Scanning method, apparatus and device

Abstract

The application discloses a scanning method, device and equipment, mainly reducing the sampling frequency of wire laying through sparse angles, and lowering the scanning dose under the premise that the image quality meets the requirements. The method comprises the following steps: determining the sparse angle sampling number N of each sampling position of a to-be-scanned object in a scanning direction; determining the exposure index of M optional exposure positions of an XY scanning section corresponding to each sampling position; screening N target exposure positions from the M exposure positions according to the exposure index, and scanning the to-be-scanned object.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of CT scan dose modulation, and in particular to a scanning method, device and equipment. BACKGROUND

[0002] Computed Tomography (CT) technology has been widely used in the medical technology field and has become an extremely common diagnostic method. With the increasing frequency of CT scanning, the X-ray radiation dose received by the scanned object during the diagnosis and treatment process is also increasing. Excessive radiation dose can increase the risk of cancer or other serious diseases for the scanned object. With the development of CT scanning technology, how to reduce the X-ray dose while ensuring image quality has become a focus topic.

[0003] Compared with conventional CT scanning technology, sparse angle sampling reduces the sampling frequency of the ray, which can reduce the received dose of the scanned object by several times, and becomes a very promising development direction in CT scanning technology. The ray position of sparse angle sampling affects the image quality, and how to select the ray position has become a research hotspot in the industry. SUMMARY

[0004] The present application provides a scanning method, device and equipment, which can reduce the sampling frequency of the ray by sparse angle, ensure the image quality while reducing the scanning dose.

[0005] According to one aspect of the present application, a scanning method is provided, which comprises:

[0006] determining the number N of sparse angle sampling of each sampling position of the scanned object in the scanning direction;

[0007] determining the exposure index of the M selectable exposure positions of the XY scan section corresponding to each sampling position, wherein the XY scan section is the cross section of the scanned object perpendicular to the scanning direction at each sampling position, M is the conventional number of CT tube exposure, and M>N;

[0008] According to the exposure index, N target exposure positions are selected from the M exposure positions, and the scanned object is scanned.

[0009] According to another aspect of the present application, a scanning device based on sparse angle sampling is provided, which comprises:

[0010] a first determination module for determining the number N of sparse angle sampling of each sampling position of the scanned object in the scanning direction;

[0011] determining exposure indexes of optional M exposure positions of an XY scanning section corresponding to each sampling position, wherein the XY scanning section is a section of the object to be scanned perpendicular to the scanning direction at the sampling position, M is a regular number of CT tube exposure, and M>N;

[0012] scanning the object to be scanned according to the N target exposure positions selected from the M exposure positions according to the exposure indexes.

[0013] According to still another aspect of the present application, a computer readable storage medium is provided, which stores a computer program, and the program is executed by a processor to implement the following steps:

[0014] determining a sparse angle sampling number N of each sampling position of the object to be scanned in a scanning direction;

[0015] determining exposure indexes of optional M exposure positions of an XY scanning section corresponding to each sampling position, wherein the XY scanning section is a section of the object to be scanned perpendicular to the scanning direction at the sampling position, M is a regular number of CT tube exposure, and M>N;

[0016] scanning the object to be scanned according to the N target exposure positions selected from the M exposure positions according to the exposure indexes.

[0017] According to still another aspect of the present application, a computer device is provided, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the following steps when executing the program:

[0018] determining a sparse angle sampling number N of each sampling position of the object to be scanned in a scanning direction;

[0019] determining exposure indexes of optional M exposure positions of an XY scanning section corresponding to each sampling position, wherein the XY scanning section is a section of the object to be scanned perpendicular to the scanning direction at the sampling position, M is a regular number of CT tube exposure, and M>N;

[0020] scanning the object to be scanned according to the N target exposure positions selected from the M exposure positions according to the exposure indexes.

[0021] By the technical scheme, the application provides a scanning method, device and equipment, which can firstly determine the sparse angle sampling number N of each sampling position of a to-be-scanned object in a scanning direction; meanwhile, the exposure index of each sampling position corresponding to M exposure positions of an XY scanning section can be determined, wherein the XY scanning section is a to-be-scanned object section perpendicular to the scanning direction at each sampling position, M is a conventional number of CT tube exposure, and M>N; further, N target exposure positions can be selected from the M exposure positions according to the exposure index, and the to-be-scanned object is scanned based on the N target exposure positions. Compared with the conventional sampling density, the technical scheme in the application can ensure accurate control and reduce the sampling frequency of the exposure line by determining the sparse angle sampling number and the target exposure position under the sparse angle sampling number, so that the image quality meets the requirements under the premise of reducing the scanning dose, and the purpose of reducing the radiation dose of the to-be-scanned object can be achieved with high efficiency.

[0022] The above description is only a summary of the technical scheme of the application, in order to more clearly understand the technical means of the application, the application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. BRIEF DESCRIPTION OF DRAWINGS

[0023] The drawings described herein are used to provide further understanding of the application, and form a part of the application. The schematic embodiments of the application and the description thereof are used to explain the application, and do not constitute an improper limitation on the application. In the drawings:

[0024] Figure 1 A flowchart of a scanning method provided by an embodiment of the application is shown;

[0025] Figure 2 A flowchart of another scanning method provided by an embodiment of the application is shown;

[0026] Figure 3 A principle diagram of a scanning based on sparse angle sampling provided by an embodiment of the application is shown;

[0027] Figure 4 Another principle diagram of a scanning based on sparse angle sampling provided by an embodiment of the application is shown;

[0028] Figure 5 A flat film and axial scan image diagram provided by an embodiment of the application is shown;

[0029] Figure 6 A structure diagram of a scanning device based on sparse angle sampling provided by an embodiment of the application is shown;

[0030] Figure 7 Fig. 2 shows a structural schematic diagram of another scanning device based on sparse angle sampling provided by an embodiment of the application;

[0031] Figure 8 Fig. 3 shows a physical structural schematic diagram of a computer device provided by an embodiment of the application. DETAILED DESCRIPTION

[0032] Hereinafter, the application will be described in detail with reference to the accompanying drawings and in conjunction with the embodiments. It should be noted that the embodiments in the application and the features in the embodiments can be combined with each other without conflict.

[0033] Some terms involved in the application are explained as follows:

[0034] view: represents a projection sampling. When scanning, a CT machine divides each scanning circle into multiple parts (generally more than 1000), and integrates and normalizes the X-ray photons in each part to obtain a projection sampling.

[0035] Flat film: the ball tube is stationary at 0°, 90°, 180° or 270° without rotation, only the scanning bed of the CT machine moves in the scanning direction, and the ball tube samples, which is generally used for positioning the interested part before tomography or spiral scanning.

[0036] In order to reduce the sampling frequency and ensure that the image quality meets the requirements, the application provides a scanning method, which provides a feasible sampling position determination scheme for sparse angle sampling, as shown in Fig. 1, the method comprises the following steps. Figure 1

[0037] 101, determining the number N of sparse angle samplings of the object to be scanned at each sampling position in the scanning direction.

[0038] For this embodiment, the number N of sparse angle samplings of the object to be scanned at each sampling position in the scanning direction can be determined according to the pre-scanning data. The object to be scanned can include a person, an animal or a phantom, etc. The pre-scanning data is medical data that can reflect the corresponding scanning information of the object to be scanned, which can include flat film data, ultra-low dose pre-scanning data, the sampling current of which is less than 50 mA, and the pre-scanning data includes the attenuation information of X-rays passing through the object to be scanned. The scanning direction is the direction from the head to the foot of the object to be scanned or the direction from the foot to the head of the object to be scanned, and the scanning direction is also the advancing and retreating direction of the scanning bed.

[0039] ​For the X-ray CT machine device, its main components include X-ray tube, X-ray detector, gantry, scan bed, etc.; the X-ray tube is the emission source of X-ray, emitting X-ray from the focal point position of the tube, reaching the detector through the object to be scanned, the detector receiving X-ray and converting it into an energy intensity signal for subsequent processing, the X-ray tube and the detector are installed on the gantry, driven by the gantry to rotate around the center of rotation, which is a geometric center determined by the CT machine design. During the CT scanning process, the object to be scanned lies on the scan bed, and the scan bed drives the object to be scanned to move, realizing the detection of different parts of the object to be scanned. With the increasing frequency of CT scanning, the X-ray radiation dose received by the object to be scanned during diagnosis and treatment also increases. Excessive radiation dose will increase the risk of cancer or other serious diseases for the object to be scanned. With the popularity and development of X-ray CT scanning, how to reduce the X-ray dose while ensuring image quality has become a focus.

[0040] One of the methods to reduce the X-ray dose is to apply a matching dose according to the size of the attenuation value of the object to be scanned, apply a large dose for large attenuation, and apply a small dose for small attenuation, so that the image noise can meet the needs of diagnosis, and the image noise is an important evaluation index of image quality. Since the attenuation value of the object to be scanned changes with the scanning direction position and the rotation position (tube angle) in one scan, the X-ray dose needs to be modulated in real time according to the scanning direction (i.e. the direction of the bed movement) position and the rotation position of the tube. The adjustment of the X-ray dose is generally realized by adjusting the mA value of the CT tube. The adjustment mainly includes two steps: first, estimating the attenuation value of the human body, and second, calculating the appropriate tube mA according to the attenuation value.

[0041] For this embodiment, taking the pre-scan data as a flat film data as an example, the sparse view sampling technology can be used to calculate the sparse view (sparse angle sampling number N) corresponding to the sampling position of each 1mm thick slice within the flat film range in the scanning direction based on the flat film data. Compared with the conventional sampling density, the sparse view reduces the sampling frequency of the radiograph, which can reduce the received dose of the object to be scanned by several times. For this embodiment, the sparse angle can be considered, mainly by estimating the sparse view number of each sampling position, and based on the estimation result, the target exposure position under the sparse view number and the determination of the scanning dose at the target exposure position are performed, so that the scanning dose is reduced under the premise that the image quality meets the requirements.

[0042] The execution entity for embodiments of this application can be a scanning system based on sparse angle sampling. Based on estimation, the number of sparse sampling views, the sampling positions of different sparse sampling views, and the dose can be determined. Specifically, after determining the number N of sparse angle samples at each sampling position of the object to be scanned along the scanning direction, the exposure index of M selectable exposure positions on the XY scanning cross-section corresponding to each sampling position can be determined. Further, based on the exposure index, N target exposure positions are selected from the M exposure positions. Finally, the object to be scanned is scanned based on the N target exposure positions.

[0043] 102. Determine the exposure index of the M selectable exposure positions for each sampling position on the XY scan section.

[0044] Wherein, the XY scan section is the cross-section of the object to be scanned perpendicular to the scanning direction at each sampling position, M is the conventional number of CT tube exposures, M>N, and N corresponds to the minimum number of samples at the sampling position. The specific values ​​of M and N can be determined according to the actual application scenario, and are not specifically limited here.

[0045] In specific application scenarios, such as Figure 3 As shown, assuming M exposures are performed for each sampling position, resulting in M ​​exposure positions, if the X-ray tube rotates once, including P sampling positions, then there are a total of P×M exposure positions in one rotation, where P is greater than 1. The M exposure positions are either uniformly distributed or non-uniformly distributed, i.e., equal-density sampling or variable-density sampling. No specific limitations are imposed here. As an optional method, to facilitate the selection of N target exposure positions with higher weights (higher weights increase the probability of being selected as laying positions) from the M exposure positions, and to ensure that the selected target exposure positions have a uniform exposure effect, M can be... The exposure positions are evenly distributed at the same exposure angle interval on each sampling position. For example, if M = 360, then the exposure angles corresponding to the M exposure positions are 0, 1, 2, 3...359. Alternatively, the M exposure positions can be irregularly distributed on the rotating scanning circle of the X-ray tube at each sampling position according to different exposure angles. For example, if M = 360, then the exposure angles corresponding to the M exposure positions can be set according to the actual exposure conditions. For example, they can be 0.5, 1.7, 2.6, 3.4...359.3.

[0046] For the present embodiment, the optional M exposure positions of each sampling position corresponding to the XY scan section can be first determined, and the M exposure positions are respectively indexed based on the specific data feature dimension, and the exposure index corresponding to the M exposure positions can be obtained. The exposure index is an evaluation index value capable of calculating the exposure weight value corresponding to each exposure position, such as exposure weight array, exposure score, exposure influence degree, etc., which is not specifically limited herein. In the following embodiment steps in the present application, the exposure index is taken as an example of the exposure weight array to illustrate the technical solutions in the present application, but it does not constitute a specific limitation on the technical solutions in the present application. For example, when the exposure index corresponds to the exposure weight array, the M exposure weight values corresponding to the M exposure positions of the same sampling position are included in the exposure weight array, the sum of the M exposure weight values in the same exposure weight array can be 1, and they are arranged in descending or ascending order according to the exposure angle, such as Wi(i=1 / 2 / 3 / 4 / 5……M).

[0047] 103、According to the exposure index, N target exposure positions are selected from the M exposure positions, and the object to be scanned is scanned.

[0048] For the present embodiment, the optional M exposure positions of each sampling position corresponding to the XY scan section can be first determined, and the M exposure positions are respectively indexed based on the specific data feature dimension, and the exposure index corresponding to the M exposure positions can be obtained. The exposure index is an evaluation index value capable of calculating the exposure weight value corresponding to each exposure position, such as exposure weight array, exposure score, exposure influence degree, etc., which is not specifically limited herein. In the following embodiment steps in the present application, the exposure index is taken as an example of the exposure weight array to illustrate the technical solutions in the present application, but it does not constitute a specific limitation on the technical solutions in the present application. For example, when the exposure index corresponds to the exposure weight array, the M exposure weight values corresponding to the M exposure positions of the same sampling position are included in the exposure weight array, the sum of the M exposure weight values in the same exposure weight array can be 1, and they are arranged in descending or ascending order according to the exposure angle, such as Wi(i=1 / 2 / 3 / 4 / 5……M).

[0049] It should be noted that for steps 102 and 103 of the embodiment, the purpose is to select N sparse target exposure positions from M exposure positions by sparse angle sampling, and the implementation manner is not limited to using the weight-based equation to obtain an optimal solution, but can also be obtained by other methods, which is not limited herein. In the embodiment of the present application, the technical solution is described by taking the weight-based assignment screening method as an example, which does not constitute a specific limitation on the technical solution in the present application.

[0050] By the scanning method in the embodiment, the sparse angle sampling number N of each sampling position of the object to be scanned in the scanning direction can be determined first, wherein the scanning direction is from the head to the foot of the object to be scanned or from the foot to the head of the object to be scanned; at the same time, the exposure index of the optional M exposure positions of the XY scanning section corresponding to each sampling position can be determined, wherein the XY scanning section is the cross section of the object to be scanned perpendicular to the scanning direction at each sampling position, M is the conventional number of CT tube exposure, and M>N; further, N target exposure positions can be selected from the M exposure positions according to the exposure index, and the object to be scanned is scanned based on the N target exposure positions. Compared with the conventional sampling density, the technical solution in the present application can determine the sparse angle sampling number and the target exposure position under the sparse angle sampling number, ensure accurate control while reducing the sampling frequency of the exposure line, reduce the scanning dose under the premise that the image quality meets the requirements, and efficiently achieve the purpose of reducing the radiation dose of the object to be scanned.

[0051] Further, as a refinement and expansion of the specific implementation manner of the above embodiment, in order to completely describe the implementation manner of the embodiment, the embodiment also provides another scanning method, as shown in Figure 2 The method comprises:

[0052] 201. Based on the pre-scanning data of the object to be scanned, obtain scanning information, create a sparse function based on the function constraint condition corresponding to the scanning information, and calculate the sparse angle sampling number N of each sampling position of the object to be scanned in the scanning direction by using the sparse function.

[0053] For example, when the pre-scanning data is a plain film data, the scanning information of the object to be scanned can include at least one of the scanning site size, the scanning site name, and the scanning eccentric distance. In a specific application scenario, when the scanning information is extracted according to the plain film data, the attenuation area of each sampling position can be calculated every 1 mm (the interval is not limited) along the in-out bed direction by scanning the plain film, and the scanning site size Size of the object to be scanned is estimated. iMeanwhile, the scanning part and the off-centre condition can be recognized through the flat film image (or other methods such as external camera), and accordingly, when extracting the scanning information of the object to be scanned according to the flat film data, the embodiment steps can specifically include: calculating the attenuation area of each sampling position of the object to be scanned in the scanning direction based on the flat film data, and estimating the scanning part size of the object to be scanned according to the attenuation area; inputting the flat film data into the pre-trained part prediction model to determine the predicted scanning part name of the object to be scanned; and identifying the flat film data by using automatic positioning to determine the off-centre distance of the object to be scanned in the X and Y directions relative to the scanning center.

[0054] In the calculation of the attenuation area of each sampling position of the object to be scanned in the scanning direction, the scanning flat film can be used, for example, the frontal flat film (the ball tube is directly above, 0°, and the angle increases clockwise), Figure 4 As shown in the figure, the scanning range is a-d, the flat film data is sampled along the Z direction (the advancing and retreating bed direction), each sampling is a cross section, the attenuation area of each cross section is calculated, and the equivalent water phantom diameter of the cross section is confirmed by the equivalent area. The definitions of the equivalent attenuation area and the equivalent water phantom diameter are as follows:

[0055]

[0056]

[0057] D scan = 2 * sqrt(mean(S) / (PI * mu water ))

[0058] In the formula, mu i l i , mu i+1 l i+1 is the cross section data corresponding to the adjacent sampling position, mu i is the average attenuation coefficient of the i-th detection channel, l i is the attenuation path of the i-th detection channel, N is the total number of detection channels, R is the distance from the positioning position to the detector, N is the number of detectors covered by the positioning position, and alpha is the fan angle of the positioning position.

[0059] After the scanning information of the to-be-scanned object is extracted, a sparse function can be created based on the function limitation condition corresponding to the scanning information. For example, the scanning information includes the scanning site size, the scanning site name, and the scanning off-center distance of the to-be-scanned object. The function limitation condition corresponding to the scanning site size can be that the larger the scanning site is, the more sampling views are. The function limitation condition corresponding to the scanning site name can be that the sampling number can be reduced for some dose-sensitive sites such as eyeballs and thyroid glands. The function limitation condition corresponding to the scanning off-center distance can be that the higher the off-center degree is, the more the sampling number can be increased. In a specific application scenario, the corresponding sampling view interval range can be set for the function limitation condition corresponding to different scanning information. Then, the sparse function can be constructed according to the function limitation condition corresponding to multiple scanning information, such as N=f(Size, Part, Off-centre). Then, the sparse angle sampling number N that meets the sampling view interval range set by the function limitation condition can be calculated according to the sparse function.

[0060] It should be noted that the scanning information of the to-be-scanned object is not limited to the scanning site size, the scanning site name, and the scanning off-center distance. Other types of scanning information that can be implemented can also be included. In the steps of the embodiment, the scanning information of the to-be-scanned object includes at least one of the scanning site size, the scanning site name, and the scanning off-center distance. The technical solutions in the present application are described, but this does not constitute a specific limitation on the technical solutions in the present application.

[0061] 202, determine the exposure index of the M exposure positions of the XY scanning section corresponding to each sampling position.

[0062] For this embodiment, when calculating the exposure index corresponding to the M exposure positions, taking the exposure index as an exposure weight array for example, the shape of the scanned object at each sampling position in the scanning direction can be estimated by AI or other methods, and then the M interest values can be obtained by scanning one circle at this position (assuming M exposures are performed in one circle). The interest values can be the maximum attenuation value, the average attenuation value, etc., which are not limited herein. M here is the conventional number of CT tube exposures, M > N ≥ the minimum sampling number of the corresponding sampling positions. Further, based on the M interest values, weights can be assigned to the M exposure positions to obtain the exposure weight array corresponding to the first exposure index: Wi(i = 1 / 2 / 3 / 4 / 5…M). Considering the influence of the rays, higher weights can be assigned to positions with large attenuation (large interest values), and the higher the weight, the higher the possibility of being selected as a ray position. In addition, as another aspect, the weights of the M exposure positions can also be set based on the distance of the M exposure positions from the scanned object to obtain the exposure weight array corresponding to the second exposure index: Li(i = 1 / 2 / 3 / 4 / 5…M). The farther the distance from the ray source, the greater the weight, and the higher the weight, the higher the possibility of being selected as a ray position. As yet another aspect, the weights of the M exposure positions can also be assigned based on the actual imaging method restrictions to obtain the exposure weight array corresponding to the third exposure index: Gi(i = 1 / 2 / 3 / 4 / 5…M). Due to different reconstruction methods, the importance of data at each angle is different. The sampling angles need to be weighted G according to the characteristics of the own reconstruction method. The weight is related to the sampling angle, denoted as G(angle). For example, when the sparse angle sampling number N is determined to be 3, taking the first exposure position (such as 0°) as the reference, the other two target exposure positions are at sampling angles of 120° and 240°, which can ensure the uniformity of the tube exposure. Therefore, a larger weight G value can be set for the sampling angles 0°, 120° and 240°, and a smaller weight G can be set for other sampling angles.

[0063] Correspondingly, the embodiment steps can specifically include: determining the influence factors of the exposure index corresponding to the M exposure positions; determining the exposure index corresponding to the M exposure positions based on the influence factors of the exposure index corresponding to the M exposure positions, wherein the influence factors of the exposure index include at least one of exposure attenuation, exposure distance, and exposure angle, and correspondingly, the exposure index includes at least one of the first exposure index, the second exposure index, and the third exposure index.

[0064] The M exposure positions correspond to exposure indexes based on the type of exposure index corresponding to the M exposure positions, including: inputting the pre-scanning data of the to-be-scanned object into the trained decay value estimation model, obtaining the exposure decay values corresponding to the M exposure positions at each sampling position, determining the first exposure index corresponding to the M exposure positions based on the exposure decay values; respectively determine the exposure distance value of the M exposure positions and the to-be-scanned object, and determine the second exposure index corresponding to the M exposure positions according to the exposure distance value; determine the first exposure position in the M exposure positions, and determine the third exposure index corresponding to the M exposure positions according to the exposure angles of the remaining M-1 exposure positions relative to the first exposure position.

[0065] In a specific application scenario, before inputting the pre-scanning data of the to-be-scanned object into the trained decay value estimation model to obtain the exposure decay values corresponding to the M exposure positions at each sampling position, the decay value estimation model needs to be trained in advance to make the decay value estimation model meet the preset training standard, and the trained decay value estimation model can be used to obtain the exposure decay values corresponding to the M exposure positions at each sampling position. The exposure decay value can be a maximum decay value, an average decay value, etc., which is not limited here.

[0066] Wherein, in training the decay value estimation model, the attached Figure 3 , specifically can include the following training steps:

[0067] Step 1: Experimental data acquisition and preliminary processing

[0068] Obtain a large amount of head data of clinical scanning (the method of other parts is the same, taking the head as an example), which includes flat data (obtained by plain scan, i.e. positioning sheet) and scanning data (which can be tomographic data or spiral data, this paper takes tomographic data as an example), respectively denoted as PltUOri and ScanUOri;

[0069] Step 2: Constructing flat and scanning data relationship

[0070] Firstly, the start and end couches of the film are obtained from the film data, denoted as PltStartCouch and PltEndCouch respectively, and the start and end couches of the scan are obtained from the scan data, denoted as ScanStartCouch and ScanEndCouch respectively. The interval length is 1mm (or other length) and the sampling points are SN, and the sampling positions are i, i = 1, 2, 3, 4…SN. The PltUOri is projected to the SN points according to the distance relationship, and finally the attenuation curves of the SN positions are obtained, denoted as PltU.

[0071] The maximum attenuation values of the M views at each position are obtained, denoted as MaxU(j), j = 1, 2, 3…M.

[0072] Step 3: Training set data acquisition

[0073] The combination of PltU and MaxU(j) is input as the training set, and the specific description is as follows:

[0074] PltU contains SN attenuation curves, denoted as PltU(i), i = 1, 2, 3…SN, and PltU(i) represents the attenuation curve at position i, containing K data (K is the number of detector channels).

[0075] Corresponding to MaxU(i, j), j = 1, 2, 3…M, represents the maximum attenuation of each of the M views at position i (or other characteristic value, not limited), containing M data.

[0076] PltU(i) and MaxU(i, j) are in one-to-one correspondence.

[0077] Step 4: Data training

[0078] The deep learning network is denoted as net w , where W is the weight in the network.

[0079] The training process is as follows

[0080] The PltU data is input into the network to obtain MaxU′:

[0081] MaxU′(i) = net W (PltU(i-n, i-n+1…, i-1, i, i+1…i+n-1, i+n))

[0082] The loss function is denoted as f loss , the training loss is loss, and the smaller the difference between MaxU'(i) and MaxU(i) is, the smaller the loss loss is. The training loss loss is:

[0083] loss=f loss (MaxU'(i), MaxU(i))

[0084] The W in net W is trained by using the BP algorithm to make the loss as small as possible. Through repeated iteration, MaxU'(i) is finally extremely close to MaxU(i), that is, the process of predicting MaxU(i,j) through PltU(i) is realized, and the training is completed.

[0085] In a specific application scenario, when training the attenuation value estimation model, as shown in Figure 5 , it is equivalent to determining the relationship between the flat film and the axial scan data at each position of the flat film through the training of the flat film and the axial scan image. Correspondingly, the embodiment steps can specifically include: obtaining sample flat film data and sample scan data of a preset object part, determining the attenuation curve of each sampling position and the preset attenuation value of M exposure positions corresponding to the attenuation curve according to the sample flat film data and the sample scan data; taking the attenuation curve as the model input feature and the preset attenuation value of the M exposure positions as the training label, training the attenuation value estimation model; obtaining the loss function of the attenuation value estimation model, and if it is judged that the loss function is less than a preset threshold, it is determined that the attenuation value estimation model is trained.

[0086] It should be noted that when training the attenuation value estimation model, in addition to training based on the maximum attenuation value MaxU, other information of the M exposure positions can also be trained, for example, MeanU, or directly training the U distribution, etc., which is not limited specifically. Correspondingly, when calculating the exposure index corresponding to the M exposure positions, it is not limited to the calculation method of the exposure index in the above three dimensions, and other calculable dimensions can also be included, which should all fall within the protection scope of the present application, and will not be enumerated here. In addition, the method for estimating the shape of the scanned part of the to-be-scanned object in the present application is not limited to flat film + AI, but can also be obtained by other methods, such as using statistical methods or modeling by camera image data, which is not limited here. How to determine the number and sampling position of the sampling view, the above gives one implementation method, there can be more factors to consider, the control method can be different, which is not limited here.

[0087] 203、According to the exposure index, N target exposure positions are selected from the M exposure positions.

[0088] For this embodiment, taking the exposure index as an example, if the exposure index is an exposure weight array determined based on only one data feature dimension, then N target exposure positions can be further selected from the M exposure positions according to the exposure weight values of the M exposure positions in descending order; if the exposure index is a plurality of exposure weight arrays determined based on a plurality of data feature dimensions, then the exposure weight values of the M exposure positions can be further calculated according to the plurality of exposure weight arrays, such as adding the exposure weight values corresponding to the same exposure position in the plurality of exposure weight arrays, and selecting N target exposure positions from the M exposure positions according to the order of the added exposure weight values in descending order; for example, the SUMPRODUCT function values of any N exposure positions in the M exposure positions can also be calculated according to the plurality of exposure weight arrays, and the N exposure positions corresponding to the maximum SUMPRODUCT function values are determined as the N target exposure positions. For example, if the exposure index is a plurality of exposure weight arrays determined based on a plurality of data feature dimensions, including: exposure weight array 1: Wi (i = 1 / 2 / 3 / 4 / 5…M), exposure weight array 2: Li (i = 1 / 2 / 3 / 4 / 5…M), exposure weight array 3: Gi (i = 1 / 2 / 3 / 4 / 5…M), when selecting N target exposure positions, N positions can be selected from the M exposure positions to maximize sum(G*W*L), and finally N target exposure positions after sparsification are obtained.

[0089] Correspondingly, when selecting N target exposure positions from the M exposure positions, the embodiment steps can specifically include: determining the exposure weight values of the M exposure positions according to at least one of the first exposure index, the second exposure index, and the third exposure index, and selecting N target exposure positions from the M exposure positions based on the exposure weight values. When selecting N target exposure positions from the M exposure positions based on the exposure weight values, the embodiment steps can specifically include: calculating the weight product sum of any N exposure positions in the M exposure positions based on the exposure weight values, and determining the N target exposure positions corresponding to the maximum weight product sum.

[0090] 204, respectively according to the target attenuation values of the N target exposure positions and the tube exposure capacity, determine the initial scan dose of the corresponding target exposure position.

[0091] The initial scanning dose of the corresponding target exposure position is modulated according to the exposure decay value and the ball tube exposure capacity of the N target exposure positions, and the exposure decay value includes at least one of the maximum decay value and the average decay value. The ball tube exposure capacity can be the mA capacity of the ball tube. In a specific application scenario, when the target scanning dose is modulated, the actual exposure capacity of the ball tube is also considered on the basis of the decay value, so that the determined initial scanning dose not only meets the theoretical scanning dose, but also ensures the normal operation of the CT machine ball tube and the image quality. The determination method of the initial scanning dose can use the existing related technology (such as CN104398266A), which is not limited here

[0092] 205. Determine the target exposure position to be corrected from the N target exposure positions, modulate the initial scanning dose of the target exposure position to be corrected based on the number of target exposure positions within a specific window length around the target exposure position to be corrected and the image quality requirement, and obtain the target scanning dose of the target exposure position to be corrected.

[0093] The above steps obtain the initial exposure dose. In order to make the modulated image quality meet the requirements more, a sparse degree function can be designed to correct the initial exposure dose for a specific image index, which includes but is not limited to at least one of noise, resolution and contrast. The following takes noise as an example:

[0094] Each target exposure position needs to be corrected for the initial scanning dose. Take one of the target exposure positions as an example. The initial exposure dose of the target exposure position C to be corrected is Dose_c. Take the target exposure position C to be corrected as the center, add a specific window length L on both sides (two sides), calculate the number of exposure positions of 2L+1 positions in the window, and mark it as N 2L . Calculate k=N 2L / 2L+1 to quantify the exposure position sparse degree around the target exposure position C to be corrected. For example, assuming that the target exposure position to be corrected is 9, the window L=4, then 5-6-7-8-9-10-11-12-13 (a total of 2*4+1=9 points) are looked at, and it is found that 8 and 10 are selected as target exposure positions. If it is found that 6, 7, 8, 9, 10, 11, 12 are selected as exposure positions, then the exposure sparse degree around position 9 is 7 / 9, k=2 / 9. Quantization is that 7 / 9 is more intensive than 2 / 9. Based on the image quality requirement of consistent noise, a correction function Y=F(k) is constructed, and the correction factor Y c of the target exposure position C to be corrected is obtained through the correction function Y=F(k), and the target exposure amount of the target exposure position to be corrected is Dose_c_R=Dose_c*Yc， i.e. initial scan dose of target exposure position to be corrected and correction factor Y c multiplication. Wherein, F(k) can be a piecewise function or other type of function, F(k) is a decreasing function, i.e. the larger the value of k, the smaller the value of Y. And the quantization method is not limited to k=N 2L / 2L+1, as long as the number of target exposure positions around the target exposure position to be corrected can be measured.

[0095] 206, based on N target exposure positions and target scan doses corresponding to the target exposure positions, scanning is performed.

[0096] The scanning method, device and equipment provided by the embodiment can first determine the sparse angle sampling number N of each sampling position of the object to be scanned in the scanning direction, wherein the scanning direction is from the head to the foot of the object to be scanned or from the foot to the head of the object to be scanned; at the same time, the exposure index of the optional M exposure positions of the XY scanning section corresponding to each sampling position can be determined, wherein the XY scanning section is the cross section of the object to be scanned at each sampling position perpendicular to the scanning direction, and M is the conventional number of CT tube exposure, M>N; further, N target exposure positions can be selected from the M exposure positions according to the exposure index, and the object to be scanned is scanned based on the N target exposure positions. Compared with the conventional sampling density, the technical solution in the application can determine the sparse angle sampling number and the target exposure position under the sparse angle sampling number, ensure accurate control, reduce the sampling frequency of the exposure line, reduce the scanning dose under the premise that the image quality meets the requirements, and efficiently reduce the radiation dose of the object to be scanned.

[0097] Further, as a specific implementation of the method shown in Figure 1 The embodiment provides a scanning device based on sparse angle sampling, as shown in Figure 6 The device comprises a first determination module 31, a second determination module 32 and a scanning module 33.

[0098] The first determination module 31 can be used to determine the sparse angle sampling number N of each sampling position of the object to be scanned in the scanning direction.

[0099] The second determination module 32 can be used to determine the exposure index of the optional M exposure positions of the XY scanning section corresponding to each sampling position, wherein the XY scanning section is the cross section of the object to be scanned at each sampling position perpendicular to the scanning direction, and M is the conventional number of CT tube exposure, M>N.

[0100] The scanning module 33 is used to select N target exposure positions from the M exposure positions according to the exposure index, and scan the object to be scanned.

[0101] In a specific application scenario, as shown in Figure 7 The first determining module 31 can specifically include an obtaining unit 311 and a calculating unit 312.

[0102] The obtaining unit 311 can be configured to obtain scanning information based on the pre-scanning data of the to-be-scanned object, the scanning information including at least one of a scanning part size, a scanning part name, and a scanning eccentric distance.

[0103] The calculating unit 312 can be configured to create a sparse function based on a function limit condition corresponding to the scanning information, and calculate the sparse angle sampling number N of each sampling position of the to-be-scanned object in the scanning direction by using the sparse function.

[0104] In a specific application scenario, when determining the exposure indexes corresponding to the M exposure positions, as shown in Figure 7 The second determining module 32 can specifically include a first determining unit 321 and a second determining unit 322.

[0105] The first determining unit 321 can be configured to determine an influencing factor of the exposure indexes corresponding to the M exposure positions.

[0106] The second determining unit 322 can be configured to determine the exposure indexes corresponding to the M exposure positions based on the influencing factor of the exposure indexes corresponding to the M exposure positions, wherein the influencing factor of the exposure indexes includes at least one of exposure attenuation, exposure distance, and exposure angle, and correspondingly, the exposure indexes include at least one of a first exposure index, a second exposure index, and a third exposure index.

[0107] The second determining unit 322 can specifically be configured to input the to-be-scanned object data into a trained attenuation value estimation model, obtain the exposure attenuation values respectively corresponding to the M exposure positions at each sampling position, determine the first exposure indexes corresponding to the M exposure positions based on the exposure attenuation values, respectively determine the exposure distance values of the M exposure positions and the to-be-scanned object, determine the second exposure indexes corresponding to the M exposure positions according to the exposure distance values, determine a first exposure position in the M exposure positions, and determine the third exposure indexes corresponding to the M exposure positions according to the exposure angles of the remaining M-1 exposure positions relative to the first exposure position.

[0108] In a specific application scenario, as shown in Figure 7 The device further includes a training module 34.

[0109] Training module 34 can be used to acquire sample flat film data and sample scan data of preset object parts, determine the attenuation curve of each sampling position based on the sample flat film data and sample scan data, and the preset attenuation value of M exposure positions corresponding to the attenuation curve; use the attenuation curve as the model input feature and the preset attenuation value of M exposure positions as the training label to train the attenuation value prediction model; obtain the loss function of the attenuation value prediction model, and if the loss function is less than the preset threshold, it is determined that the attenuation value prediction model training is complete.

[0110] In specific application scenarios, when selecting N target exposure locations from M exposure locations based on the exposure index, such as... Figure 7 As shown, the scanning module 33 may specifically include: a filtering unit 331;

[0111] The filtering unit 331 can be used to determine the exposure weight value of each of the M exposure positions based on at least one of the first exposure index, the second exposure index, and the third exposure index, and to filter out N target exposure positions from the M exposure positions based on the exposure weight value.

[0112] In specific application scenarios, the filtering unit 331 can be used to calculate the sum of the weighted products of any N exposure positions out of M exposure positions based on the exposure weight value, and determine the N uniform exposure positions with the largest sum of the corresponding weighted products as N target exposure positions.

[0113] In specific application scenarios, such as Figure 7 As shown, the scanning module 33 may further include: a modulation unit 332;

[0114] The modulation unit 332 can be used to modulate the initial scan dose of the corresponding target exposure position according to the exposure attenuation value of N target exposure positions and the X-ray tube exposure capability, and determine the target exposure position to be corrected from the N target exposure positions. Based on the number of target exposure positions within a specific window length around the target exposure position to be corrected and the image quality requirements, the initial scan dose of the target exposure position to be corrected is modulated to obtain the target scan dose of the target exposure position to be corrected.

[0115] Accordingly, the exposure attenuation value includes at least one of the maximum attenuation value and the average attenuation value. The modulation unit 332 can be specifically used to determine the exposure attenuation value of the N target exposure positions and the target X-ray tube mA that matches the X-ray tube exposure capability.

[0116] The modulation unit 332 can further be used to determine the number N of target exposure positions within a specific window length L on both sides, centered on the target exposure position to be corrected. 2L, a specific window length L is arranged according to a spacing sampling of M exposure positions per sampling position; based on the image quality requirement of consistent noise at each sampling position, a correction function Y=F(k) is constructed, k=N 2L / 2L+1, F(k) is a decreasing function; based on the correction function Y=F(k), a correction factor Y of a target exposure position to be corrected is obtained c ; the correction factor Y c is multiplied by the target exposure position to be corrected, and a target scanning dose of the target exposure position to be corrected is obtained.

[0117] It should be noted that other corresponding descriptions of the functions of the scanning device based on sparse angle sampling provided by the embodiments of the present application can refer to the corresponding descriptions of the method shown in Figure 1 , and will not be described here.

[0118] Based on the method shown in Figure 1 , accordingly, the embodiments of the present application also provide a computer readable storage medium having a computer program stored thereon, which is executed by a processor to implement the following steps: determining the sparse angle sampling number N of each sampling position of the object to be scanned in the scanning direction, wherein the scanning direction is from the head to the foot of the object to be scanned or from the foot to the head of the object to be scanned; determining the exposure index of the M exposure positions corresponding to each sampling position of the XY scanning section, wherein the XY scanning section is the cross section of the object to be scanned perpendicular to the scanning direction at each sampling position, M is the conventional number of CT tube exposure, and M>N; according to the exposure index, selecting N target exposure positions from the M exposure positions to scan the object to be scanned.

[0119] Based on the method shown in Figure 1 , Figure 2 and the device shown in Figure 6 , Figure 7 , the embodiments of the present application also provide a physical structure diagram of a computer device, as shown in Figure 8As shown, the computer device comprises a processor 41, a memory 42, and a computer program stored on the memory 42 and executable on the processor, wherein the memory 42 and the processor 41 are arranged on a bus 43, and the processor 41 implements the following steps when executing the program: determining a sparse angle sampling number N of each sampling position of a to-be-scanned object in a scanning direction, wherein the scanning direction is a direction from a head of the to-be-scanned object to a foot or a direction from the foot of the to-be-scanned object to the head; determining an exposure index of each sampling position corresponding to M exposure positions of an XY scanning section, wherein the XY scanning section is a section of the to-be-scanned object at each sampling position perpendicular to the scanning direction, M is a conventional number of CT tube exposure, and M>N; and selecting N target exposure positions from the M exposure positions according to the exposure index, and scanning the to-be-scanned object.

[0120] Through the technical scheme of the present application, the sparse angle sampling number N of each sampling position of the to-be-scanned object in the scanning direction can be determined first, wherein the scanning direction is a direction from a head of the to-be-scanned object to a foot or a direction from the foot of the to-be-scanned object to the head; at the same time, the exposure index of each sampling position corresponding to M exposure positions of an XY scanning section can be determined, wherein the XY scanning section is a section of the to-be-scanned object at each sampling position perpendicular to the scanning direction, M is a conventional number of CT tube exposure, and M>N; further, N target exposure positions can be selected from the M exposure positions according to the exposure index, and the to-be-scanned object is scanned based on the N target exposure positions. Compared with the conventional sampling density, the technical scheme in the present application can determine the sparse angle sampling number and the target exposure position under the sparse angle sampling number, ensure accurate control, reduce the sampling frequency of the exposure line, reduce the scanning dose under the premise that the image quality meets the requirements, and can efficiently reduce the radiation dose of the to-be-scanned object.

[0121] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application can be realized by a general computing device, which can be concentrated on a single computing device or distributed on a network composed of multiple computing devices, and alternatively, they can be realized by program codes executable by the computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in different orders, or they can be manufactured into individual integrated circuit modules, or multiple modules or steps among them can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any particular combination of hardware and software.

[0122] The above merely describes the preferred embodiments of the present application, and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A scanning method, characterized in that, include: Determine the number N of sparse angle samples at each sampling position of the object to be scanned along the scanning direction; Determine the exposure index of M exposure positions corresponding to each sampling position on the XY scanning section, wherein the XY scanning section is the cross section of the object to be scanned perpendicular to the scanning direction at each sampling position, M is the conventional number of CT tube exposures, and M>N; Based on the exposure index, N target exposure positions are selected from the M exposure positions, and the object to be scanned is scanned. The scanning of the object to be scanned includes: The initial scan dose for each of the N target exposure positions is determined based on the exposure attenuation value and the X-ray tube exposure capability, wherein the exposure attenuation value includes at least one of the maximum attenuation value and the average attenuation value. The target exposure position to be corrected is determined from N target exposure positions. Based on the number of target exposure positions within a specific window length around the target exposure position to be corrected and the image quality requirements, the initial scan dose of the target exposure position to be corrected is modulated to obtain the target scan dose of the target exposure position to be corrected.

2. The method according to claim 1, characterized in that, The determination of the sparse angle sampling number N of the object to be scanned at each sampling position in the scanning direction includes: Based on the pre-scan data of the object to be scanned, scanning information is obtained, which includes at least one of the following: scanned area size, scanned area name, and scan offset distance; A sparse function is created based on the function constraints corresponding to the scanning information, and the sparse function is used to calculate the number of sparse angle samples N at each sampling position of the object to be scanned in the scanning direction.

3. The method according to claim 2, characterized in that, Determining the exposure index of the M exposure positions corresponding to each sampling position on the XY scan section includes: Determine the factors influencing the exposure index corresponding to the M exposure positions; Based on the influencing factors of the exposure index corresponding to the M exposure positions, the exposure index corresponding to the M exposure positions is determined; The factors influencing the exposure index include at least one of the following: exposure attenuation value, exposure distance, and exposure angle.

4. The method according to claim 3, characterized in that, The factors influencing the determination of the exposure index corresponding to the M exposure positions include: The pre-scanning data of the object to be scanned is input into the trained attenuation value prediction model to obtain the exposure attenuation values ​​corresponding to the M exposure positions at each sampling position.

5. The method according to claim 4, characterized in that, Before inputting the pre-scanned data of the object to be scanned into the trained attenuation prediction model, the process also includes: Acquire sample flat film data and sample scan data of a preset object area, and determine the attenuation curve of each sampling position and the preset attenuation value of M exposure positions corresponding to the attenuation curve based on the sample flat film data and the sample scan data. The attenuation curve is used as the model input feature, and the preset attenuation values ​​of the M exposure positions are used as training labels to train the attenuation value prediction model. Obtain the loss function of the decay value prediction model. If the loss function is less than a preset threshold, then the decay value prediction model is determined to have completed training.

6. The method according to claim 3, characterized in that, The step of selecting N target exposure locations from the M exposure locations based on the exposure index includes: Based on at least one of the exposure indices, determine the exposure weight value of each of the M exposure locations, and select N target exposure locations from the M exposure locations based on the exposure weight values.

7. The method according to claim 6, characterized in that, The step of selecting N target exposure positions from the M exposure positions based on the exposure weight value includes: Based on the exposure weight values, the weights of any N exposure positions among the M exposure positions are summed, and the N exposure positions with the largest corresponding weight sums are determined as the N target exposure positions.

8. The method according to claim 1, characterized in that, The process of modulating the initial scan dose of the target exposure position to be corrected based on the number of target exposure positions within a specific window length surrounding the target exposure position to be corrected and the image quality requirements, to obtain the target scan dose, includes: Centered on the target exposure position to be corrected, determine the number N of target exposure positions within a specific window length L on both sides. 2L The specific window length L is sampled at intervals of M exposure positions arranged at each sampling position; Based on the requirement of consistent image quality despite noise, a correction function Y=F(k), k=N is constructed. 2L / 2L+1, F(k) is a decreasing function; Based on the correction function Y=F(k), the correction factor Y of the target exposure position to be corrected is obtained. c ; The correction factor Y c The target scan dose at the target exposure location to be corrected is obtained by multiplying it by the initial scan dose.

9. A scanning device based on sparse angular sampling, characterized in that, include: The first determining module is used to determine the number N of sparse angle samples at each sampling position of the object to be scanned in the scanning direction; The second determining module is used to determine the exposure index of M exposure positions of the XY scanning section corresponding to each sampling position, wherein the XY scanning section is the cross section of the object to be scanned perpendicular to the scanning direction at each sampling position, M is the conventional number of CT tube exposures, and M>N. The scanning module is used to select N target exposure positions from the M exposure positions based on the exposure index, and scan the object to be scanned. The scanning module further includes: a modulation unit; The modulation unit is configured to determine the initial scan dose of the corresponding target exposure position based on the exposure attenuation value of the N target exposure positions and the X-ray tube exposure capability, wherein the exposure attenuation value includes at least one of the maximum attenuation value and the average attenuation value; determine the target exposure position to be corrected from the N target exposure positions; and modulate the initial scan dose of the target exposure position to be corrected based on the number of target exposure positions within a specific window length around the target exposure position to be corrected and the image quality requirements, thereby obtaining the target scan dose of the target exposure position to be corrected.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

11. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

Images