, the long and short axis parameters include at least one of the following: the length of the long axis, the length of the short axis, and the ratio of the long axis to the short axis. The effective region where the scanned object is located in the cross-sectional image can be approximately regarded as an ellipse, and the length of the long axis and the length of the short axis of the ellipse can be equivalent to the width and thickness of the scanned object. The effective region may include bone tissue, soft tissue, blood vessels, etc.

[0127] Step 208: Perform an association process on the long and short axis parameters of the effective region and the cross-sectional image.

[0128] In this embodiment, first, the effective region where the sample object in the cross-sectional image in the cross-sectional information is segmented by the boundary threshold, that is, the background region in the cross-sectional image is removed, and then the number of pixels occupied by the sample object in the x and y directions on the cross-sectional image is calculated. According to the ratio of the number of pixels in the effective region to the total number of pixels in the entire cross-sectional image (effective region + background region) and the width and height corresponding to the total number of pixels at the current bed height, the major axis length (width) and minor axis length (thickness) of the effective region at this time are calculated respectively. At the same time, the major-minor axis ratio is determined based on the major axis length and minor axis length, so as to obtain the major-minor axis parameters of the effective region. Then, the major-minor axis parameters are associated with the cross-sectional information. This facilitates quickly distinguishing different parts of the sample object corresponding to different cross-sectional images according to the major-minor axis parameters, and is convenient for analyzing the body type of the target object. Moreover, after training the target model with the cross-sectional information associated with the major-minor axis parameters, the target model can directly obtain the major-minor axis parameters associated with the cross-sectional information from the input plain film data, realizing the rapid and accurate acquisition of the major-minor axis parameters, and further improving the scanning efficiency of the CT device.

[0129] Specifically, taking the calculation of the major axis length as an example, the following formula is used:

[0130]

[0131] Among them, H 长轴 represents the major axis length of the effective region, H 宽 represents the total width corresponding to the current bed height, that is, the scanning width. The total width can be determined according to the parameters of the scanning bed of the CT device. x 有效 represents the number of pixels in the effective region, and x 总 represents the total number of pixels in the cross-sectional image.

[0132] Similarly, the following formula is used to calculate the minor axis length:

[0133]

[0134] Among them, H 短轴 represents the minor axis length of the effective region, and H 高 represents the total height corresponding to the current bed height. The total height can be determined according to the parameters of the scanning bed of the CT device. x 有效 represents the number of pixels in the effective region, and x 总 represents the total number of pixels in the cross-sectional image.

[0135] Step 103: Based on the estimated cross-sectional information, control the CT device to scan the part of the target object to be scanned.

[0136] In this embodiment, the body shape of the target object can be estimated based on the estimated cross-section information, and the final scan is completed by combining the body shape and the estimated attenuation domain information of the estimated cross-section information. Therefore, during the scanning process, appropriate scanning parameters can be set for different parts to be scanned of the same target object, which not only meets the clinical requirements of the reconstructed image but also reduces the damage of X-rays to the target object.

[0137] Further, as a refinement and extension of the specific implementation manner of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, step 103, that is, based on the estimated cross-section information, controlling the CT device to scan the part to be scanned of the target object, specifically includes the following methods:

[0138] Method 1: Adjust the scanning dose of the radiation source of the CT device based on the estimated cross-section information.

[0139] Among them, the scanning dose can be the scanning dose output by the radiation source, the scanning dose received by the target object, or the current value of the radiation source.

[0140] In this embodiment, by calculating the attenuation domain information of the part to be scanned of the target object based on the estimated cross-section information, dose modulation can be realized for different parts during the scanning of different parts of the same target object. On the basis of improving the imaging quality, the damage of X-rays to the human body is reduced. At the same time, the scanning operation steps are streamlined, the time cost is saved, and the adaptability and applicability of the CT device are greatly improved.

[0141] Method 2: When the estimated cross-section information includes an estimated cross-section image, adjust the bed height of the scanning bed of the CT device based on the long and short axis parameters associated with the estimated cross-section image.

[0142] Specifically, the long and short axis parameters include at least one of the following: the length of the long axis, the length of the short axis, and the ratio of the long axis to the short axis. The effective area where the scanned object is located in the cross-section image can be approximately considered as an ellipse, and the length of the long axis and the length of the short axis of the ellipse can be equivalent to the width and thickness of the scanned object.

[0143] In this embodiment, based on the short axis length associated with the estimated cross-section image corresponding to the thickness of the target object and the long axis length corresponding to the width of the target object, the initial bed height of the scanning bed can be adjusted according to the thickness of the target object, so that the longitudinal center of the target object is located at the center of the field of view, ensuring the maximum ray utilization rate and further ensuring the imaging quality of the CT device.

[0144] It can be understood that the scanning method further includes: obtaining the long and short axis parameters associated with the estimated cross-section image based on the estimated cross-section image; or, when inputting the plain film data into the target model, simultaneously obtaining the long and short axis parameters associated with the estimated cross-section image.

[0145] In this embodiment, the estimated cross-sectional image is recognized and processed to identify the long and short axis parameters related to the estimated cross-sectional image by the number of pixels in the effective region where the target object is located in the estimated cross-sectional image. Or when training the target model with the sample cross-sectional information associated with the long and short axis parameters, after inputting the plain film data into the target model, the target model can not only output the estimated cross-sectional information, but also read the related long and short axis parameters through the estimated cross-sectional information. This facilitates the rapid acquisition of the long and short axis parameters associated with the estimated cross-sectional image and provides a reliable data basis for subsequent scan control.

[0146] Further, as Figure 3 shown, in one embodiment, based on the estimated cross-sectional information, the scanning dose of the radiation source of the CT device is adjusted, which specifically includes:

[0147] Step 301, based on the estimated cross-sectional information, obtain the current modulation curve of the part to be scanned of the target object;

[0148] Step 302, adjust the scanning dose based on the current modulation curve.

[0149] In this embodiment, based on the estimated cross-sectional information, the attenuation domain information of different parts to be scanned can be predicted, and based on this, the target current values required for different parts to be scanned can be analyzed. The target current values are used to draw the current modulation curve of the part to be scanned to control the radiation source of the CT device to complete the scan. Thus, by calculating the attenuation domain information of the part to be scanned, a current modulation curve can be determined. When scanning different parts of the same target object, different currents can be used for scanning different parts, which improves the imaging quality and reduces the damage of X-rays to the human body. At the same time, the scanning operation steps are streamlined, the time cost is saved, and the adaptability and applicability of the CT device are greatly improved.

[0150] Further, in the case where the estimated cross-sectional information includes the estimated cross-sectional image, step 301 specifically includes:

[0151] Step 301-1, project the estimated cross-sectional image of the part to be scanned of the target object at multiple radiation positions to determine the estimated attenuation domain information of the part to be scanned of the target object at multiple radiation positions;

[0152] It should be noted that different parts of the target object have their respective corresponding attenuation coefficients, and the attenuation domain information is the integral value of the attenuation coefficient on the path passed by the ray, that is, the value obtained after air correction of the value received by the CT device detector.

[0153] Step 301-2, configure the current modulation curve of the part to be scanned of the target object according to the estimated attenuation domain information.

[0154] In this embodiment, the attenuation domain information of the part to be scanned of the target object in the predicted cross-section information is predicted through projection processing. According to the predicted attenuation domain information of the part to be scanned, the current modulation curves required for all the wire placement positions of the part to be scanned are configured. The radiation source of the CT device is controlled according to the current modulation curves to complete the scan dose modulation. By calculating the attenuation domain information of the part to be scanned to determine a current modulation curve, dose modulation for different parts can be achieved when scanning different parts of the same target object. On the basis of improving the imaging quality, the damage of X-rays to the human body is reduced. At the same time, the scanning operation steps are streamlined, the time cost is saved, and the adaptability and applicability of the CT device are greatly improved.

[0155] Specifically, step 301-2, that is, according to the predicted attenuation domain information, configuring the current modulation curve of the part to be scanned of the target object, specifically includes:

[0156] Step 301-2-a, calculating the target current value of the part to be scanned of the target object at each wire placement position according to the predicted attenuation domain information;

[0157] Step 301-2-b, determining the current modulation curve of the part to be scanned of the target object according to the target current values of the part to be scanned of the target object at each wire placement position.

[0158] In this embodiment, for a part to be scanned of the target object, the target current values required for each wire placement position are calculated according to the predicted attenuation domain information at each of the multiple different wire placement positions. The target current values required for all the wire placement positions corresponding to this part to be scanned constitute the current range of this part to be scanned, that is, the current modulation curve.

[0159] It can be understood that if multiple different parts to be scanned need to be scanned, only by fitting the current modulation curves of each part to be scanned through curve fitting technology, the current modulation curve of the target object that is smooth, continuous and easy to implement can be obtained. As Figure 6 shown, the current modulation curve can be a curve that changes along the scanning direction or the z direction and is composed of the target current values required for each cross-sectional image. The target current values required for each part of each scanning part can be intuitively obtained through the current modulation curve.

[0160] It is worth mentioning that by using the equivalent water phantom corresponding to the cross-sectional image, the target current value of the part to be scanned of the target object at each wire placement position is calculated according to the attenuation domain information, and the following formula is used:

[0161]

[0162] In the formula, mAs ACS represents the target current value, μ waterRepresents the attenuation coefficient of water, D ref Represents the preset equivalent water phantom diameter, μ i D i Represents the attenuation domain information of the target object, that is, the attenuation domain information of bone tissue / soft tissue / air, mAs ref Represents the preset reference current value, and adjCoef represents the preset exponential adjustment parameter.

[0163] It is worth mentioning that before step 301-2, configuring the current modulation curve of the part to be scanned of the target object according to the estimated attenuation domain information, the scanning method further includes:

[0164] Step 301-3, when the CT scan mode is helical scan, calculate the error amount according to the actual attenuation domain information and the estimated attenuation domain information of the already laid wire position of the estimated cross-sectional image;

[0165] Step 301-4, correct the estimated attenuation domain information of the unlaid wire position adjacent to the already laid wire position in the estimated cross-sectional information according to the error amount.

[0166] In this embodiment, if the CT scan mode is tomographic scan, the current modulation curve can be directly used for scanning. If the CT scan mode is helical scan, then according to the actual attenuation domain information and the estimated attenuation domain information of the already laid wire position of the estimated cross-sectional information, and according to the principle of light reversibility, in the rotation direction of the helical scan, correct the estimated attenuation domain information of the unlaid wire position adjacent to the already laid wire position, so as to improve the accuracy of the attenuation domain information and contribute to more accurate scanning.

[0167] Specifically, for example, after the estimated attenuation domain information of the wth View (laid wire position) determined by the target model, after scanning the wth View position, calculate the actual attenuation domain information of the wth View, compare it with the estimated attenuation domain information of the wth View, obtain the relative error of the estimated attenuation, and use this error to correct the estimated attenuation domain information of the un-scanned (w + 1)th view determined by the target model to obtain the corrected estimated attenuation domain information, so as to improve the accuracy of the data to be used. The specific formula is as follows:

[0168] PreAtt(w + 1) = PreAtt(w + 1)′ × (RealAtt(w) / PreAtt(w)′),

[0169] Among them, PreAtt(w + 1) represents the estimated attenuation domain information of the (w + 1)-th View (the unlaid position adjacent to the w-th laid position) after correction, PreAtt(w + 1)' represents the estimated attenuation domain information of the (w + 1)-th View before correction, PreAtt(w)' represents the estimated attenuation domain information of the w-th View before correction, and RealAtt(w) represents the actual attenuation domain information of the w-th View.

[0170] It is worth mentioning that, as Figure 7 shown, each View (laid position) has a total of 5 channels from c1 to c5. According to the shown rotation direction, View i is after View j and View k. At this time, if we want to estimate the maximum channel attenuation of View i, we first need to estimate the attenuation of all channels of View i and then take the maximum value from them. To further improve the metering modulation efficiency, the estimated attenuation domain information of one channel can be used to estimate the estimated attenuation domain information of the channel opposite to it. For example, as Figure 7 shown, since the paths of the channels through the scanned object are similar, the attenuation of the c4 channel of View j can be used as the estimated value (estimated attenuation domain information) of the attenuation of the c2 channel of View i, and the attenuation of the c1 channel of View k can be used as the estimated value of the attenuation of the c5 channel of View i, and the two pairs of channels are symmetric about the central channel. By analogy, the attenuation of other channels can also be estimated in this way.

[0171] In an actual application scenario, before projecting the estimated cross-sectional images of the part to be scanned of the target object according to multiple laid positions, the scanning method further includes: obtaining the part to be scanned of the target object based on the plain film data; or, in the case where the estimated cross-sectional information includes an estimated cross-sectional image, performing part classification processing on the estimated cross-sectional image according to the long and short axis parameters associated with the estimated cross-sectional image to obtain the part to be scanned of the target object.

[0172] In this embodiment, the part to be scanned of the target object can be obtained through the plain film data or the long and short axis parameters associated with the estimated cross-sectional image, so as to facilitate distinguishing different parts of the target object, which helps to set different scanning parameters for different parts, thereby improving the imaging quality of the CT device and reducing the damage of X-rays to the human body. At the same time, it is convenient for users to query the scanning data of different parts, providing convenience for the research and clinical application of the scanning data.

[0173] As Figure 4 shown, in another embodiment, in the case where the estimated cross-sectional information includes an estimated cross-sectional image, based on the estimated cross-sectional information, adjusting the scanning dose of the radiation source of the CT device specifically includes:

[0174] Step 401: Determine the dose ratio between different parts in the to-be-scanned part of the target object based on the long and short axis parameters associated with the estimated cross-sectional image.

[0175] In this embodiment, due to the differences in the long and short axis parameters of the effective regions of the cross-sectional images of different parts of the target object, the parts of the target object corresponding to each estimated cross-sectional image can be quickly distinguished by the long and short axis parameters associated with the cross-sectional images in the estimated cross-sectional information. At the same time, the area of the effective region containing the target object in each estimated cross-sectional image can be calculated based on the long and short axis parameters. The dose ratio between different parts of the target object can be determined by the areas of the effective regions of the target object in different estimated cross-sections, which is beneficial to improving the scanning efficiency.

[0176] Step 402: Adjust the scanning dose based on the dose ratio.

[0177] In this embodiment, when the to-be-scanned part involves multiple different parts, based on the scanning dose required for any one part, the scanning doses for the other parts except any one part among the multiple different parts are configured according to the dose ratio. Thus, when scanning different parts of the same target object, the dynamic modulation of the scanning dose is performed for scanning different parts. On the basis of improving the imaging quality, the damage of X-rays to the human body is reduced. At the same time, the scanning operation steps are streamlined, the time cost is saved, and the adaptability and applicability of the CT device are greatly improved.

[0178] Further, as Figure 5 shown, as a specific implementation of the above scanning method, an embodiment of the present application provides a scanning device 500, which includes an acquisition module 501, a data processing module 502, and a control module 503.

[0179] Among them, the acquisition module 501 is used to acquire the plain film data of the target object; the data processing module 502 is used to input the plain film data into the target model to obtain the estimated cross-sectional information of the target object; the control module 503 is used to control the CT device to scan the to-be-scanned part of the target object based on the estimated cross-sectional information.

[0180] In this embodiment, the sample plain film data of a large number of sample objects is used as input, and the sample cross-section information associated with the sample plain film data is used as a label to train a deep learning network, obtaining a target model based on deep learning. The target model is used to match the plain film data of a target object with the possible estimated cross-section information of multiple different parts of the target object. Through the estimated cross-section information, the body shape of the target object can be estimated, and combined with the body shape and the parameters related to the estimated cross-section information, the scanning of the part to be scanned of the target object is completed. On the one hand, the estimated cross-section information can be obtained using big data, and then the body shape of the target object can be analyzed, improving the accuracy of body shape estimation, and there is no need to add additional detection equipment, which is beneficial to reducing equipment costs. On the other hand, the different parts of the target object can be accurately analyzed through the body shape, so that during the scanning process, appropriate scanning parameters can be set for different parts to be scanned of the same target object, which not only meets the clinical requirements of the reconstructed image but also reduces the damage of X-rays to the target object.

[0181] Further, the control module 503 is specifically configured to adjust the scanning dose of the radiation source of the CT device based on the estimated cross-section information; and / or in the case where the estimated cross-section information includes an estimated cross-section image, adjust the bed height of the scanning bed of the CT device based on the major and minor axis parameters associated with the estimated cross-section image.

[0182] Further, the control module 503 is specifically configured to obtain the current modulation curve of the part to be scanned of the target object based on the estimated cross-section information, and adjust the scanning dose based on the current modulation curve.

[0183] Further, the scanning device 500 further includes: a first determination module (not shown in the figure), where the first determination module is configured to determine the dose ratio between different parts in the part to be scanned of the target object based on the major and minor axis parameters associated with the estimated cross-section image in the case where the estimated cross-section information includes an estimated cross-section image; the control module 503 is specifically configured to adjust the scanning dose based on the dose ratio.

[0184] Further, in the case where the estimated cross-section information includes an estimated cross-section image, the scanning device 500 further includes: a second determination module (not shown in the figure), where the second determination module is configured to perform projection processing on the estimated cross-section image of the part to be scanned of the target object according to multiple radiation positions, determine the estimated attenuation domain information of the part to be scanned of the target object at multiple radiation positions; and configure the current modulation curve of the part to be scanned of the target object according to the estimated attenuation domain information.

[0185] Further, the scanning device 500 further includes: a third determination module (not shown in the figure), where the third determination module is configured to obtain the part to be scanned of the target object based on the plain film data; or, in the case where the predicted cross-sectional information includes a predicted cross-sectional image, perform part classification processing on the predicted cross-sectional image according to the long and short axis parameters associated with the predicted cross-sectional image to obtain the part to be scanned of the target object.

[0186] Further, the second determination module is specifically configured to calculate the target current value of the part to be scanned of the target object at each wire laying position according to the predicted attenuation domain information; determine the current modulation curve of the part to be scanned of the target object according to the target current value of the part to be scanned of the target object at each wire laying position; calculate the target current value of the part to be scanned of the target object at each wire laying position according to the attenuation domain information, using the following formula:

[0187]

[0188] In the formula, mAs ACS represents the target current value, μ water represents the attenuation coefficient of water, D ref represents the preset equivalent water phantom diameter, μ i D i represents the attenuation domain information, mAs ref represents the preset reference current value, and adjCoef represents the preset exponential adjustment parameter.

[0189] Further, the scanning device 500 further includes: a correction module (not shown in the figure), where the correction module is configured to calculate the error amount according to the actual attenuation domain information and the predicted attenuation domain information of the wire laying position of the predicted cross-sectional image in the case where the CT scanning mode is spiral scanning; correct the predicted attenuation domain information of the non-wire laying position adjacent to the wire laying position in the predicted cross-sectional image according to the error amount;

[0190] Correct the predicted attenuation domain information of the non-wire laying position adjacent to the wire laying position in the predicted cross-sectional image according to the error amount, using the following formula:

[0191] PreAtt(w + 1) = PreAtt(w + 1)' × (RealAtt(w) / PreAtt(w)'),

[0192] In the formula, PreAtt(w + 1) represents the predicted attenuation domain information of the non-wire laying position adjacent to the w-th wire laying position after correction, PreAtt(w + 1)' represents the predicted attenuation domain information of the non-wire laying position adjacent to the w-th wire laying position before correction, PreAtt(w)' represents the predicted attenuation domain information of the w-th wire laying position, and RealAtt(w) represents the actual attenuation domain information of the w-th wire laying position.

[0193] Further, the scanning device 500 further includes: an identification module (not shown in the figure), configured to obtain the long and short axis parameters associated with the estimated cross-sectional image based on the estimated cross-sectional image.

[0194] Further, the data processing module 502 is further configured to obtain the long and short axis parameters associated with the estimated cross-sectional image while inputting the plain film data into the target model.

[0195] Further, the obtaining module 501 is further configured to obtain multiple plain film data and multiple groups of cross-sectional information of a sample object, where the cross-sectional information of the sample object includes tomographic cross-sectional information or helical cross-sectional information; the scanning device 500 further includes: an association module (not shown in the figure), and the association module is configured to perform association processing on each row in the plain film image corresponding to the plain film data of each sample object and each cross-sectional information in each group of cross-sectional information of the sample object according to the bed code value; the obtaining module 501 is further configured to divide the plain film data of multiple sample objects into sample plain film data and test plain film data according to a preset ratio, and determine the cross-sectional information associated with the sample plain film data as sample cross-sectional information, and determine the cross-sectional information associated with the test plain film data as test cross-sectional information; the scanning device 500 further includes: a training module (not shown in the figure), and the training module is configured to train the neural network model according to the sample plain film data and the sample cross-sectional information; evaluate the trained neural network model according to the test plain film data and the test cross-sectional information to obtain an evaluation index; and when the evaluation index meets the convergence condition of the loss function, confirm the trained neural network model as the target model.

[0196] Further, the scanning device 500 further includes: a fourth determination module (not shown in the figure), and the fourth determination module is configured to determine the long and short axis parameters of the effective area according to the number of pixels of the sample object in the effective area of the cross-sectional image of the cross-sectional information of the sample object; the association module is further configured to perform association processing on the long and short axis parameters of the effective area and the cross-sectional image.

[0197] Further, the training module is specifically configured to input the test plain film data into the trained neural network model to obtain the cross-sectional information to be tested; calculate the evaluation index according to the first number of pixels of the scanned object in the effective area of the cross-sectional image to be tested in the cross-sectional information to be tested and the second number of pixels of the scanned object in the effective area of the cross-sectional image in the test cross-sectional information.

[0198] For the specific limitations of the scanning device, reference can be made to the limitations of the scanning method in the foregoing text, which will not be elaborated here. Each module in the above scanning device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in or independent of the processor in the CT device in the form of hardware, or stored in the memory of the CT device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0199] Based on the above such as Figures 1 to 4 shown in the method, correspondingly, an embodiment of the present application further provides a readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the above such as Figures 1 to 4 shown in the scanning method.

[0200] Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions to enable a CT device to execute the methods in various implementation scenarios of the present application.

[0201] Based on the above such as Figures 1 to 4 shown in the method, and Figure 5 shown in the virtual device embodiment, in order to achieve the above object, an embodiment of the present application further provides a CT device, which can specifically be a personal computer, a server, a network device, etc. The CT device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to implement the above such as Figures 1 to 4 shown in the scanning method.

[0202] Specifically, as Figure 8 shown, the CT device can include a scanning gantry 510, a scanning bed 520, a console 530, etc.

[0203] The scanning gantry 510 is mainly used to generate X-rays in the scanning hole, collect data, and transmit the collected data to the console 530. The scanning gantry 510 can include a radiation source 511 and a detector 512. The radiation source 511 is used to generate X-rays during the CT scan. The radiation source 511 forms a radiation beam 5114 through an aperture 5112. The radiation beam 5114 can pass through a target object 600 at the central position of the scanning gantry 510. After passing through the target object 600, the radiation beam 5114 hits the detector 512. The detector 512 is arranged at a position opposite to the radiation source on the scanning gantry 510. The detector 512 is divided into multiple layers along the Z direction and multiple receiving channels along the X direction. Thus, the detector 512 is an array-like detection device composed of multiple detection units. The detection units in the same layer are arranged along the X direction, and the detection units in the same receiving channel are arranged along the Z direction.

[0204] The scanning bed 520 is a tool that cooperates with the scanning gantry 510 to complete the scanning task. It is used to support the scanning object 600, position and control the target object 600, control the up and down movement of the scanning bed, and move in and out of the scanning hole. Generally, the moving direction of the scanning bed 520 is the Z-axis direction.

[0205] The console 530 may include an input panel through which the user can input relevant information. The console 530 may include a processor, a memory, and a display. Computer instructions and data are stored in the memory, and the processor can read the computer program instructions from the memory into the memory for operation to implement the scanning method in the embodiments of the present application.

[0206] It should be noted that the console 530 may further include at least one of the following: a touch panel, an emergency stop button, a speaker, data and subject information input, scanning parameter setting, image reconstruction and display, a tape drive, and photographic control, etc. Among them, some or all of the processing functions in the console 530 can be set on the scanning gantry 510, and some or all of the processing functions on the scanning gantry 510 can also be set on the console 530. When all the processing functions in the console 530 are set on the scanning gantry 510 or all the processing functions on the scanning gantry 510 are set on the console 530, the console 530 and the scanning gantry 510 can be integrated. That is to say, the scanning method in the embodiments of the present invention can be executed by the console 530, or by the scanning gantry 510, or by the cooperation of the console 530 and the scanning gantry 510. For example, the console 530 and the scanning gantry 510 are used as two parts of a distributed system.

[0207] It can be understood that the console 530 includes the scanning device provided in the above embodiments.

[0208] When using Figure 8 the CT device shown to scan the target object 600, the scanning bed 520 controls the target object 600 to enter the scanning hole. The scanning gantry 510 rotates around the target object 600 as the scanning bed 520 moves along the Z-axis, and the X-ray starts to scan the target object 600. Further, during the spiral scan of the CT device, the radiation source 511, the aperture 5112, and the detector 512 rotate around the rotation axis. During the rotation, the radiation position (also known as View) of the radiation source 511 changes continuously. Each radiation position corresponds to multiple receiving channels on the detector 512.

[0209] Those skilled in the art can understand that the structure of a CT device provided in this embodiment does not limit the CT device, and it may include more or fewer components, or combine some components, or have different component arrangements.

[0210] The storage medium may further include an operating system and a network communication module. The operating system is a program for managing and storing the hardware and software resources of the CT device, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to implement communication between components inside the storage medium, as well as communication with other hardware and software in the entity device.

[0211] Through the description of the above embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or the flat-panel data of the target object can be obtained through hardware; the flat-panel data is input into the target model to obtain the estimated cross-sectional information of the target object; based on the estimated cross-sectional information, the CT device is controlled to scan the part to be scanned of the target object. In one aspect of the embodiments of the present application, it is possible to use big data to obtain the estimated cross-sectional information, and then analyze the body shape of the target object, improving the accuracy of body shape estimation, and there is no need to add additional detection equipment, which is beneficial to reducing equipment costs. On the other hand, different parts of the target object can be accurately analyzed through the body shape, so that during the scanning process, appropriate scanning parameters can be set for different parts to be scanned of the same target object, which not only meets the clinical requirements of the reconstructed image, but also reduces the damage of X-rays to the target object.

[0212] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the drawings are not necessarily essential for implementing the present application. Those skilled in the art can understand that the modules in the device in the implementation scenario can be distributed in the device in the implementation scenario according to the description of the implementation scenario, or can be correspondingly changed and located in one or more devices different from this implementation scenario. The modules in the above implementation scenario can be combined into one module, or further split into multiple sub-modules.

[0213] The above serial numbers of the present application are only for description and do not represent the advantages or disadvantages of the implementation scenarios. The above disclosure is only several specific implementation scenarios of the present application. However, the present application is not limited thereto, and any changes that can be thought of by those skilled in the art should fall within the protection scope of the present application.

Claims

1. A scanning method, characterized in that, The method includes: Obtaining the plain film data of the target object; Inputting the plain film data into the target model to obtain the predicted cross-sectional information of the target object. The target model is trained based on the multiple plain film data and multiple groups of cross-sectional information of the sample objects. For each sample object, each row in the plain film image corresponding to the plain film data and each cross-sectional information in each group of the cross-sectional information of the sample object are associated based on the bed code value; Based on the predicted cross-sectional information, controlling the CT device to scan the part to be scanned of the target object.

2. The scanning method according to claim 1, characterized in that The controlling the CT device to scan the part to be scanned of the target object based on the predicted cross-sectional information includes: Based on the predicted cross-sectional information, adjusting the scanning dose of the radiation source of the CT device; and / or, When the predicted cross-sectional information includes a predicted cross-sectional image, based on the major and minor axis parameters associated with the predicted cross-sectional image, adjusting the bed height of the scanning bed of the CT device.

3. The scanning method according to claim 2, characterized in that, The adjusting the scanning dose of the radiation source of the CT device based on the predicted cross-sectional information includes: Based on the predicted cross-sectional information, obtaining the current modulation curve of the part to be scanned of the target object, and adjusting the scanning dose based on the current modulation curve; and / or, When the predicted cross-sectional information includes a predicted cross-sectional image, determining the dose ratio between different parts in the part to be scanned of the target object based on the major and minor axis parameters associated with the predicted cross-sectional image, and adjusting the scanning dose based on the dose ratio.

4. The scanning method according to claim 3, wherein, When the predicted cross-sectional information includes a predicted cross-sectional image, the obtaining the current modulation curve of the part to be scanned of the target object based on the predicted cross-sectional information includes: Performing projection processing on the predicted cross-sectional image of the part to be scanned of the target object according to multiple radiation positions to determine the predicted attenuation domain information of the part to be scanned of the target object at the multiple radiation positions; Configuring the current modulation curve of the part to be scanned of the target object according to the predicted attenuation domain information.

5. The scanning method according to claim 4, wherein Before performing the projection processing on the predicted cross-sectional image of the part to be scanned of the target object according to multiple radiation positions, the method further includes: Based on the plain film data, obtaining the part to be scanned of the target object; or, When the predicted cross-sectional information includes a predicted cross-sectional image, performing part classification processing on the predicted cross-sectional image according to the major and minor axis parameters associated with the predicted cross-sectional image to obtain the part to be scanned of the target object.

6. The scanning method according to claim 4, wherein The configuring the current modulation curve of the part to be scanned of the target object according to the predicted attenuation domain information includes: Calculating the target current value of the part to be scanned of the target object at each radiation position according to the predicted attenuation domain information; Determining the current modulation curve of the part to be scanned of the target object according to the target current value of the part to be scanned of the target object at each radiation position; The calculating the target current value of the part to be scanned of the target object at each radiation position according to the attenuation domain information uses the following formula: ， In the formula, represents the target current value, represents the attenuation coefficient of water, represents the preset equivalent water phantom diameter, represents the attenuation domain information, represents the preset reference current value, represents the preset exponential adjustment parameter.

7. The scanning method according to claim 4, wherein Before configuring the current modulation curve of the part to be scanned of the target object according to the predicted attenuation domain information, the method further includes: When the CT scan mode is spiral scan, calculate an error amount according to the actual attenuation domain information of the actual wire placement position in the predicted cross-sectional image and the predicted attenuation domain information; Correct the predicted attenuation domain information of the un-wired position adjacent to the wired position in the predicted cross-sectional image according to the error amount; The correction of the predicted attenuation domain information of the un-wired position adjacent to the wired position in the predicted cross-sectional image according to the error amount is performed using the following formula: PreAtt ( w +1)= PreAtt ( w +1)'×( RealAtt ( w ) / PreAtt ( w )'), In the formula, PreAtt ( w +1) represents the estimated attenuation domain information of the unlaid position adjacent to the w-th laid position after correction, PreAtt ( w +1)ʹ represents the estimated attenuation domain information of the unlaid position adjacent to the w-th laid position before correction, PreAtt ( w ) ʹ represents the estimated attenuation domain information of the w-th laid position, RealAtt ( w ) represents the actual attenuation domain information of the w-th laid position.

8. The scanning method according to any one of claims 2 to 7, characterized in that, The method further includes: Based on the predicted cross-sectional image, obtain the long and short axis parameters associated with the predicted cross-sectional image; or, When inputting the flat film data into the target model, simultaneously obtain the long and short axis parameters associated with the predicted cross-sectional image.

9. The scanning method according to any one of claims 1 to 7, characterized in that The method further includes: Obtain multiple flat film data and multiple groups of cross-sectional information of the sample object, where the cross-sectional information of the sample object includes tomographic cross-sectional information or spiral cross-sectional information; Perform an association process on each row in the flat film image corresponding to the flat film data of each sample object and each cross-sectional information in each group of cross-sectional information of the sample object according to the bed code value; Divide the flat film data of multiple sample objects into sample flat film data and test flat film data according to a preset ratio, and determine the cross-sectional information associated with the sample flat film data as sample cross-sectional information, and determine the cross-sectional information associated with the test flat film data as test cross-sectional information; Train a neural network model according to the sample flat film data and the sample cross-sectional information; Evaluate the trained neural network model according to the test flat film data and the test cross-sectional information to obtain evaluation indicators; When the evaluation indicators meet the convergence conditions of the loss function, confirm the trained neural network model as the target model.

10. The scanning method according to claim 9, wherein The method further includes: Determine the long and short axis parameters of the effective area according to the number of pixels in the effective area where the sample object is located in the cross-sectional image of the cross-sectional information of the sample object; Perform an association process on the long and short axis parameters of the effective area and the cross-sectional image.

11. A scanning device, characterized in that, The device includes: An acquisition module, configured to acquire the flat film data of the target object; A data processing module, configured to input the flat film data into a target model to obtain the predicted cross-sectional information of the target object, where the target model is trained according to multiple flat film data and multiple groups of cross-sectional information of the sample object, and each row in the flat film image corresponding to the flat film data of each sample object and each cross-sectional information in each group of cross-sectional information of the sample object are associated based on the bed code value; A control module, configured to control the CT device to scan the part to be scanned of the target object based on the predicted cross-sectional information.

12. A readable storage medium, on which a program or instructions are stored, characterized in that, When the program or instruction is executed by a processor, it implements the steps of the scanning method according to any one of claims 1 to 10.

13. A CT device, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the program, it implements the scanning method described in any one of claims 1 to 10.

Images