Panoramic imaging method and device based on dual energy

Abstract

The invention relates to the technical field of oral cavity imaging, and discloses a method for applying a dual-energy technology to oral cavity panoramic imaging, which comprises the following steps of: dual-energy panoramic projection data acquisition: acquiring a high-low-energy panoramic narrow-edge projection drawing by utilizing technologies such as area array filtering, high-pressure fast switching or multi-layer flat plate; performing panoramic projection domain substance decomposition: decomposing the high and low energy projection to obtain a target base substance density integrogram; variable center finite angle reconstruction: reconstructing a base material density integral image to generate a corresponding panoramic base image; and application extension: generating various clinical related images according to the panoramic base image. The invention further provides a panoramic imaging device based on dual energy. According to the invention, the expressive force of the panoramic image can be improved, and more application scenes can be expanded.

Description

Technical Field

[0001] The present invention relates to the technical field of oral imaging, and in particular to a panoramic imaging method and device based on dual energy. Background Art

[0002] In the field of oral diagnosis, panoramic X-ray imaging technology (Panoramic X-ray) is a key tool for obtaining a comprehensive view of teeth and jaws. This technology uses single-energy (kVp) X-rays to perform variable-center limited-angle exposure scanning along a preset scanning trajectory to generate a two-dimensional image containing the upper and lower jaws, all teeth and other oral structures. Although it is easy to operate, cost-effective and widely used.

[0003] Existing oral panoramic imaging methods perform panoramic reconstruction by scanning at a variable center limited non-uniform angle along a preset scanning trajectory according to a preset dental arch curve. In actual applications, different manufacturers use different scanning trajectories and reconstruction algorithms. These result in the tissue grayscale values ​​generated by panoramic reconstruction having no unified or clear physical meaning, and cannot represent or accurately measure the properties of the material at that location. Clinical disease diagnosis can only be made based on the relative relationship between pixels. Dual-energy imaging technology collects X-ray projection data at different energy levels and utilizes the differences in attenuation characteristics of different materials at different energies to significantly improve tissue differentiation and the accuracy of material classification and identification. This technology shows great potential in improving the accuracy of CT image diagnosis and expanding the scope of clinical application. However, existing panoramic imaging methods are difficult to perform accurate material decomposition in conventional ways, and have certain limitations in material resolution, image detail expression and clinical application scope, which may lead to poor imaging results during panoramic imaging. Summary of the invention

[0004] The present invention provides a dual-energy based panoramic imaging method and device, the main purpose of which is to solve the problem of poor expressiveness of imaging results during panoramic imaging.

[0005] To achieve the above object, the present invention provides a dual-energy based panoramic imaging method, comprising:

[0006] Generate dual-energy rays by using a high-voltage fast-cut method, and use the dual-energy rays for projection imaging to obtain a high- and low-energy sparse projection image group;

[0007] Performing sparse compensation on the high and low energy sparse projection image group to obtain a high and low energy projection image group;

[0008] Performing energy spectrum measurement on the dual-energy rays to obtain high-energy and low-energy spectrum curves;

[0009] Generate a dual-energy decomposition mapping table according to a preset virtual base material model and the high- and low-energy energy spectrum curves, and perform projection domain decomposition on the high- and low-energy projection image group according to the dual-energy decomposition mapping table to obtain a base material density integral decomposition result;

[0010] The base material density integral decomposition result is reconstructed with a variable center finite angle to obtain a panoramic base map group.

[0011] Optionally, the performing energy spectrum measurement on the dual-energy rays to obtain high-energy and low-energy spectrum curves includes:

[0012] Using the dual-energy rays to perform exposure scanning on a preset pure measurement phantom to obtain a projection data set;

[0013] Selecting projection data in the projection data group one by one as target projection data, performing back-projection reconstruction, three-dimensional segmentation, re-assignment and forward projection on the target projection data to obtain an integral path length;

[0014] Initialize the initial energy spectrum curve, perform energy spectrum attenuation calculation according to the initial energy spectrum curve and the integral path length, and obtain the predicted ray attenuation value:

[0015]

[0016] Wherein, P refers to the predicted ray attenuation value, refers to the initial energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the pure measurement phantom for energy E, μ is the linear attenuation coefficient, ρ is the density symbol, l is the integral path length, d is the differential symbol, and E is the energy value;

[0017] calculating a projection attenuation difference between the predicted ray attenuation value and the target projection data;

[0018] The initial energy spectrum curve is iteratively updated according to the projection attenuation difference to obtain an energy spectrum curve, and high and low energy spectrum curves are generated according to the energy spectrum curves of all target projection data in the projection data group.

[0019] Optionally, performing back-projection reconstruction, three-dimensional segmentation, re-assignment and forward projection on the target projection data to obtain the integral path length includes:

[0020] Performing back-projection reconstruction on the target projection data to obtain a reconstruction result;

[0021] Performing threshold segmentation on the reconstruction result according to a preset attenuation coefficient interval to obtain a segmented region phantom;

[0022] Reassigning voxels of the segmented region phantom to obtain a voxel-reassigned phantom;

[0023] The voxel re-values ​​are forward projected to obtain the integrated path length.

[0024] Optionally, generating a dual-energy decomposition mapping table according to a preset virtual base material model and the high- and low-energy spectrum curves includes:

[0025] Performing forward projection on a preset virtual base material model to obtain a first base material density integral and a second base material density integral;

[0026] Extracting a low energy spectrum curve and a high energy spectrum curve from the high and low energy spectrum curves;

[0027] The low-energy ray attenuation value is calculated according to the first base material density integral, the second base material density integral and the low-energy spectrum curve:

[0028]

[0029] m A =∫ρ A (l)dl

[0030] m B =∫ρ B (l)dl

[0031] Among them, P L It refers to the low energy ray attenuation value, L refers to low energy, S L (E) refers to the low energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the first base material A preset in the virtual base material model, μ A is the linear attenuation coefficient of the first base material A preset in the virtual base material phantom, ρ A is the density of the first base material A preset in the virtual base material model, m A refers to the density integral of the first basis substance, is the mass attenuation coefficient of the second base material B preset in the virtual base material model, μ B is the linear attenuation coefficient of the second base material B preset in the virtual base material phantom, ρ B is the density of the first base material B preset in the virtual base material model, m B refers to the density integral of the second basis material, d is the differential sign, E is the energy value, and l is the preset integral path length;

[0032] The high-energy ray attenuation value is calculated according to the first base material density integral, the second base material density integral and the high-energy spectrum curve:

[0033]

[0034] Among them, P H It refers to the attenuation value of the high-energy ray, H refers to high energy, S H (E) refers to the high energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the first base material A preset in the virtual base material model, μ A is the linear attenuation coefficient of the first base material A preset in the virtual base material phantom, ρ A is the density of the first base material A preset in the virtual base material model, m A refers to the density integral of the first basis substance, is the mass attenuation coefficient of the second base material B preset in the virtual base material model, μ B is the linear attenuation coefficient of the second base material B preset in the virtual base material phantom, ρ B is the density of the first base material B preset in the virtual base material model, m B refers to the density integral of the second basis material, d is the differential sign, E is the energy value, and l is the preset integral path length;

[0035] Generate a primary mapping table according to the first base material density integral, the second base material density integral, the low-energy ray attenuation value, and the high-energy ray attenuation value;

[0036] The primary mapping table is uniformly interpolated with equal spacing to obtain a dual-energy decomposition mapping table.

[0037] Optionally, performing projection domain decomposition on the high and low energy projection image group according to the dual energy decomposition mapping table to obtain a basis material density integral decomposition result includes:

[0038] Performing attenuation calculation on the high- and low-energy projection image group to obtain a high- and low-energy attenuation value pair;

[0039] According to the method of nearest search, a neighborhood value group closest to the high and low energy attenuation values ​​is screened out from the dual energy decomposition mapping table to obtain a nearest attenuation value group;

[0040] Performing bilinear interpolation on the group of most recent attenuation values ​​to obtain a density decomposition value;

[0041] Density reconstruction is performed according to the density decomposition value to obtain a base material density integral decomposition result.

[0042] Optionally, performing variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base image group includes:

[0043] Performing ramp filtering on the base material density integral decomposition result to obtain a filtered base material decomposition result;

[0044] Performing weighted calculation on the filtering basis material decomposition result to obtain a weighted basis material decomposition result;

[0045] Perform multi-layer dental arch curve back-projection on the weighted base material decomposition result to obtain a hierarchical panoramic base image group;

[0046] Automatically focus and fuse the hierarchical panoramic base image group to obtain a panoramic base image group.

[0047] In order to solve the above problems, the present invention further provides a dual-energy based panoramic imaging device, the device comprising:

[0048] A projection acquisition module, used to generate dual-energy rays by using a high-voltage fast-cut method, and use the dual-energy rays for projection imaging to obtain a high- and low-energy sparse projection image group;

[0049] A sparse compensation module, used for performing sparse compensation on the high and low energy sparse projection image group to obtain a high and low energy projection image group;

[0050] An energy spectrum measurement module, used for measuring the energy spectrum of the dual-energy rays to obtain high-energy and low-energy spectrum curves;

[0051] A mapping decomposition module, used to generate a dual-energy decomposition mapping table according to a preset virtual base material model and the high- and low-energy energy spectrum curves, and perform projection domain decomposition on the high- and low-energy projection image group according to the dual-energy decomposition mapping table to obtain a base material density integral decomposition result;

[0052] The base image reconstruction module is used to perform variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base image group.

[0053] The present invention utilizes the dual-energy rays for projection imaging and can utilize the specific attenuation characteristics of different substances under X-rays at different energy levels to decompose materials, which can significantly improve the recognition accuracy of specific oral tissues. By performing sparse compensation, deep learning or linear interpolation methods can be used to interpolate projection images at intermediate angles, thereby avoiding quality degradation during subsequent image reconstruction caused by angular sparseness of the projection image.

[0054] By performing energy spectrum measurement on the dual-energy rays, the accuracy of energy spectrum measurement can be improved, the calibration process can be simplified, and reliable reference data can be provided. By performing relationship mapping and projection domain decomposition, noise and false sound in different energy images can be eliminated, image contrast can be enhanced, and it is convenient to distinguish different substances, realize material identification and accurate quantification, thereby improving the accuracy of subsequent panoramic imaging. By performing variable center limited angle reconstruction and panoramic base map imaging, the reconstruction accuracy can be improved, local area optimization can be achieved, and high-contrast and high-detail oral structures can be fully displayed, thereby improving the accuracy of panoramic images. Therefore, the dual-energy-based panoramic imaging method and device proposed in the present invention can solve the problem of poor imaging results when performing oral panoramic imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic diagram of a flow chart of a dual-energy based panoramic imaging method provided by an embodiment of the present invention;

[0056] Figure 2 A schematic diagram of the unit structure of a filtering area array unit of an area array filtering scanner provided by an embodiment of the present invention;

[0057] Figure 3 A schematic diagram of a structure for high-pressure rapid cutting and collection provided by an embodiment of the present invention;

[0058] Figure 4 A schematic diagram of a scanning energy spectrum curve provided by an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of a dual energy decomposition mapping table provided by an embodiment of the present invention;

[0060] Figure 6 A functional module diagram of a dual-energy panoramic imaging device provided by an embodiment of the present invention;

[0061] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0062] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0063] The embodiment of the present application provides a dual-energy based panoramic imaging method. The execution subject of the dual-energy based panoramic imaging method includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided in the embodiment of the present application. In other words, the dual-energy based panoramic imaging method can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms.

[0064] Reference Figure 1 FIG. 1 is a flow chart of a dual-energy panoramic imaging method according to an embodiment of the present invention. In this embodiment, the dual-energy panoramic imaging method includes:

[0065] S1. Generate dual-energy rays using a high-voltage fast-cut method, and use the dual-energy rays to perform projection imaging to obtain a high- and low-energy sparse projection image group.

[0066] Specifically, the dual-energy ray refers to X-rays of both high energy and low energy, and the dual-energy ray can be generated by a filtered area array scanner, a double-layer flat-panel detector or a high-voltage fast cutting method, and the high- and low-energy sparse projection image group includes a high-energy sparse projection image and a low-energy sparse projection image, wherein the high-energy sparse projection image refers to a sparse projection image obtained by projecting high-energy X-rays, and the low-energy sparse projection image refers to a sparse projection image obtained by projecting low-energy X-rays.

[0067] In detail, the filtering area array scanner refers to an X-ray scanner equipped with a filtering area array unit, wherein the filtering area array unit evenly covers each detector channel of the X-ray scanner. The filtering area array unit can refer to Figure 2 As shown, black represents strong filtering materials, such as aluminum, copper, etc., and white represents weak filtering or even no filtering materials. When scanning, the exposure parameter is set to high kV, and the pixels corresponding to the black strong filtering area of ​​the filtering module in the collected projection image are low-energy sparse projection images, and the white area is a high-energy sparse projection image. At this time, each sparse projection image contains high and low energy projection values, but they are alternating and staggered distribution.

[0068] Specifically, the double-layer flat-panel detector is a method for simultaneously acquiring high and low energy spectrum data in a single exposure. Since the data is acquired in a single exposure, motion artifacts are avoided and the temporal resolution is improved. By using two closely arranged flat-panel detector layers, each flat-panel detector layer has different sensitivities to X-rays of different energies. The upper flat-panel detector is usually more sensitive to low-energy X-rays, and a thinner absorption layer or a material with a higher absorption efficiency for low-energy X-rays can be used. The lower flat-panel detector is more sensitive to high-energy X-rays, and a thicker absorption layer or a material with a higher absorption efficiency for high-energy X-rays can be used.

[0069] Specifically, refer to Figure 3 As shown, the method of generating dual-energy rays by high-voltage fast cutting refers to the use of a flat-panel detector to alternately collect high and low energy data by quickly switching the high-voltage tube voltage. The projection data of the high and low energy sparse projection image groups are arranged in the order of "low, high, low, high..." When collecting, the rotation axis is not fixed, but moves along a preset rotation center trajectory. Common designs include a fixed x-axis that moves back and forth on the y-axis, and the trajectory is in the shape of a "1"; or the x-axis also adds movement in a fixed direction, and the trajectory is in the shape of a "human", such as the panoramic preset dental arch trajectory in the figure.

[0070] In the embodiment of the present invention, by using the dual-energy rays for projection imaging, the specific attenuation characteristics of different substances under X-rays at different energy levels can be used to decompose the substances, which can significantly improve the recognition accuracy of specific oral tissues.

[0071] S2. Perform sparse compensation on the high and low energy sparse projection image group to obtain a high and low energy projection image group.

[0072] In detail, there is an acquisition time difference between adjacent high and low energy projections obtained by high-voltage fast cutting. Although it is very small, the ray path still changes. The projection data is not pixel-level corresponding and has a certain sparsity. Therefore, it is necessary to perform sparse compensation on the high and low energy sparse projection image group.

[0073] In the embodiment of the present invention, sparse compensation is performed on the high and low energy sparse projection image group to obtain a high and low energy projection image group:

[0074] Using a pre-trained deep interpolation model to perform feature convolution and feature pooling on the high and low energy sparse projection image group to obtain a sparse projection feature group;

[0075] Using the deep interpolation model to perform skip connections and full connections on the sparse projection feature group to obtain a standard projection feature group;

[0076] The deep interpolation model is used to perform feature activation and feature upsampling on the standard projection feature group to obtain a high- and low-energy projection image group.

[0077] In detail, in another embodiment of the present invention, sparse compensation can be performed by a continuous linear interpolation method to obtain a group of high and low energy projection images. The continuous linear interpolation, for example, collects projection images of n projection angles and positions, and the low energy sparse projection images are numbered 1, 3, 5, ..., n-1, and the high energy sparse projection images are numbered 2, 4, 6, ..., n. The adjacent odd-numbered data are used to interpolate the intermediate even-numbered data, that is, for low energy data, 1 and 3 are interpolated into 2, 3 and 5 are interpolated into 4, and both high and low energy can interpolate complete projection images with numbers 1, 2, 3, ..., n, a total of n angles.

[0078] Specifically, the deep interpolation model can be a convolutional neural network of U-Net, and the deep interpolation model includes a convolution layer, a pooling layer, a jump connection layer, a fully connected layer, an activation sampling layer and an upsampling layer, wherein the convolution layer is used to extract local features of the input image and provide the network with rich image information. The pooling layer is used to reduce the spatial dimension of the features while improving the network's abstraction ability to help the model capture higher-level visual information. The jump connection layer, in some architectures, such as U-Net, is used to use a jump connection to directly connect the feature map in the encoder to the corresponding layer of the decoder, which helps to retain more details and position information during image reconstruction. The fully connected layer is in the deep layer of the network and is used to map the learned features to the output space to generate a complete data representation. The activation sampling layer includes an activation function, such as ReLU, which enhances the expressive power of the model by introducing nonlinearity and allows the network to learn more complex data relationships. The upsampling layer is in the decoder part and is used to gradually restore the spatial resolution of the image and generate high-resolution output.

[0079] In detail, before performing feature convolution and feature pooling on the high and low energy sparse projection image group using a pre-trained deep interpolation model to obtain a sparse projection feature group, the method further includes:

[0080] The pre-collected standard sparse projection map is split into data sets to obtain training sparse projection maps and real sparse projection maps;

[0081] Using a convolutional neural network to perform feature convolution and feature pooling on the training sparse projection image to obtain training projection features;

[0082] Performing skip connection and full connection on the training projection features to obtain connection projection features;

[0083] Performing feature activation and feature upsampling on the connection projection features to obtain a predicted sparse projection graph;

[0084] The composite loss value between the predicted sparse projection graph and the true sparse projection graph is calculated using the following composite loss algorithm:

[0085]

[0086] Wherein, Loss refers to the composite loss value, α and β are preset balance weights, N refers to the total number of pixels of the predicted oral projection image, and the total number of pixels of the predicted sparse projection image is equal to the total number of pixels of the real sparse projection image, i is the pixel number, refers to the i-th pixel value in the predicted sparse projection image, refers to the i-th pixel value in the true sparse projection image, and SSIM() is the symbol of the structural similarity index function;

[0087] The convolutional neural network is iteratively parameter updated according to the composite loss value to obtain a deep interpolation model.

[0088] In detail, the standard sparse projection image is a high-energy projection image and a low-energy projection image obtained in advance by projection imaging using dual-energy rays. The data set splitting refers to selecting a projection image for training and a corresponding projection image used as a real data set. The composite loss dispersion can reduce pixel-level errors and generate prediction results that are visually closer to the projection image while maintaining image structural features.

[0089] In an embodiment of the present invention, by performing sparse compensation, a projection image at an intermediate angle can be interpolated using a deep learning or linear interpolation method, thereby avoiding quality degradation during subsequent image reconstruction due to angular sparseness of the projection image.

[0090] S3. Measure the energy spectrum of the dual-energy rays to obtain high-energy and low-energy spectrum curves.

[0091] For details, refer to Figure 4 As shown, the high- and low-energy spectrum curve refers to the number distribution curve of photons of different energies (keV) in the X-ray of the tube at a certain voltage. It is usually necessary to normalize it into a weight curve with an integral of 1. When designing high voltage, it is necessary to consider making the overlap of the high- and low-energy spectrum curves as low as possible. The spectrum of the high voltage at different voltages can be directly measured by a spectrum measuring instrument, or an actual pure model can be designed for scanning and calculated through algorithm iteration. The high- and low-energy spectrum curves include a high-energy spectrum curve and a low-energy spectrum curve, wherein the dotted line portion is the low-energy spectrum curve, and the solid line portion is the high-energy spectrum curve.

[0092] In the embodiment of the present invention, the energy spectrum measurement of the dual-energy rays to obtain high-energy and low-energy spectrum curves includes:

[0093] Using the dual-energy rays to perform exposure scanning on a preset pure measurement phantom to obtain a projection data set;

[0094] Selecting projection data in the projection data group one by one as target projection data, performing back-projection reconstruction, three-dimensional segmentation, re-assignment and forward projection on the target projection data to obtain an integral path length;

[0095] Initialize the initial energy spectrum curve, perform energy spectrum attenuation calculation according to the initial energy spectrum curve and the integral path length, and obtain the predicted ray attenuation value:

[0096]

[0097] Wherein, P refers to the predicted ray attenuation value, refers to the initial energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the pure measurement phantom for energy E, μ is the linear attenuation coefficient, ρ is the density symbol, l is the integral path length, d is the differential symbol, and E is the energy value;

[0098] calculating a projection attenuation difference between the predicted ray attenuation value and the target projection data;

[0099] The initial energy spectrum curve is iteratively updated according to the projection attenuation difference to obtain an energy spectrum curve, and high and low energy spectrum curves are generated according to the energy spectrum curves of all target projection data in the projection data group.

[0100] In detail, the pure measurement phantom refers to a measurement phantom composed of a known pure base material with known parameters, and the known pure base material may be water or bone, and the projection data set includes projection data of high-energy rays and projection data of low-energy rays.

[0101] Specifically, performing back-projection reconstruction, three-dimensional segmentation, re-assignment and forward projection on the target projection data to obtain the integral path length includes:

[0102] Performing back-projection reconstruction on the target projection data to obtain a reconstruction result;

[0103] Performing threshold segmentation on the reconstruction result according to a preset attenuation coefficient interval to obtain a segmented region phantom;

[0104] Reassigning voxels of the segmented region phantom to obtain a voxel-reassigned phantom;

[0105] The voxel re-values ​​are forward projected to obtain the integrated path length.

[0106] Specifically, the back-projection reconstruction method may be a Feldkamp-Davis-Kress projection reconstruction method, the segmented region phantom may be a polymethyl methacrylate (PMMA) region phantom, the voxel reassignment refers to setting the voxel value of the projection interest region to 1 and the voxel value of the remaining region to 0, and the Radon algorithm may be used for forward projection.

[0107] Specifically, the primary energy spectrum curve is a discrete and normalized initial energy spectrum weight curve simulated by software such as Spectrum. The projected attenuation density algorithm can more accurately calculate the attenuation of X-rays when they pass through an object by integrating X-rays of different energies.

[0108] In the embodiment of the present invention, by performing energy spectrum measurement on the dual-energy rays, the accuracy of energy spectrum measurement can be improved, the calibration process can be simplified, and reliable benchmark data can be provided.

[0109] S4. Generate a dual-energy decomposition mapping table according to the preset virtual base material model and the high- and low-energy spectrum curves, and perform projection domain decomposition on the high- and low-energy projection image group according to the dual-energy decomposition mapping table to obtain a base material density integral decomposition result.

[0110] In detail, the base material density integral decomposition result includes the density of each type of material in the oral projection and the corresponding image area, and the base material density integral decomposition result can improve the accuracy of subsequent panoramic imaging.

[0111] In the embodiment of the present invention, the projection domain decomposition of the high and low energy projection image group is performed according to the dual energy decomposition mapping table to obtain the basis material density integral decomposition result, including:

[0112] Performing attenuation calculation on the high- and low-energy projection image group to obtain a high- and low-energy attenuation value pair;

[0113] According to the method of nearest search, a neighborhood value group closest to the high and low energy attenuation values ​​is screened out from the dual energy decomposition mapping table to obtain a nearest attenuation value group;

[0114] Performing bilinear interpolation on the group of most recent attenuation values ​​to obtain a density decomposition value;

[0115] Density reconstruction is performed according to the density decomposition value to obtain a base material density integral decomposition result.

[0116] Specifically, the attenuation calculation refers to respectively splitting the high and low energy spectra of the enhanced oral projection image and extracting the energy spectrum attenuation to obtain high energy attenuation values ​​and low energy attenuation values, and aggregating the high energy attenuation values ​​and the low energy attenuation values ​​into a high and low energy attenuation value pair.

[0117] Specifically, the method of nearest search is used to filter out the neighborhood value group closest to the high and low energy attenuation values ​​from the dual-energy decomposition mapping table. Obtaining the nearest attenuation value group means finding the four nearest attenuation values ​​from the dual-energy decomposition mapping table and gathering them into the nearest attenuation value group. The density decomposition substance refers to the density integral decomposition result of the two substances corresponding to the dual-energy decomposition mapping table.

[0118] In detail, the generating of the dual energy decomposition mapping table according to the preset virtual base material model and the high and low energy spectrum curves includes:

[0119] Performing forward projection on a preset virtual base material model to obtain a first base material density integral and a second base material density integral;

[0120] Extracting a low energy spectrum curve and a high energy spectrum curve from the high and low energy spectrum curves;

[0121] The low-energy ray attenuation value is calculated according to the first base material density integral, the second base material density integral and the low-energy spectrum curve:

[0122]

[0123] m A =∫ρ A (l)dl

[0124] m B =∫ρ B (l)dl

[0125] Among them, P L It refers to the low energy ray attenuation value, L refers to low energy, S L (E) refers to the low energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the first base material A preset in the virtual base material model, μ A is the linear attenuation coefficient of the first base material A preset in the virtual base material phantom, ρ A is the density of the first base material A preset in the virtual base material model, m A refers to the density integral of the first basis substance, is the mass attenuation coefficient of the second base material B preset in the virtual base material model, μ B is the linear attenuation coefficient of the second base material B preset in the virtual base material phantom, ρ B is the density of the first base material B preset in the virtual base material model, m B refers to the density integral of the second basis material, d is the differential sign, E is the energy value, and l is the preset integral path length;

[0126] The high-energy ray attenuation value is calculated according to the first base material density integral, the second base material density integral and the high-energy spectrum curve:

[0127]

[0128] Among them, P H It refers to the attenuation value of the high-energy ray, H refers to high energy, S H (E) refers to the high energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the first base material A preset in the virtual base material model, μ A is the linear attenuation coefficient of the first base material A preset in the virtual base material phantom, ρ A is the density of the first base material A preset in the virtual base material model, m A refers to the density integral of the first basis substance, is the mass attenuation coefficient of the second base material B preset in the virtual base material model, μ B is the linear attenuation coefficient of the second base material B preset in the virtual base material phantom, ρ B is the density of the first base material B preset in the virtual base material model, m B refers to the density integral of the second basis material, d is the differential sign, E is the energy value, and l is the preset integral path length;

[0129] Generate a primary mapping table according to the first base material density integral, the second base material density integral, the low-energy ray attenuation value, and the high-energy ray attenuation value;

[0130] The primary mapping table is uniformly interpolated with equal spacing to obtain a dual-energy decomposition mapping table.

[0131] In detail, the virtual base material phantom is obtained by selecting a first base material A and a second base material B (such as water, calcium, etc.) for decomposition, and designing two three-dimensional phantoms of the same volume with density distributions of A and B respectively. The density values ​​of the two require that the density integral on the longest integral path must exceed the maximum possible density integral value of the actual scanned object, to ensure that the density integral values ​​of materials A and B are as evenly distributed as possible within the maximum density integral value range.

[0132] In detail, the virtual base matter phantom is a conceptual tool used in dual-energy CT and multi-energy CT imaging technologies. The virtual base matter phantom is used to decompose and analyze the material composition inside an object by using a model based on known materials to describe and calibrate the absorption and attenuation characteristics of different substances to X-rays. The core idea of ​​the virtual base matter phantom is to use the known physical properties of the base substance to infer the composition and distribution of unknown substances.

[0133] Specifically, by integrating X-rays of different energies, the attenuation of the X-rays when passing through an object can be calculated more accurately, thereby obtaining a more accurate dual-energy decomposition mapping table.

[0134] For details, refer to Figure 5 As shown, it is a schematic diagram of the dual energy decomposition mapping table, where P H The value refers to the high energy projection attenuation value, P L The value refers to the low-energy projection attenuation value. refers to the uth P L value, refers to the vth P H value, It means P L Value is And P H Value is The first basis material density integral at time , It means P L Value is And P H Value is The second basis material density integral at .

[0135] In another embodiment of the present invention, the dual-energy decomposition mapping table can be reversely established according to the high- and low-energy spectrum curves by a fixed grid method, including:

[0136] Obtaining a first basis material density integral interval and a second basis material density integral interval;

[0137] Generate a first basis material density sequence according to the first basis material density integral interval, and generate a second basis material density sequence according to the second basis material density integral interval;

[0138] Calculate a high-energy projection attenuation value sequence and a low-energy projection attenuation value sequence according to the first basis material density sequence and the second basis material density sequence;

[0139] Generate a primary mapping table according to the first basis material density sequence, the second basis material density sequence, the high energy projection attenuation value sequence, and the low energy projection attenuation value sequence;

[0140] Perform bilinear interpolation on the primary mapping table to obtain a dual-energy decomposition mapping table.

[0141] In the embodiment of the present invention, by performing relationship mapping and projection domain decomposition, noise and artifacts in images of different energies can be eliminated, image contrast can be enhanced, and different substances can be easily distinguished, material identification and accurate quantification can be achieved, thereby improving the accuracy of subsequent panoramic imaging.

[0142] S5. Perform variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base map group.

[0143] In detail, the panoramic base image set refers to an image set of panoramic base images of oral cavity projection, wherein the panoramic base image is an image that provides a panoramic view of the oral cavity area, covering all teeth and related structures from one side to another.

[0144] In the embodiment of the present invention, the step of performing variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base image group includes:

[0145] Performing ramp filtering on the base material density integral decomposition result to obtain a filtered base material decomposition result;

[0146] Performing weighted calculation on the filtering basis material decomposition result to obtain a weighted basis material decomposition result;

[0147] Perform multi-layer dental arch curve back-projection on the weighted base material decomposition result to obtain a hierarchical panoramic base image group;

[0148] Automatically focus and fuse the hierarchical panoramic base image group to obtain a panoramic base image group.

[0149] In detail, the weighted calculation refers to weighted accumulation of the integral values ​​of the base material density of all preset reconstruction points in the filtered base material decomposition result to obtain a panoramic base map of the material, and the multi-layer dental arch curve back projection refers to using preset multi-layer parallel dental arch curves to calculate panoramic images of different depths corresponding to the weighted base material decomposition result to obtain a hierarchical panoramic base map group.

[0150] Specifically, the automatic focus fusion refers to calculating the clarity of each part of each level of the panoramic base image in the hierarchical panoramic base image group using a local clarity detection algorithm, and selecting each local image with the highest clarity to fuse and reorganize into a panoramic base image group.

[0151] In detail, other images with tissue-specific identification functions, such as virtual conventional images, virtual decalcified images, virtual monoenergetic images, and electron density images, can be generated based on the panoramic base image group.

[0152] In the embodiment of the present invention, by performing variable-center limited-angle reconstruction and panoramic base map imaging, the reconstruction accuracy can be improved and local area optimization can be achieved, thereby fully displaying high-contrast and high-detail oral structures and improving the accuracy of panoramic images.

[0153] The present invention utilizes the dual-energy rays for projection imaging and can utilize the specific attenuation characteristics of different substances under X-rays at different energy levels to decompose materials, which can significantly improve the recognition accuracy of specific oral tissues. By performing sparse compensation, deep learning or linear interpolation methods can be used to interpolate projection images at intermediate angles, thereby avoiding quality degradation during subsequent image reconstruction caused by angular sparseness of the projection image.

[0154] By performing energy spectrum measurement on the dual-energy rays, the accuracy of energy spectrum measurement can be improved, the calibration process can be simplified, and reliable benchmark data can be provided. By performing relationship mapping and projection domain decomposition, noise and false sound in different energy images can be eliminated, image contrast can be enhanced, and different substances can be easily distinguished, material identification and accurate quantification can be achieved, thereby improving the accuracy of subsequent panoramic imaging. By performing variable center limited angle reconstruction and panoramic base map imaging, the reconstruction accuracy can be improved, local area optimization can be achieved, and high-contrast and high-detail oral structures can be fully displayed, thereby improving the accuracy of panoramic images. Therefore, the dual-energy-based panoramic imaging method proposed in the present invention can solve the problem of poor expressiveness of imaging results when performing oral panoramic imaging.

[0155] like Figure 6 , is a functional module diagram of a dual-energy based panoramic imaging device provided in one embodiment of the present invention.

[0156] The dual-energy based panoramic imaging device 100 of the present invention can be installed in an electronic device. According to the functions to be implemented, the dual-energy based panoramic imaging device 100 can include a projection acquisition module 101, a sparse compensation module 102, an energy spectrum measurement module 103, a mapping decomposition module 104 and a base map reconstruction module 105. The module of the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.

[0157] In this embodiment, the functions of each module / unit are as follows:

[0158] The projection acquisition module 101 is used to generate dual-energy rays by using a high-voltage fast-cut method, and use the dual-energy rays to perform projection imaging to obtain a high-energy and low-energy sparse projection image group;

[0159] The sparse compensation module 102 is used to perform sparse compensation on the high and low energy sparse projection image group to obtain a high and low energy projection image group;

[0160] The energy spectrum measurement module 103 is used to measure the energy spectrum of the dual-energy ray to obtain high-energy and low-energy spectrum curves;

[0161] The mapping decomposition module 104 is used to generate a dual-energy decomposition mapping table according to a preset virtual base material model and the high- and low-energy energy spectrum curves, and perform projection domain decomposition on the high- and low-energy projection image group according to the dual-energy decomposition mapping table to obtain a base material density integral decomposition result;

[0162] The base image reconstruction module 105 is used to perform variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base image group.

[0163] In detail, each module described in the dual-energy panoramic imaging device 100 described in the embodiment of the present invention is used in the same manner as described above. Figure 1 The technical means are the same as the dual-energy panoramic imaging method described in, and can produce the same technical effects, so they will not be repeated here.

[0164] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0165] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0166] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0167] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0168] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.

[0169] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application device that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0170] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units or devices stated in the device embodiment can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

[0171] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A dual-energy based panoramic imaging method, characterized in that: The method comprises: Generate dual-energy rays by using a high-voltage fast-cut method, and use the dual-energy rays for projection imaging to obtain a high- and low-energy sparse projection image group; Performing sparse compensation on the high and low energy sparse projection image group to obtain a high and low energy projection image group; Performing energy spectrum measurement on the dual-energy rays to obtain high-energy and low-energy spectrum curves; Generate a dual-energy decomposition mapping table according to a preset virtual base material model and the high- and low-energy energy spectrum curves, and perform projection domain decomposition on the high- and low-energy projection image group according to the dual-energy decomposition mapping table to obtain a base material density integral decomposition result; The base material density integral decomposition result is reconstructed with a variable center finite angle to obtain a panoramic base map group.

2. The dual-energy based panoramic imaging method according to claim 1, characterized in that: The step of measuring the energy spectrum of the dual-energy rays to obtain high-energy and low-energy spectrum curves includes: Using the dual-energy rays to perform exposure scanning on a preset pure measurement phantom to obtain a projection data set; Selecting projection data in the projection data group one by one as target projection data, performing back-projection reconstruction, three-dimensional segmentation, re-assignment and forward projection on the target projection data to obtain an integral path length; Initialize the initial energy spectrum curve, perform energy spectrum attenuation calculation according to the initial energy spectrum curve and the integral path length, and obtain the predicted ray attenuation value: Wherein, P refers to the predicted ray attenuation value, refers to the initial energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the pure measurement phantom for energy E, μ is the linear attenuation coefficient, ρ is the density symbol, l is the integral path length, d is the differential symbol, and E is the energy value; calculating a projection attenuation difference between the predicted ray attenuation value and the target projection data; The initial energy spectrum curve is iteratively updated according to the projection attenuation difference to obtain an energy spectrum curve, and high and low energy spectrum curves are generated according to the energy spectrum curves of all target projection data in the projection data group.

3. The dual-energy based panoramic imaging method according to claim 2, characterized in that: The step of performing back-projection reconstruction, three-dimensional segmentation, re-assignment and forward projection on the target projection data to obtain an integral path length includes: Performing back-projection reconstruction on the target projection data to obtain a reconstruction result; Performing threshold segmentation on the reconstruction result according to a preset attenuation coefficient interval to obtain a segmented region phantom; Reassigning voxels of the segmented region phantom to obtain a voxel-reassigned phantom; The voxel re-values ​​are forward projected to obtain the integrated path length.

4. The dual-energy based panoramic imaging method according to claim 1, characterized in that: The generating of the dual energy decomposition mapping table according to the preset virtual base material model and the high and low energy spectrum curves includes: Performing forward projection on a preset virtual base material model to obtain a first base material density integral and a second base material density integral; Extracting a low energy spectrum curve and a high energy spectrum curve from the high and low energy spectrum curves; The low-energy ray attenuation value is calculated according to the first base material density integral, the second base material density integral and the low-energy spectrum curve: m A =∫ρ A (l)dl m B =∫ρ B (l)dl Among them, P L refers to the low energy ray attenuation value, L refers to low energy, S L (E) refers to the low energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the first base material A preset in the virtual base material model, μ A is the linear attenuation coefficient of the first base material A preset in the virtual base material phantom, ρ A is the density of the first base material A preset in the virtual base material model, mA refers to the integral of the density of the first base material, is the mass attenuation coefficient of the second base material B preset in the virtual base material model, μ B is the linear attenuation coefficient of the second base material B preset in the virtual base material phantom, ρ B is the density of the first base material B preset in the virtual base material model, m B refers to the density integral of the second basis material, d is the differential sign, E is the energy value, and l is the preset integral path length; The high-energy ray attenuation value is calculated according to the first base material density integral, the second base material density integral and the high-energy spectrum curve: Among them, P H It refers to the attenuation value of the high-energy ray, H refers to high energy, S H (E) refers to the high energy spectrum curve for energy E, exp is the exponential function symbol, is the mass attenuation coefficient of the first base material A preset in the virtual base material model, μ A is the linear attenuation coefficient of the first base material A preset in the virtual base material phantom, ρ A is the density of the first base material A preset in the virtual base material model, m A refers to the density integral of the first basis substance, is the mass attenuation coefficient of the second base material B preset in the virtual base material model, μ B is the linear attenuation coefficient of the second base material B preset in the virtual base material phantom, ρ B is the density of the first base material B preset in the virtual base material model, m B refers to the density integral of the second basis material, d is the differential sign, E is the energy value, and l is the preset integral path length; Generate a primary mapping table according to the first base material density integral, the second base material density integral, the low-energy ray attenuation value, and the high-energy ray attenuation value; The primary mapping table is uniformly interpolated with equal spacing to obtain a dual-energy decomposition mapping table.

5. The dual-energy based panoramic imaging method according to claim 1, characterized in that: The step of performing projection domain decomposition on the high and low energy projection image group according to the dual energy decomposition mapping table to obtain a basis material density integral decomposition result includes: Performing attenuation calculation on the high- and low-energy projection image group to obtain a high- and low-energy attenuation value pair; According to the method of nearest search, a neighborhood value group closest to the high and low energy attenuation values ​​is screened out from the dual energy decomposition mapping table to obtain a nearest attenuation value group; Performing bilinear interpolation on the group of most recent attenuation values ​​to obtain a density decomposition value; Density reconstruction is performed according to the density decomposition value to obtain a base material density integral decomposition result.

6. The dual-energy based panoramic imaging method according to claim 1, characterized in that: The step of performing variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base image group includes: Performing ramp filtering on the base material density integral decomposition result to obtain a filtered base material decomposition result; Performing weighted calculation on the filtering basis material decomposition result to obtain a weighted basis material decomposition result; Perform multi-layer dental arch curve back-projection on the weighted base material decomposition result to obtain a hierarchical panoramic base image group; Automatically focus and fuse the hierarchical panoramic base image group to obtain a panoramic base image group.

7. A dual-energy panoramic imaging device, characterized in that: The device comprises: A projection acquisition module, used to generate dual-energy rays by using a high-voltage fast-cut method, and use the dual-energy rays for projection imaging to obtain a high- and low-energy sparse projection image group; A sparse compensation module, used for performing sparse compensation on the high and low energy sparse projection image group to obtain a high and low energy projection image group; An energy spectrum measurement module, used for measuring the energy spectrum of the dual-energy rays to obtain high-energy and low-energy spectrum curves; A mapping decomposition module, used to generate a dual-energy decomposition mapping table according to a preset virtual base material model and the high- and low-energy energy spectrum curves, and perform projection domain decomposition on the high- and low-energy projection image group according to the dual-energy decomposition mapping table to obtain a base material density integral decomposition result; The base image reconstruction module is used to perform variable center finite angle reconstruction on the base material density integral decomposition result to obtain a panoramic base image group.

Images