Model training method, method for determining density distribution of material elements, device, equipment and medium

Abstract

The present disclosure provides a model training method, a method for determining the elemental density distribution of a substance, a device, an electronic device, and a storage medium, relating to the fields of physics and computer science, and particularly to the application of physics and computer technologies in the field of radiation medicine. The model training method includes: obtaining the characteristic parameters of each of the M scanning energy spectra of a computed tomography (CT) device; determining M sample CT images of each of the N sample materials under the M scanning energy spectra based on the M characteristic parameters and the elemental density distributions of the N sample materials, obtaining M×N sample CT images; constructing a sample data set based on the M sample CT images and the elemental density distributions of each of the N sample materials; and training an initial model using the sample data set to obtain a target model corresponding to the CT device, the target model being used to determine the elemental density distribution based on the CT image obtained by scanning with the CT device.

Description

Technical Field

[0001] The present disclosure relates to the fields of physics and computer science, and more particularly to the application of physics and computer technologies in the field of radiation medicine. More specifically, the present disclosure relates to a model training method, a method for determining the density distribution of material elements, an apparatus, an electronic device, and a storage medium. Background Art

[0002] Computed Tomography (CT) imaging technology has been widely used in the field of medical diagnosis. Through CT imaging technology and image reconstruction technology, the internal structure of the detected object can be presented. For example, based on CT images, the density distribution of material elements can be obtained, and the element density distribution is of great significance in radiation diagnosis and radiation therapy. Summary of the Invention

[0003] The present disclosure provides a model training method, a method for determining the density distribution of material elements, an apparatus, an electronic device, a storage medium, and a program product.

[0004] According to one aspect of the present disclosure, there is provided a model training method, including: obtaining the characteristic parameters of each of the M scanning energy spectra of a Computed Tomography (CT) device; determining M sample CT images of each of the N sample materials under the M scanning energy spectra based on the M characteristic parameters and the element density distributions of the N sample materials, to obtain M×N sample CT images; constructing a sample data set based on the M sample CT images and the element density distributions of each of the N sample materials; and training an initial model using the sample data set to obtain a target model corresponding to the CT device, where the target model is used to determine the element density distribution based on the CT image scanned by the CT device.

[0005] According to another aspect of the present disclosure, there is provided a method for determining the density distribution of material elements, including: scanning a material to be measured using each of the M scanning energy spectra of a CT device to obtain M CT images; and determining the element density distribution of the material to be measured based on the M CT images using an element analysis model, where the element analysis model is a target model trained using the model training method provided in the embodiments of the present disclosure.

[0006] According to another aspect of the present disclosure, there is provided a model training apparatus, including: an acquisition module configured to acquire characteristic parameters of each of M scanning energy spectra of a computed tomography (CT) device; a first determination module configured to determine M sample CT images of each of N sample materials under the M scanning energy spectra based on the M characteristic parameters and the elemental density distributions of the N sample materials, thereby obtaining M×N sample CT images; a construction module configured to construct a sample data set based on the M sample CT images and the respective elemental density distributions of each of the N sample materials; and a training module configured to train an initial model using the sample data set to obtain a target model corresponding to the CT device, where the target model is used to determine the elemental density distribution based on a CT image obtained by scanning with the CT device.

[0007] According to another aspect of the present disclosure, there is provided an apparatus for determining the elemental density distribution of a substance, including: a scanning module configured to respectively scan a substance to be measured using M scanning energy spectra of a CT device to obtain M CT images; and a second determination module configured to determine the elemental density distribution of the substance to be measured based on the M CT images using an elemental analysis model, where the elemental analysis model is a target model trained using the model training apparatus provided in an embodiment of the present disclosure.

[0008] Another aspect of the present disclosure provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; where the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the model training method and / or the method for determining the elemental density distribution provided by the present disclosure.

[0009] According to another aspect of the embodiments of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to cause a computer to execute the model training method and / or the method for determining the elemental density distribution provided by the present disclosure.

[0010] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. Description of the Drawings

[0011] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0012] Figure 1 is a schematic flowchart of the model training method according to an embodiment of the present disclosure;

[0013] Figure 2 is a schematic diagram of the scenario of the model training method according to an embodiment of the present disclosure;

[0014] Figure 3 Schematic diagram of the principle of the target model according to an embodiment of the present disclosure;

[0015] Figure 4 Flow chart of the method for determining the density distribution of material elements according to an embodiment of the present disclosure;

[0016] Figure 5 Block diagram of the structure of the model training device according to an embodiment of the present disclosure;

[0017] Figure 6 Block diagram of the structure of the device for determining the density distribution of material elements according to an embodiment of the present disclosure; and

[0018] Figure 7 Schematic block diagram of an exemplary electronic device for implementing the embodiments of the present disclosure. Detailed implementation manners

[0019] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, descriptions of well-known functions and structures are omitted below for clarity and conciseness.

[0020] The terms used herein are merely for describing specific embodiments and are not intended to limit the present disclosure. The terms "including", "comprising", etc. used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0021] In the technical solution of the present disclosure, the processing of data (such as including but not limited to user personal information) such as collection, storage, use, processing, transmission, provision, disclosure, and application complies with the provisions of relevant laws and regulations, takes necessary confidentiality measures, and does not violate public order and good customs.

[0022] Figure 1 Flow chart of the model training method according to an embodiment of the present disclosure.

[0023] As Figure 1 shown, the model training method 100 of this embodiment may include operation S110 to operation S140.

[0024] In operation S110, obtain the characteristic parameters of each of the M scanning energy spectra of a computed tomography (CT) device, where M is a positive integer.

[0025] In the embodiments of the present disclosure, when a CT device scans the same tissue material based on multiple scanning energy spectra, the obtained CT images may be different. The CT image includes the distribution of HU values of the tissue material, and the HU value can reflect the interaction between the substance and photons.

[0026] The different scanning energy spectra may be formed based on different radiation sources, and the energy components included in the multiple scanning energy spectra may be different. When the CT device scans the same tissue material based on different scanning energy spectra, the HU values may be the same or different.

[0027] For example, a dual-energy CT device has two scanning energy spectra. After scanning multiple tissue materials based on a single radiation source, the elemental compositions of the multiple tissue materials are different, but the HU values under a single energy spectrum may be the same. In this case, it is impossible to further distinguish between the two tissue materials in the CT image, which may lead to deviation in the diagnostic results. Therefore, by using dual-energy CT, two CT images obtained by scanning two tissue materials respectively based on the two energy spectra can be compared. When the HU values of the two tissue materials are the same under a certain energy spectrum, the HU values under another energy spectrum are often different, so as to further distinguish between the two tissue materials.

[0028] In the embodiments of the present disclosure, the characteristic parameters can describe the energy combination of multiple emission sources of the CT device. For example, the emission source may be composed of multiple sub-emission sources with different energies, the photon energy ranges emitted by the multiple sub-emission sources may be within 20-120 keV, and the photon energies emitted by the multiple sub-emission sources form corresponding scanning energy spectra.

[0029] For example, among the multiple emission sources, some emission sources emit more 40 keV photon energy, some emission sources emit more 60 keV photon energy, and some emission sources emit more 80 keV photon energy.

[0030] Based on the characteristic parameters of the scanning energy spectrum, the situation of the photon energies emitted by the multiple emission sources can be determined.

[0031] In operation S120, based on M characteristic parameters and the elemental density distributions of N sample materials, M sample CT images of each of the N sample materials under M scanning energy spectra are determined, obtaining M*N sample CT images, where N is a positive integer.

[0032] In the embodiments of the present disclosure, the elemental density distribution of the sample material can be obtained from a database in advance. The elemental density distribution can describe the types of elements included in the sample material, the density of each element in the sample material, and the distribution of multiple elements in the sample material.

[0033] Based on the characteristic parameters, the composition of the photon energy emitted by the emission source corresponding to the scanning energy spectrum can be determined. Based on the elemental density distribution of the sample material, the absorption degree of photons with different energies by the sample material can be determined, so as to calculate the sample CT image of the sample material.

[0034] Since different CT devices are based on different scanning energy spectra, the CT images obtained by scanning the same material tissue are different. In addition, the known database only records the relevant parameters of the sample material and cannot obtain the real tissue with the same characteristics as the sample material. Therefore, it is impossible to use the CT device to scan the real tissue of the sample material to determine the CT image of the sample material obtained based on the CT device scan. Therefore, the present disclosure simulates the virtual scanning environment of the CT device based on the characteristic parameters and simulates the virtual tissue of the sample material based on the elemental density distribution of the sample material. In this case, calculate the sample CT image obtained by scanning the virtual tissue in the virtual scanning environment.

[0035] In the embodiments of the present disclosure, the sample CT image is calculated based on the characteristic parameters of the elemental density distribution and the scanning energy spectrum. Based on the characteristic parameters, the scanning environment of the CT device can be simulated to determine the photon energy emitted by the emission source. Based on the elemental density distribution, the elemental density at different positions in the sample material can be simulated, so as to calculate the absorption ability of different positions in the sample material to the photon energy, and then calculate the HU values at different positions in the sample material based on the absorption ability to obtain the sample CT image.

[0036] In operation S130, a sample data set is constructed based on the M sample CT images and the respective elemental density distributions of N sample materials.

[0037] In the embodiments of the present disclosure, the sample CT image includes multiple HU values, and the elemental density distribution includes the elemental density of the region corresponding to each HU value. For example, each sample material includes multiple voxels. In the M sample CT images of each sample material, each voxel corresponds to M HU values. Understandably, the M HU values correspond to the elemental density of the voxel.

[0038] For example, the voxels characterized by the M HU values determined under different scanning energy spectra represent the same elemental density. The sample data set constructed based on the sample CT image and the elemental density includes multiple sets of corresponding relationships between the HU values and the elemental density.

[0039] In operation S140, the initial model is trained using the sample data set to obtain the target model corresponding to the CT device.

[0040] In an embodiment of the present disclosure, the initial model may be a fully connected neural network. By training the fully connected neural network using a sample data set, the correlation between HU values and elemental densities can be extracted using fully connected operations, such that the trained target model can determine the elemental density distribution based on the CT images obtained by scanning with a CT device.

[0041] For example, the fully connected neural network can extract the features of HU values and the features of elemental densities, and perform fully connected operations on the extracted features to determine the correspondence between HU values and elemental densities.

[0042] In the target model, the HU value is the input data and the elemental density is the output data. Therefore, by inputting the CT image of the substance to be measured into the target model, the elemental densities of different regions in the substance to be measured can be output, thereby determining the elemental density distribution.

[0043] Combined with Figure 2 the illustrated scenario for description. Figure 2 FIG. is a schematic diagram of a scenario of a model training method according to an embodiment of the present disclosure.

[0044] As Figure 2 shown, the sample material 200 can be divided into multiple objects. For example, the sample material 200 may have a cubic structure, and the sample material 200 is evenly divided into multiple objects, and each object also has a cubic structure.

[0045] In an embodiment of the present disclosure, each object may be a voxel of the sample material 200, and each object can be equivalent to a point inside the sample material 200. According to the elemental densities of the multiple objects respectively, the elemental density distribution of the sample material 200 can be determined.

[0046] For example, the sample material 200 is divided into 5×3×3 objects. The elemental density distribution may include the density distributions of carbon, hydrogen, oxygen, nitrogen, phosphorus, and calcium elements. For example, the carbon elemental density distribution of the sample material 200 can be represented by a 5×3×3 three-dimensional matrix.

[0047] Based on the characteristic parameters of multiple scanning energy spectra and the elemental density distribution of the sample material 200, the HU values of 30 objects respectively can be calculated. Among them, based on M scanning energy spectra, M HU values of each object can be calculated. The M HU values of each object all correspond to the same elemental density.

[0048] For example, based on M scanning energy spectra, M HU values of the object 201 can be calculated. The M HU values of the object 201 are correlated with the elemental density of the object 201. Based on the 30 sets of correlated HU values and elemental densities of the sample material 200, a sample data set can be constructed.

[0049] Through the embodiments of the present disclosure, a sample dataset for a CT device can be constructed based on the elemental density distribution of a sample material and the spectral characteristic parameters of the CT device. Using the data pairs of HU values and elemental densities included in the sample dataset to train an initial model, so that the initial model can learn the correlation between the HU values and the elemental densities, and a target model is obtained. After learning the correlation between the HU values and the elemental densities under the scanning energy spectrum constraint of the CT device, the target model can output the elemental density distribution of the substance to be measured corresponding to the CT image obtained by scanning the CT device. In addition, the model training method provided by the present disclosure can be implemented based on a fully connected neural network, with less hardware resources and computing resources required for model training, which can reduce resource overhead.

[0050] In some embodiments, the characteristic parameters of the scanning energy spectrum of the CT device can be determined by a known standard phantom. For example, scanning a reference material with M scanning energy spectrum distributions of the CT device to obtain M scanning data of the reference material; and comparing the M scanning data with the reference data of the reference material obtained in advance to obtain M characteristic parameters of the M scanning energy spectra.

[0051] In the embodiments of the present disclosure, the reference material can be a known standard phantom. For example, information such as the elemental density distribution of the reference material, the total density of the material, and the CT image obtained based on the known scanning energy spectrum are all known.

[0052] For example, the reference material can be a Gammex phantom and a CIRS phantom, etc. Scanning the Gammex phantom with M scanning energy spectra by the CT device to obtain M CT images of the Gammex phantom as scanning data. Based on the known CT image, elemental density distribution and total material density of the Gammex phantom, the absorption degree of each voxel in the Gammex phantom for a specific photon energy can be determined. Based on the absorption degree and the scanning data, the photon energy emitted by the emission source of the CT device is calculated, thereby determining the characteristic energy spectrum of the CT device.

[0053] For example, the absorption ability of a voxel with a certain material in the Gammex phantom for a ray with a specific energy can be considered to be fixed. Based on the known CT image of the Gammex phantom, the corresponding relationship between the HU value and the photon energy of the energy spectrum is determined. Based on this corresponding relationship, based on the HU value in the scanning data, the photon energy of the energy spectrum is calculated, so that M characteristic parameters of the M scanning energy spectra can be determined based on the M CT images in the scanning data.

[0054] In some embodiments, the method for calculating M*N sample CT images obtained by a CT device scanning N sample materials may include: determining the energy absorption capabilities of each of the N sample materials for M scanning energy spectra by using M characteristic parameters and the elemental density distributions of the N sample materials; and determining M sets of CT values for each of the N sample materials based on the energy absorption capabilities, where the M sets of CT values form M CT images.

[0055] In the embodiments of the present disclosure, each sample material may be divided into a plurality of voxels, and the absorption ability of each voxel for photon energy is determined by the density and elemental distribution of the voxel. For example, if the densities of calcium and phosphorus elements in a voxel are large, the voxel may be bone tissue, with a strong absorption ability for photon energy and a relatively large corresponding HU value. If the densities of carbon, oxygen, hydrogen, and nitrogen elements in a voxel are large, the voxel may be soft tissue, with a weak absorption ability for photon energy and a relatively small corresponding HU value.

[0056] Based on the absorption ability of each voxel in the sample material for photon energy and the correspondence between the HU value and the photon energy of the energy spectrum, calculate the HU value of each voxel in the sample material, and based on the HU value, the CT image of the sample material can be reconstructed.

[0057] The absorption abilities of each voxel in the sample material for different photon energies are different. Therefore, for the scanning energy spectra described by different characteristic parameters, different HU values can be calculated for each voxel. For M characteristic parameters, M HU values can be calculated for each voxel.

[0058] In some embodiments, the method for constructing a sample data set may include, for each of the N sample materials: determining M CT values of each of the multiple sample objects in the sample material based on M sample CT images; obtaining M elemental densities corresponding to the M CT values based on the elemental density distribution; determining the sample data of each of the multiple sample objects based on the M CT values and the M elemental densities; and constructing a sample data set of the sample material based on the sample data.

[0059] In the embodiments of the present disclosure, the sample object may be a voxel in the sample material, and the CT value is the HU value of each voxel in the CT image. The HU values of multiple voxels in the sample material may be represented in the form of a three-dimensional matrix, and the elemental density distribution may also be the elemental density of each voxel represented in the form of a three-dimensional matrix. In the three-dimensional matrix, the HU value and the elemental density of the same voxel have the same coordinates. Among the three-dimensional matrices of the HU values of multiple M sample CT images, the M HU values with the same coordinates correspond to the elemental density at this coordinate in the three-dimensional matrix of the elemental density distribution.

[0060] The M HU values at the same coordinates and the corresponding elemental densities form data pairs for the voxel. These data pairs can be used to describe the data on the correlation between the HU values and elemental densities of the voxel. The data pairs of each voxel in N sample materials form a sample data set. The sample data set describes the correlation between the HU values and elemental densities of each voxel.

[0061] In some embodiments, data augmentation can also be performed on the sample data of each sample material to improve the accuracy of the sample data. For example, within a preset range, multiple sets of augmented elemental densities are generated based on the M elemental densities. Each set of augmented elemental densities includes M augmented elemental densities, and the difference between the M elemental densities and each set of augmented elemental densities is within the preset range; and based on the M CT values of each of the multiple sample objects and the multiple sets of augmented elemental densities, the sample data of each of the multiple sample objects is determined.

[0062] In the embodiments of the present disclosure, the accuracy of the elemental density distribution output by the target model is related to the quality of the sample data. Performing data augmentation on the sample data can increase the richness of the sample data, thereby improving the quality of the target model.

[0063] For example, due to errors, uncertainties in the elemental density and HU values, sample augmentation can be performed on the elemental density of each voxel in the sample material, so that the M HU values of each voxel can correspond to multiple elemental densities, in order to increase the data pairs used to characterize the correlation between the HU values and elemental densities of each voxel, and to reduce the error of the data pairs before data augmentation.

[0064] For example, assume that the elemental density of the voxel of the sample material follows a Gaussian distribution. For a sample material, 100 sets of elemental densities are randomly generated with the elemental density of the voxel as the central value and 10% of the central value as the standard deviation. Each set of elemental densities includes the densities of multiple elements. Based on these 100 sets of elemental densities and the M HU values of the voxel, 100 data pairs are constructed, so that the original set of data pairs can be augmented to 100 data pairs.

[0065] For example, data augmentation can also be performed on the overall density of the sample material, and the overall density of the sample material and the CT image are constructed into data pairs. The data pairs of the overall density of the sample material and the CT image can describe the correlation between the overall density of the sample material and the CT image (all HU values), and can also reflect the influence of the correlation between the HU values of adjacent voxels on the overall density.

[0066] For example, for a sample material, 100 overall densities are randomly generated with the overall density of the sample material as the central value and 10% of the central value as the standard deviation. Based on the 100 overall densities and M sample CT images of the sample material, 100 data pairs are constructed, so that a single original data pair can be enhanced to 100 data pairs.

[0067] For example, data augmentation can also be performed on the elemental density distribution of the sample material, and the elemental density distribution and the CT image are constructed into a data pair. Constructing the elemental density distribution and the CT image into a data pair can describe the correlation between the overall elemental density distribution of the sample material and the CT image, and can also reflect the influence of the correlation between the elemental densities of adjacent voxels on the HU value.

[0068] For example, for a sample material, 100 elemental density distributions are randomly generated with the elemental density distribution of the sample material as the central value and 10% of the central value as the standard deviation. Based on the 100 elemental density distributions and M sample CT images of the sample material, 100 data pairs are constructed, so that a single original data pair can be enhanced to 100 data pairs.

[0069] In the embodiments of the present disclosure, data augmentation can also be performed on the overall density, elemental density distribution, and elemental density of a single voxel of the sample material respectively, and a data pair is constructed based on at least one of the enhanced overall density, enhanced elemental density distribution, and enhanced elemental density of a single voxel and the CT image or the HU value of a single voxel.

[0070] Based on this, the data augmentation operation can expand the data volume, reduce data errors, improve the uncertainty in the data, thereby optimizing the correlation between the two data in the data pair, and increasing the robustness of the model training against CT image noise and the uncertainty of the elemental density distribution of tissue materials. In addition, constructing data pairs based on the overall density and elemental density distribution can fully consider the correlation and mutual influence between voxels, and avoid ignoring the correlation between data in the data pairs constructed based on a single voxel, thereby improving the quality of the target model.

[0071] In some embodiments, training an initial model using a sample data set to obtain a target model corresponding to a CT device may include: determining a sample CT image in the sample data set as input data for the initial model, and determining an elemental density distribution in the sample data set as output data for the initial model; using the distribution characteristics of multiple elements in the elemental density distribution as hint information; and training the initial model based on the hint information, input data, and output data to obtain the target model.

[0072] In the embodiments of the present disclosure, a sample CT image is used as the input data of the model, and the elemental density of the sample material is used as the output data of the model, so that the model can calculate and output the elemental density distribution of the substance to be measured based on the CT image of the substance to be measured.

[0073] For example, the M HU values of a single voxel can be used as the input data, and the elemental density of a single voxel can be used as the output data, so that the model can calculate and output the elemental density of each voxel based on the HU value of each voxel in the substance to be measured, thereby determining the elemental density distribution of the substance to be measured.

[0074] The initial model can learn the correlation between the HU value and the elemental density of a single voxel described by the data pair in the sample dataset, can also learn the influence of the elemental density between adjacent voxels on the HU value, can also learn the influence of the overall elemental density distribution of the sample material on the HU value of a single voxel, and can also learn the influence of the overall density of the sample material on the HU value of a single voxel.

[0075] In the embodiments of the present disclosure, the distribution characteristics of each of the multiple elements in the elemental density distribution can be the law of how the HU value changes with the elemental density. For example, for different elements, the trend of how the elemental density changes with the HU value is different. Based on the distribution characteristics of each element as prompt information, the accuracy of the elemental density of each element output by the target model based on the HU value can be improved.

[0076] For example, based on the HU values of each voxel in the sample material, the relative electron density ρ e and the effective atomic number Z eff of each voxel can be calculated to determine the variation law between the relative electron density ρ e , the effective atomic number Z eff and the elemental density of multiple voxels.

[0077] For example, based on the relative electron density ρ e and the effective atomic number Z eff of the voxel, it can be determined whether the voxel belongs to bone tissue or soft tissue. For example, if the effective atomic number Z eff is greater than 8.2, it is determined that the material type of the substance to be measured is bone tissue. If the effective atomic number Z eff is less than or equal to 8.2, it is determined that the material type of the substance to be measured is soft tissue. For example, the substance to be measured can be animal tissue. Bone tissue can be bones, etc. There are a large amount of calcium and phosphorus elements distributed in bone tissue, as well as a certain amount of carbon, oxygen, hydrogen, and nitrogen elements distributed. Soft tissue includes a large amount of fat and protein, etc. There are a large amount of carbon, oxygen, hydrogen, and nitrogen elements distributed in soft tissue, as well as a small amount of calcium and phosphorus elements. The content of calcium and phosphorus elements in soft tissue is usually less than 0.2%, so the calcium and phosphorus elements included in soft tissue can be ignored.

[0078] For example, for bone tissue, the sum of the mass fractions of oxygen and carbon elements is negatively correlated with the relative electron density ρ e and the effective atomic number Z eff respectively. The mass fraction of calcium element is positively correlated with the relative electron density ρ e and the effective atomic number Z eff respectively. The mass fraction of phosphorus element is positively correlated with the relative electron density ρ e and the effective atomic number Z eff respectively. The mass fractions of carbon and hydrogen elements change with the relative electron density ρ e and the effective atomic number Z eff respectively, and can satisfy the constraints of the arctangent function. The mass fraction of nitrogen element can be determined by subtracting the sum of the mass fractions of calcium, phosphorus, carbon, oxygen, and hydrogen elements from 100%. The sum of the mass fractions of carbon, oxygen, hydrogen, nitrogen, calcium, and phosphorus elements is greater than 99% and less than 100%.

[0079] For soft tissue, the mass fractions of oxygen and carbon elements are negatively correlated with each other. The mass fraction of carbon element decreases as the oxygen element increases.

[0080] Based on the mass fractions of calcium, phosphorus, carbon, oxygen, hydrogen, and nitrogen elements, the density distributions of calcium, phosphorus, carbon, oxygen, hydrogen, and nitrogen elements can be determined.

[0081] In the embodiments of the present disclosure, based on the rule that the mass fractions of the elements in the voxel change with the relative electron density ρ e and the effective atomic number Z eff , the rule that the mass fractions of the elements in the voxel change with the HU value is determined, and then the rule that the element densities of the elements in the voxel change with the HU value is determined. The initial model learns the rule that the element density changes with the HU value, so that the target model can improve the accuracy of calculating the element densities of the voxels based on the HU value based on this change rule.

[0082] For example, when the target model calculates the element densities of multiple voxels based on the HU values of the multiple voxels, the element densities of each voxel can be optimized and adjusted based on the change rule described in the prompt information. The target model can also assist in calculating the element density that is difficult to calculate based on the HU value based on the change rule described in the prompt information and the element density that has been calculated based on the HU value.

[0083] Combined Figure 3 An exemplary description of the training process is given Figure 3Schematic diagram of the principle of the target model according to an embodiment of the present disclosure.

[0084] In an embodiment of the present disclosure, a CT device is used to scan a substance to be measured based on M scanning energy spectra to obtain M CT images. Based on the M CT images, M HU values HU 1 、HU 2 、…、HU M of each voxel in the substance to be measured are determined.

[0085] The input data 301 of the target model 303 is the M HU values HU 1 、HU 2 、…、HU M of each voxel, and the output data 302 is the elemental density ρ j of each voxel. j can be H, C, N, O, P, and Ca, respectively representing the elemental densities of hydrogen, carbon, nitrogen, oxygen, phosphorus, and calcium elements.

[0086] In an embodiment of the present disclosure, the CT device for determining the CT image of the substance to be measured and the CT device for determining the sample CT image of the sample material are the same CT device. The target model 303 is trained based on the sample CT images simulated according to the characteristic parameters of the CT device, and the target model 303 also needs to calculate the elemental density distribution based on the CT images scanned by the CT device. Therefore, the target model 303 is associated with the CT device. When the CT device changes, the initial model needs to be retrained to obtain a new target model.

[0087] In an embodiment of the present disclosure, the target model 303 is trained based on the M CT images of each sample material because, in order to ensure the calculation accuracy of the target model 303, the M CT images scanned based on the M scanning energy spectra of the CT device need to be used as the input data 301.

[0088] Figure 4 Schematic diagram of the process of the method for determining the elemental density distribution of a substance according to an embodiment of the present disclosure.

[0089] As Figure 4 shown, the method 400 for determining the elemental density distribution of the substance in this embodiment may include operation S410 to operation S420.

[0090] In operation S410, the substance to be measured is scanned respectively using the M scanning energy spectra of the CT device to obtain M CT images.

[0091] In operation S420, an elemental analysis model is used to determine the elemental density distribution of the substance to be measured based on the M CT images.

[0092] In an embodiment of the present disclosure, the elemental analysis model is a target model trained by using the model training method 100 provided in the embodiment of the present disclosure.

[0093] For example, the HU values included in the M CT images are output to the elemental analysis model, and the elemental analysis model can output the elemental density distribution of the substance to be measured.

[0094] In an embodiment of the present disclosure, through the elemental analysis model, based on the M CT images, the process of determining the elemental density distribution of the substance to be measured is similar to the training process described above. For the sake of brevity, it will not be elaborated here.

[0095] Based on the model training method provided in the present disclosure, the present disclosure also provides a model training device, which will be described in detail below in combination with Figure 5 This device will be described in detail.

[0096] Figure 5 is a structural block diagram of a model training device according to an embodiment of the present disclosure.

[0097] As Figure 5 shown, the model training device 500 of this embodiment may include an acquisition module 510, a first determination module 520, a construction module 530, and a training module 540.

[0098] The acquisition module 510 is configured to acquire the characteristic parameters of each of the M scanning energy spectra of a computed tomography (CT) device. In one embodiment, the acquisition module 510 may be configured to perform the operation S110 described above, which will not be elaborated here.

[0099] The first determination module 520 is configured to determine, based on the M characteristic parameters and the elemental density distributions of the N sample materials, the M sample CT images of each of the N sample materials under the M scanning energy spectra, so as to obtain M×N sample CT images. In one embodiment, the first determination module 520 is configured to perform the operation S120 described above, which will not be elaborated here.

[0100] The construction module 530 is configured to construct a sample data set based on the M sample CT images of each of the N sample materials and their respective elemental density distributions. In one embodiment, the construction module 530 may be configured to perform the operation S130 described above, which will not be elaborated here.

[0101] The training module 540 is configured to train an initial model by using the sample data set to obtain a target model corresponding to the CT device, and the target model is used to determine the elemental density distribution based on the CT images scanned by the CT device. In one embodiment, the training module 540 may be configured to perform the operation S140 described above, which will not be elaborated here.

[0102] According to an embodiment of the present disclosure, the acquisition module 510 acquires characteristic parameters of each of the M scanning energy spectra of a computed tomography (CT) device, including: scanning a reference material using the M scanning energy spectra distributions of the CT device to obtain M scanning data of the reference material; and comparing the M scanning data with reference data of the reference material acquired in advance to obtain M characteristic parameters of the M scanning energy spectra.

[0103] According to an embodiment of the present disclosure, the construction module 530 is configured to construct a sample data set based on M sample CT images and respective element density distributions of N sample materials, including: for each of the N sample materials, performing the following operations: based on the M sample CT images, determining M CT values of multiple sample objects in the sample material; based on the element density distribution, acquiring M element densities corresponding to the M CT values; based on the M CT values and the M element densities, determining sample data of multiple sample objects; and constructing a sample data set of the sample material based on the sample data.

[0104] According to an embodiment of the present disclosure, the construction module 530 is configured to determine sample data of multiple sample objects based on M CT values and element density, including: within a preset range, generating multiple groups of enhanced element densities based on the M element densities, each group of enhanced element densities including M enhanced element densities, and the difference between the M element densities and each group of enhanced element densities being within the preset range; and determining sample data of multiple sample objects based on the M CT values of multiple sample objects and the multiple groups of enhanced element densities.

[0105] According to an embodiment of the present disclosure, the first determination module 520 is configured to determine M sample CT images of N sample materials under the M scanning energy spectra based on the M characteristic parameters and the element density distributions of the N sample materials, to obtain M×N sample CT images, including: using the M characteristic parameters and the element density distributions of the N sample materials to determine the energy absorption capabilities of the N sample materials with respect to the M scanning energy spectra respectively; and based on the energy absorption capabilities, determining M groups of CT values of the N sample materials, and the M groups of CT values form M CT images.

[0106] According to an embodiment of the present disclosure, the training module 540 is configured to train an initial model using the sample data set to obtain a target model corresponding to the CT device, and the target model is configured to determine an element density distribution based on a CT image scanned by the CT device, including: determining the sample CT images in the sample data set as input data of the initial model, and determining the element density distributions in the sample data set as output data of the initial model; using the distribution characteristics of multiple elements in the element density distribution as hint information; and training the initial model based on the hint information, the input data, and the output data to obtain the target model.

[0107] Figure 6 It is a structural block diagram of a device for determining the density distribution of material elements according to an embodiment of the present disclosure.

[0108] As Figure 6 shown, the device 600 for determining the density distribution of material elements in this embodiment may include a scanning module 610 and a second determination module 620.

[0109] The scanning module 610 is configured to scan a material to be measured using M scanning energy spectra of a CT device to obtain M CT images. In one embodiment, the scanning module 610 may be used to perform the operation S410 described above, which will not be elaborated here.

[0110] The second determination module 620 is configured to use an element analysis model to determine the element density distribution of the material to be measured based on the M CT images. In one embodiment, the second determination module 620 is used to perform the operation S420 described above, which will not be elaborated here.

[0111] In the embodiment of the present disclosure, the element analysis model is a target model trained using, for example, the model training device 500 provided in the embodiment of the present disclosure.

[0112] It should be noted that in the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision, disclosure, and application of the user's personal information involved all comply with the provisions of relevant laws and regulations, necessary confidentiality measures are taken, and it does not violate public order and good customs. In the technical solution of the present disclosure, the authorization or consent of the user is obtained before obtaining or collecting the user's personal information.

[0113] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0114] Figure 7 A schematic block diagram of an example electronic device 700 that can be used to implement the method of the embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0115] As Figure 7As shown, the electronic device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the electronic device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0116] Multiple components in the electronic device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a magnetic disk, an optical disc, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the electronic device 700 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0117] The computing unit 701 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 701 include but are not limited to a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 701 executes the various methods and processes described above, such as the method for determining the density distribution of material elements and / or the model training method. For example, in some embodiments, the method for determining the density distribution of material elements and / or the model training method can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as the storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the method for determining the density distribution of material elements and / or the model training method described above can be executed. Alternatively, in other embodiments, the computing unit 701 can be configured to execute the method for determining the density distribution of material elements and / or the model training by any other appropriate means (e.g., by means of firmware).

[0118] The various embodiments of the systems and techniques described above in this specification can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), systems-on-chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which may be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0119] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0120] In the context of the present disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0121] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0122] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0123] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. Among them, the server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services ("Virtual Private Server", or simply "VPS"). The server can also be a server of a distributed system, or a server combined with a blockchain.

[0124] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and no limitation is made herein.

[0125] The above specific embodiments do not constitute a limitation to the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. A model training method, comprising: Obtaining characteristic parameters of M scanning energy spectra of a computed tomography (CT) device, where M is a positive integer; Based on the M characteristic parameters and the element density distribution of the N sample materials, determine M sample CT images of the N sample materials under the M scanning energy spectra to obtain M*N sample CT images, where N is a positive integer; constructing a sample data set based on the M sample CT images of the N sample materials and the element density distribution of each; as well as The sample data set is used to train an initial model to obtain a target model corresponding to the CT device, and the target model is used to determine element density distribution based on a CT image scanned by the CT device.

2. The method according to claim 1, wherein: The step of obtaining characteristic parameters of M scanning energy spectra of a computer tomography (CT) device comprises: Scanning a reference material using the M scanning energy spectrum distributions of the CT device to obtain M scanning data of the reference material; and The M scanning data are respectively compared with the reference data of the reference material acquired in advance to obtain M characteristic parameters of the M scanning energy spectra.

3. The method according to claim 1, wherein: The constructing a sample data set based on the M sample CT images and the element density distributions of the N sample materials respectively includes: For each of the N sample materials, do the following: Based on the M sample CT images, determining M CT values ​​of each of a plurality of sample objects in the sample material; Based on the element density distribution, obtaining M element densities corresponding to the M CT values; Determining sample data of each of the plurality of sample objects based on the M CT values ​​and the M element densities; and Based on the sample data, a sample data set of the sample material is constructed.

4. The method according to claim 3, wherein: The determining, based on the M CT values ​​and the element density, the sample data of each of the plurality of sample objects comprises: generating a plurality of groups of enhancement element densities based on the M element densities within a preset range, each group of enhancement element densities including M enhancement element densities, and a difference between the M element densities and each group of enhancement element densities being within the preset range; and Based on the M CT values ​​of each of the plurality of sample objects and the plurality of sets of enhancement element densities, sample data of each of the plurality of sample objects are determined.

5. The method according to claim 1, wherein: The determining, based on the M characteristic parameters and the element density distribution of the N sample materials, M sample CT images of the N sample materials under the M scanning energy spectra to obtain M*N sample CT images includes: Determining the energy absorption capacity of each of the N sample materials for the M scanning energy spectra by using the M characteristic parameters and the element density distributions of the N sample materials; and Based on the energy absorption capabilities, M groups of CT values ​​are determined for each of the N sample materials, and the M groups of CT values ​​form M CT images.

6. The method according to claim 1, wherein: The using the sample data set to train the initial model to obtain the target model corresponding to the CT device includes: Determining a sample CT image in the sample data set as input data of the initial model, and determining an element density distribution in the sample data set as output data of the initial model; using distribution characteristics of each of the plurality of elements in the element density distribution as prompt information; and Based on the prompt information, the input data and the output data, the initial model is trained to obtain the target model.

7. A method for determining the density distribution of a material element, comprising: The material to be tested is scanned respectively using the M scanning energy spectra of the CT device to obtain M CT images, where M is a positive integer; as well as Determining the element density distribution of the substance to be tested based on the M CT images using an element analysis model; Wherein, the element analysis model is a target model trained using the method described in any one of claims 1-6.

8. A model training device, comprising: An acquisition module, used for acquiring characteristic parameters of M scanning energy spectra of a computed tomography (CT) device, where M is a positive integer; A first determination module is used to determine, based on the M characteristic parameters and the element density distribution of the N sample materials, the M sample CT images of the N sample materials under the M scanning energy spectra to obtain M*N sample CT images, where N is a positive integer; A construction module, configured to construct a sample data set based on the M sample CT images and the element density distributions of the respective N sample materials; as well as A training module is used to train an initial model using the sample data set to obtain a target model corresponding to the CT device, wherein the target model is used to determine element density distribution based on a CT image scanned by the CT device.

9. A device for determining the density distribution of material elements, comprising: A scanning module, used to use the M scanning energy spectra of the CT device to respectively scan the material to be tested to obtain M CT images; as well as A second determination module is used to determine the element density distribution of the substance to be tested based on the M CT images using an element analysis model; Wherein, the element analysis model is a target model trained using the device as described in claim 8.

10. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

Images