Body weight prediction method and device, computer equipment and storage medium

Abstract

The invention relates to a weight prediction method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a medical image of a detection object; determining a density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is a numerical value of a parameter capable of characterizing the density information of the detection object; according to the density characterization parameter value and a predetermined mapping relation between the density characterization parameter value and a density value, determining a mapping weight of the detection object; and through a predetermined correction model, carrying out correction processing on the mapping body weight to obtain a target body weight of the detection object. By adopting the method, the accuracy of the obtained target weight of the detected object can be improved.

Description

Technical Field

[0001] The present application relates to the field of computer technologies, and particularly to a method, apparatus, computer device, storage medium, and computer program product for predicting body weight. Background Art

[0002] Body weight is an important parameter for calculating the standardized uptake value (SUV) in quantitative research of positron emission tomography (PET, a nuclear medicine imaging). When there is a situation of missing or incorrect filling of body weight in the clinical PET data acquisition scenario, it may affect the SUV quantitative calculation of PET images, thereby affecting normal clinical diagnosis and treatment. Therefore, a body weight assessment model is needed to accurately assess the body weight of a human body in whole-body PET / CT (Computed Tomography) data.

[0003] In related technologies, the method adopted is the spinal index regression analysis method based on abdominal CT slice 2D images. According to the significant statistical correlation between somatometric parameters (such as height and weight) and various spinal indices, the corresponding spinal indices are calculated from abdominal CT images, and the method of multiple linear regression is used to predict somatometric parameters such as human height and weight.

[0004] However, the spinal index mainly estimates body weight based on the morphological characteristics of the spine, while the body weight of a human body is mainly composed of fat, muscle, bone, and other tissues. Therefore, relying solely on the spinal index to estimate body weight may not comprehensively consider the composition of the body. Therefore, the error of the body weight predicted by the existing spinal index method is relatively large. Summary of the Invention

[0005] Based on this, in view of the technical problem of relatively large error in the body weight predicted by the above method, it is necessary to provide a method, apparatus, computer device, computer-readable storage medium, and computer program product for predicting body weight.

[0006] In a first aspect, the present application provides a method for predicting body weight. The method includes:

[0007] Obtain a medical image of a detection object;

[0008] Determine a density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is the numerical value of a parameter that can characterize the density information of the detection object;

[0009] Determine a mapped body weight of the detection object according to the density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value;

[0010] The mapped body weight is corrected through a pre-determined correction model to obtain the target body weight of the detection object.

[0011] In one embodiment, the determining of the mapped body weight of the detection object according to the density characterization parameter value and the pre-determined mapping relationship between the density characterization parameter value and the density value includes:

[0012] Determine the density value of each voxel of the detection object according to the density characterization parameter value and the mapping relationship;

[0013] Predict the weight of each voxel according to the density value of each voxel and the size of each voxel;

[0014] Obtain the mapped body weight of the detection object based on the weight of each voxel.

[0015] In one embodiment, the obtaining of the mapped body weight of the detection object based on the weight of each voxel includes:

[0016] Perform a summation process on the weights of each voxel to obtain the mapped body weight of the detection object.

[0017] In one embodiment, the medical image includes a computed tomography image; the determining of the density characterization parameter value of the detection object according to the medical image includes:

[0018] Extract the local image corresponding to the detection object in the medical image;

[0019] Extract the gray value of the local image to obtain the gray value of each voxel in the local image;

[0020] Convert the gray value of each voxel into a computed tomography value, and determine the computed tomography value as the density characterization parameter value of the detection object.

[0021] In one embodiment, the medical image includes a magnetic resonance image; the determining of the density characterization parameter value of the detection object according to the medical image includes:

[0022] Extract the local image corresponding to the detection object in the medical image;

[0023] Obtain the signal intensity value of each voxel in the local image;

[0024] Determine a magnetic resonance value based on the signal intensity value of each voxel, and determine the magnetic resonance value as the density characterization parameter value of the detection object.

[0025] In one embodiment, the mapping relationship between the density characterization parameter value and the density value is determined by the following method:

[0026] Obtain the density characterization parameter values and density values of different tissues of the sample object;

[0027] Perform fitting processing on the density characterization parameter values and density values of the different tissues to obtain the mapping relationship between the density characterization parameter value and the density value.

[0028] In one embodiment, the correction model is determined by the following method:

[0029] Obtain a sample data set; the sample data set includes multiple groups of weight data, and each group of weight data includes a sample true weight and a corresponding sample mapped weight;

[0030] Train an initial correction model according to the sample true weight and the sample mapped weight in each group of weight data in the sample data set to obtain the correction model.

[0031] In a second aspect, the present application further provides a weight prediction device. The device includes:

[0032] An image acquisition module, configured to acquire a medical image of a detection object;

[0033] A parameter value determination module, configured to determine the density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is the numerical value of a parameter that can characterize the density information of the detection object;

[0034] A weight determination module, configured to determine the mapped weight of the detection object according to the density characterization parameter value and the mapping relationship between the density characterization parameter value and the density value determined in advance;

[0035] A weight correction module, configured to perform correction processing on the mapped weight through a correction model determined in advance to obtain the target weight of the detection object.

[0036] In one embodiment, the weight determination module is further configured to determine the density value of each voxel of the detection object according to the density characterization parameter value and the mapping relationship; predict the weight of each voxel according to the density value of each voxel and the size of each voxel; and obtain the mapped weight of the detection object based on the weight of each voxel.

[0037] In one embodiment, the weight determination module is further configured to sum the weights of each voxel to obtain the mapped weight of the detection object.

[0038] In one embodiment, the medical image includes a computed tomography (CT) image; the parameter value determination module is further configured to extract a local image corresponding to the detection object from the medical image; extract the gray value of the local image to obtain the gray value of each voxel in the local image; convert the gray value of each voxel into a computed tomography value, and determine the computed tomography value as the density characterization parameter value of the detection object.

[0039] In one embodiment, the medical image includes a magnetic resonance (MR) image; the parameter value determination module is further configured to extract a local image corresponding to the detection object from the medical image; obtain the signal intensity value of each voxel in the local image; based on the signal intensity value of each voxel, determine a magnetic resonance value, and determine the magnetic resonance value as the density characterization parameter value of the detection object.

[0040] In one embodiment, the device further includes a mapping relationship determination module, configured to obtain the density characterization parameter values and density values of different tissues of a sample object; perform a fitting process on the density characterization parameter values and density values of different tissues to obtain the mapping relationship between the density characterization parameter values and density values.

[0041] In one embodiment, the device further includes a model determination module, configured to obtain a sample data set; the sample data set includes multiple groups of body weight data, and each group of body weight data includes a sample true body weight and a corresponding sample mapped body weight; train an initial correction model according to the sample true body weight and the sample mapped body weight in each group of body weight data in the sample data set to obtain the correction model.

[0042] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0043] Obtain a medical image of a detection object;

[0044] Determine the density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is the numerical value of a parameter that can characterize the density information of the detection object;

[0045] Determine the mapped body weight of the detection object according to the density characterization parameter value and the mapping relationship between the density characterization parameter value and the density value determined in advance;

[0046] Perform a correction process on the mapped body weight through a correction model determined in advance to obtain the target body weight of the detection object.

[0047] Fourth aspect, the present application also provides a computer-readable storage medium. On the computer-readable storage medium, a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0048] Obtain a medical image of a detection object;

[0049] According to the medical image, determine a density characterization parameter value of the detection object; the density characterization parameter value is a numerical value of a parameter that can characterize the density information of the detection object;

[0050] According to the density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value, determine a mapped body weight of the detection object;

[0051] Through a pre-determined correction model, perform a correction process on the mapped body weight to obtain a target body weight of the detection object.

[0052] Fifth aspect, the present application also provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0053] Obtain a medical image of a detection object;

[0054] According to the medical image, determine a density characterization parameter value of the detection object; the density characterization parameter value is a numerical value of a parameter that can characterize the density information of the detection object;

[0055] According to the density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value, determine a mapped body weight of the detection object;

[0056] Through a pre-determined correction model, perform a correction process on the mapped body weight to obtain a target body weight of the detection object.

[0057] The above weight prediction method, device, computer device, storage medium, and computer program product determine the density characterization parameter value of a detection object by detecting the medical image of the object. After determining the mapped weight of the detection object according to the density characterization parameter value and the pre-determined mapping relationship between the density characterization parameter value and the density value, the mapped weight is corrected through a pre-determined correction model to obtain the target weight of the detection object. This method first determines the density characterization parameter value based on the medical image of the detection object, and determines the mapped weight of the detection object through the density characterization parameter value that can represent the density information of the detection object, realizing weight determination based on density and improving the accuracy of the determined mapped weight. Moreover, after determining the mapped weight, the mapped weight is further corrected through a pre-determined correction model, further improving the accuracy of the target weight of the detection object obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a schematic flowchart of the weight prediction method in an embodiment;

[0059] Figure 2 is a schematic flowchart of the steps for determining the mapped weight of a detection object in an embodiment;

[0060] Figure 3 is a schematic flowchart of the steps for determining the mapping relationship between the density characterization parameter value and the density value in an embodiment;

[0061] Figure 4 is a schematic flowchart of the weight prediction method in another embodiment;

[0062] Figure 5 is a schematic diagram including predicted weight and actual weight data in an embodiment;

[0063] Figure 6 is a structural block diagram of the weight prediction device in an embodiment;

[0064] Figure 7 is an internal structure diagram of a computer device in an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] In order to make the objectives, technical solutions, and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0066] In one embodiment, as Figure 1As shown, a body weight prediction method is provided. In this embodiment, the method is exemplified by being applied to a terminal. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. Among them, the terminal can be, but is not limited to, various personal computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server can be implemented by an independent server or a server cluster composed of multiple servers. In this embodiment, the method includes the following steps:

[0067] Step S110, obtain the medical image of the detection object.

[0068] Among them, the detection object refers to the object for which the body weight needs to be determined.

[0069] Among them, the medical image is an image containing three-dimensional volume data information of the detection object. For example, the medical image can be a Computed Tomography (CT) image, a Magnetic Resonance (MR) image, an attenuation coefficient image, an infrared image, a CT scout image, etc. Among them, the attenuation coefficient map is an image used to describe the absorption capacity differences of different tissues and structures in a medical image. It reflects the attenuation degree of X-rays or magnetic fields when passing through different tissues and structures in imaging technologies such as X-rays, CT, and MRI. Different tissues and structures have different absorption capacities for X-rays or magnetic fields, so they show different gray values or signal intensities in the image. Among them, the CT scout map refers to the positioning image in Computed Tomography (CT scan), which is usually used to determine the scanning position and range of the detection object. The CT scout map can also be called a positioning image or a positioning scan.

[0070] In specific implementation, different types of medical images are obtained by different methods. Specifically, the medical image of the detection object can be collected through a medical device, or the medical image collected by the medical device can be processed to obtain the medical image of the detection object.

[0071] For example, for a Computed Tomography (CT) image, it can be collected by a Computed Tomography (CT) device. For a Magnetic Resonance image, it can be collected by a Magnetic Resonance device. For a CT scout map, it can be reconstructed based on the projection data of X-rays passing through the detection object at different angles. For the attenuation coefficient map, after collecting the CT image, based on the gray values of the CT image, the attenuation coefficients of different tissues and structures are calculated; the attenuation coefficient map is generated according to the attenuation coefficients.

[0072] Step S120: Determine the density characterization parameter value of the detection object according to the medical image.

[0073] Among them, the density characterization parameter value is the numerical value of the parameter that can characterize the density information of the detection object. For example, the density characterization parameter value can be the HU (Hounsfield Unit) value, also known as the CT value, and the HU value reflects the degree of X-ray absorption of the tissue. Similarly, the density characterization parameter value can also be the magnetic resonance value (MR value), attenuation coefficient value, etc.

[0074] Among them, each voxel in the medical image has a corresponding density characterization parameter value.

[0075] In specific implementation, after obtaining the medical image of the detection object, the density characterization parameter value of the detection object can be determined based on this medical image. Specifically, it is to determine the density characterization parameter value corresponding to each voxel in the medical image.

[0076] It can be understood that for different types of medical images, the types of corresponding density characterization parameter values will also be different. Correspondingly, the methods for determining the density characterization parameter values will also be different. That is, there is a corresponding relationship between the type of medical image and the type of density characterization parameter value.

[0077] For example, if the medical image is a CT image, the determined density characterization parameter value corresponds to the HU value. If the medical image is an MR image, the determined density characterization parameter value corresponds to the MR value. If the medical image is an attenuation coefficient map, the determined density characterization parameter value corresponds to the attenuation coefficient value.

[0078] Step S130: Determine the mapped body weight of the detection object according to the density characterization parameter value and the pre-determined mapping relationship between the density characterization parameter value and the density value.

[0079] In specific implementation, since the detection object is represented in the medical image by a number of voxels, therefore, the body weight of the detection object needs to be determined according to the voxel value of each voxel in the medical image. Specifically, the density value of each voxel of the detection object can be obtained by querying the pre-determined mapping relationship between the density characterization parameter value and the density value according to the density characterization parameter value. Based on the density value of each voxel of the detection object, the mapped body weight of the detection object is determined.

[0080] Step S140: Perform a correction process on the mapped body weight through a pre-determined correction model to obtain the target body weight of the detection object.

[0081] Among them, the correction model can be determined based on the sample mapped body weight and the sample true body weight. The correction model can be understood as a model that characterizes the relationship between the mapped body weight and the true body weight, and is used to convert the mapped body weight into the true body weight.

[0082] In a specific implementation, multiple groups of paired sample true weights and sample mapped weights can be obtained in advance, and a correction model can be determined based on the paired sample true weights and sample mapped weights of each group.

[0083] After determining the mapped weight of the detection object, the mapped weight is input into the correction model, and is converted into the true weight by the correction model for output. The weight value output by the correction model is used as the corrected weight of the detection object, that is, the target weight.

[0084] In the above weight prediction method, the density characterization parameter value of the detection object is determined through the medical image of the detection object. And after determining the mapped weight of the detection object according to the density characterization parameter value and the pre-determined mapping relationship between the density characterization parameter value and the density value, the mapped weight is corrected through the pre-determined correction model to obtain the target weight of the detection object. This method first determines the density characterization parameter value based on the medical image of the detection object, and determines the mapped weight of the detection object through the density characterization parameter value that can represent the density information of the detection object, realizing density-based weight determination, which can improve the accuracy of the determined mapped weight. And, after determining the mapped weight, the mapped weight is further corrected through the pre-determined correction model, further improving the accuracy of the target weight of the detection object obtained.

[0085] In an exemplary embodiment, as Figure 2 shown, in the above step S130, determining the mapped weight of the detection object according to the density characterization parameter value and the pre-determined mapping relationship between the density characterization parameter value and the density value includes:

[0086] Step S131, determining the density value of each voxel of the detection object according to the density characterization parameter value and the mapping relationship.

[0087] Specifically, the area occupied by the detection object can be first determined from the medical image as a local image. The mapped weight of the detection object is determined based on the density characterization parameter values of each voxel in this local image.

[0088] More specifically, the density characterization parameter values of each voxel in the local image occupied by the detection object can be obtained, and the pre-determined mapping relationship between the density characterization parameter value and the density value is queried to obtain the density value corresponding to the density characterization parameter value of each voxel in the local image, as the density value of each voxel of the detection object.

[0089] Step S132, predicting the weight of each voxel according to the density value of each voxel and the size of each voxel.

[0090] Wherein, the size of the voxel is specifically the volume of the voxel.

[0091] In a specific implementation, each voxel has a corresponding volume. After determining the density value of each voxel, the weight of each voxel can be predicted based on the density value and volume of each voxel.

[0092] More specifically, the density value and volume of each voxel can be multiplied to obtain the weight of each voxel. It can be expressed by the formula: mass (g) = density (g / cc) * volume (cc).

[0093] Step S133: Obtain the mapped weight of the detection object based on the weight of each voxel.

[0094] Specifically, after determining the weights of the respective voxels that make up the detection object, the weight of the detection object can be further obtained based on the weights of the respective voxels as the mapped weight of the detection object.

[0095] More specifically, in one embodiment, obtaining the mapped weight of the detection object based on the weight of each voxel includes: performing a summation process on the weights of each voxel to obtain the mapped weight of the detection object. That is, the sum obtained by adding the weights of the respective voxels of the detection object is used as the mapped weight of the detection object.

[0096] In this embodiment, first, the density values of the respective voxels of the detection object are determined, combined with the sizes of the respective voxels to obtain the weights of the respective voxels. Further, based on the weights of the respective voxels, the mapped weight of the detection object is determined. This method takes into account the weights of the respective voxels that make up the detection object, and calculates the weight of the detection object accurately to the voxel level, thereby improving the accuracy of the determined mapped weight.

[0097] In an exemplary embodiment, when the medical image of the detection object is a computed tomography image (CT image), correspondingly, in the above step S120, according to the medical image, determining the density characterization parameter value of the detection object includes: extracting the local image corresponding to the detection object in the medical image; performing gray value extraction on the local image to obtain the gray value of each voxel in the local image; converting the gray value of each voxel into a computed tomography value, and determining the computed tomography value as the density characterization parameter value of the detection object.

[0098] In a specific implementation, the weight of the detection object is only related to the voxels of the detection object and has nothing to do with other background regions in the medical image. Therefore, when determining the density characterization parameter value of the detection object according to the medical image, the region occupied by the detection object can be first extracted from the medical image as the local image corresponding to the detection object. For example, the region occupied by the detection object is extracted through an object detection algorithm. Further, the density characterization parameter values of the respective voxels in the local image are determined as the density characterization parameter values of the detection object.

[0099] More specifically, when the medical image is a computed tomography (CT) image, determining the density characterization parameter values of each voxel in the local image means determining the Hounsfield unit (HU) value of each voxel, or the CT value. Specifically, the gray values of each voxel in the local image can be extracted first, and then the gray values are further converted into HU values to obtain the HU values of each voxel, which are used as the density characterization parameter values of the detection object.

[0100] In this embodiment, by extracting the local image corresponding to the detection object in the medical image and only calculating the computed tomography values of each voxel in the local image corresponding to the detection object as the density characterization parameter values of the detection object, it is possible to avoid wasting computing resources caused by calculating the computed tomography values of irrelevant background regions in the medical image. At the same time, by calculating the computed tomography value of each voxel of the detection object, it is convenient to achieve voxel-level body weight prediction, which can improve the accuracy of the predicted body weight.

[0101] In an exemplary embodiment, when the medical image of the detection object is a magnetic resonance (MR) image, correspondingly, in the above step S120, determining the density characterization parameter values of the detection object according to the medical image includes: extracting the local image corresponding to the detection object in the medical image; obtaining the signal intensity value of each voxel in the local image; based on the signal intensity value of each voxel, determining the magnetic resonance value, and determining the magnetic resonance value as the density characterization parameter value of the detection object.

[0102] In specific implementation, similar to the method of determining the density characterization parameter values of the detection object when the medical image is a computed tomography image, when the medical image of the detection object is a magnetic resonance (MR) image, it is also necessary to first extract the area occupied by the detection object from the medical image as the local image corresponding to the detection object. Then, the magnetic resonance value of each voxel in the local image is determined as the density characterization parameter value of the detection object.

[0103] More specifically, when the medical image is a magnetic resonance image, determining the density characterization parameter values of each voxel in the local image means determining the magnetic resonance (MR) value of each voxel. Specifically, the signal intensity values of each voxel in the local image can be extracted first, and then the signal intensity values are further converted into magnetic resonance values to obtain the magnetic resonance values of each voxel, which are used as the density characterization parameter values of the detection object.

[0104] In this embodiment, by extracting the local image corresponding to the detection object in the medical image and only calculating the magnetic resonance values of each voxel in the local image corresponding to the detection object as the density characterization parameter values of the detection object, it is possible to avoid wasting computing resources caused by calculating the magnetic resonance values of irrelevant background regions in the medical image. At the same time, by calculating the magnetic resonance value of each voxel of the detection object, it is convenient to achieve voxel-level body weight prediction, which can improve the accuracy of the predicted body weight.

[0105] In an exemplary embodiment, as Figure 3 shown, in the above step S130, the mapping relationship between the density characterization parameter value and the density value is determined by the following method:

[0106] Step S310, obtain the density characterization parameter values and density values of different tissues of the sample object;

[0107] Step S320, perform fitting processing on the density characterization parameter values and density values of different tissues to obtain the mapping relationship between the density characterization parameter value and the density value.

[0108] Among them, the sample object and the detection object belong to the same type of object.

[0109] It can be understood that objects with different densities will correspond to different density characterization parameter values (for example, HU values). In human tissues, different voxels will also correspond to different HU values, and therefore, their corresponding densities are different. For example, for lung tissue, its density is close to that of water, but due to the presence of air in the lungs, the average density of the entire lungs is relatively low. Therefore, different parts inside the lungs will correspond to different HU values.

[0110] In a specific implementation, the density characterization parameter values and density values of different tissues of different sample objects can be obtained in advance. Taking the density characterization parameter value of the voxel as the input feature and the density value of the voxel as the output feature, a mapping function model is constructed. Through this mapping function model, fitting processing is performed on the density characterization parameter values and density values of different tissues to obtain the mapping relationship between the density characterization parameter value and the density value.

[0111] More specifically, since there is a bilinear relationship between the density characterization parameter value and the density value, a bilinear function can be constructed as the mapping function model. It can be understood that in practical applications, in addition to the bilinear function, other function types such as linear functions and piecewise functions can also be used to construct the mapping function model.

[0112] In this embodiment, by performing fitting processing on the density characterization parameter values and density values of different tissues of the sample object, the mapping relationship between the density characterization parameter value and the density value is obtained, so as to facilitate the subsequent determination of the density values of different tissues of the detection object according to this mapping relationship.

[0113] In an exemplary embodiment, the correction model in the above step S140 is determined by the following method: obtain a sample data set; the sample data set includes multiple groups of body weight data, and each group of body weight data includes a sample true body weight and the corresponding sample mapped body weight; according to the sample true body weight and the sample mapped body weight in each group of body weight data in the sample data set, train the initial correction model to obtain the correction model.

[0114] In a specific implementation, after determining the mapping relationship between the density characterization parameter value and the density value, for the medical image of any object, the mapped weight of the object can be determined. Therefore, a batch of sample medical images of sample objects and the sample true weights of the sample objects can be obtained in advance. Based on the sample medical image of each sample object and the determined mapping relationship between the density characterization parameter value and the density value, the mapped weight for each sample object is determined as the sample mapped weight. Thus, the sample mapped weight of each sample object is obtained. Further, the sample mapped weight and the sample true weight of each sample object are used as a set of weight data, and thus multiple sets of weight data are obtained, constituting a sample data set.

[0115] After determining multiple sets of weight data including the sample mapped weight and the sample true weight, a correction function can be constructed. Specifically, a correction function with the mapped weight as the input feature and the true weight as the output feature can be constructed as an initial correction model.

[0116] Further, a regression analysis is performed on the sample mapped weight and the sample true weight through the correction function to obtain a correction model. Specifically, the coefficient of the correction function is determined through regression analysis, and this coefficient is assigned to the correction function to obtain the correction model. More specifically, when determining the coefficient of the correction function through regression analysis, methods such as the least squares method can be used to fit an optimal coefficient to minimize the sum of the squared residuals between the predicted true weight of the correction model and the sample true weight.

[0117] It should be noted that the regression analysis method can adopt regression analysis methods such as linear regression, logistic regression, and least absolute shrinkage and selection operator regression (Lasso). The correction function can adopt function types such as linear functions, polynomial functions, and exponential functions. This application does not make specific limitations in this regard.

[0118] In this embodiment, through the sample true weight and the sample mapped weight, a correction model representing the relationship between the true weight and the mapped weight is trained, so that the correction model can be used to correct the mapped weight determined based on the medical image of the detection object, improving the accuracy of the determined target weight.

[0119] In one embodiment, to facilitate those skilled in the art to understand the embodiments of this application, specific examples in conjunction with the drawings will be described below. Refer to Figure 4 , which shows a specific flowchart of a weight prediction method. In this embodiment, the method includes the following steps:

[0120] Step S401, obtain the medical image of the detection object;

[0121] Step S402: Extract the local image corresponding to the detection object from the medical image;

[0122] Step S403: Determine the density characterization parameter value of the detection object according to the local image;

[0123] Step S403a: When the medical image is a computed tomography (CT) image, extract the gray value of the local image to obtain the gray value of each voxel in the local image; convert the gray value of each voxel into a CT value, and determine the CT value as the density characterization parameter value of the detection object;

[0124] Step S403b: When the medical image is a magnetic resonance (MR) image, obtain the signal intensity value of each voxel in the local image; based on the signal intensity value of each voxel, determine the MR value, and determine the MR value as the density characterization parameter value of the detection object;

[0125] Step S404: Determine the density value of each voxel of the detection object according to the density characterization parameter value and the pre-determined mapping relationship between the density characterization parameter value and the density value;

[0126] Step S405: Predict the weight of each voxel according to the density value of each voxel and the size of each voxel;

[0127] Step S406: Sum up the weights of each voxel to obtain the mapped body weight of the detection object;

[0128] Step S407: Perform a correction process on the mapped body weight through a pre-determined correction model to obtain the target body weight of the detection object.

[0129] The body weight prediction method proposed in this embodiment first determines the density characterization parameter value based on the medical image of the detection object, and then determines the density value of each voxel of the detection object through the density characterization parameter value that can represent the density information of the detection object. Combining the sizes of each voxel, the weights of each voxel are obtained. Further, according to the weights of each voxel, the mapped body weight of the detection object is determined. This method takes into account the weights of each voxel that makes up the detection object, and calculates the body weight of the detection object to the voxel level, thereby improving the accuracy of the determined mapped body weight. And, after determining the mapped body weight, further correct the mapped body weight through a pre-determined correction model, which further improves the accuracy of the obtained target body weight of the detection object.

[0130] Moreover, the principle of this method is simple, with high evaluation accuracy and strong interpretability. It can be used for accurate weight assessment of various special cases. For example, patients with abnormal Body Mass Index (BMI), amputees, metal artifacts, and infant data. And through experiments, it is found that the prediction error of the weight prediction model provided by this application is less than 1 kg. For example, in Figure 5 in the example shown in (a) of Figure 5 , the error between the predicted weight and the actual weight is: 69 - 68.95 = 0.05 kg. In

[0131] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the indications of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0132] Based on the same inventive concept, the embodiments of this application also provide a weight prediction device for implementing the above-mentioned weight prediction method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the weight prediction device provided below can refer to the limitations on the weight prediction method in the above text, and will not be repeated here.

[0133] In one embodiment, as Figure 6 shown, a weight prediction device is provided, including: an image acquisition module 610, a parameter value determination module 620, a weight determination module 630, and a weight correction module 640, where:

[0134] The image acquisition module 610 is configured to acquire a medical image of a detection object;

[0135] The parameter value determination module 620 is configured to determine a density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is the numerical value of a parameter that can characterize the density information of the detection object;

[0136] A weight determination module 630, configured to determine a mapped weight of a detection object according to a density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value;

[0137] A weight correction module 640, configured to perform a correction process on the mapped weight through a pre-determined correction model to obtain a target weight of the detection object.

[0138] In one embodiment, the weight determination module 630 is further configured to determine the density value of each voxel of the detection object according to the density characterization parameter value and the mapping relationship; predict the weight of each voxel according to the density value of each voxel and the size of each voxel; and obtain the mapped weight of the detection object based on the weight of each voxel.

[0139] In one embodiment, the weight determination module 630 is further configured to perform a summation process on the weights of each voxel to obtain the mapped weight of the detection object.

[0140] In one embodiment, the medical image includes a computed tomography image; the parameter value determination module 620 is further configured to extract a local image corresponding to the detection object in the medical image; extract the gray value of the local image to obtain the gray value of each voxel in the local image; convert the gray value of each voxel into a computed tomography value, and determine the computed tomography value as the density characterization parameter value of the detection object.

[0141] In one embodiment, the medical image includes a magnetic resonance image; the parameter value determination module 620 is further configured to extract a local image corresponding to the detection object in the medical image; obtain the signal intensity value of each voxel in the local image; determine a magnetic resonance value based on the signal intensity value of each voxel, and determine the magnetic resonance value as the density characterization parameter value of the detection object.

[0142] In one embodiment, the device further includes a mapping relationship determination module, configured to obtain the density characterization parameter values and density values of different tissues of a sample object; perform a fitting process on the density characterization parameter values and density values of different tissues to obtain a mapping relationship between the density characterization parameter value and the density value.

[0143] In one embodiment, the device further includes a model determination module, configured to obtain a sample data set; the sample data set includes multiple groups of weight data, and each group of weight data includes a sample true weight and a corresponding sample mapped weight; train an initial correction model according to the sample true weights and sample mapped weights in each group of weight data in the sample data set to obtain a correction model.

[0144] Each module in the above weight prediction device can be implemented in whole or in part by software, hardware, or a combination thereof. Each of the above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or stored in a memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0145] In one embodiment, a computer device is provided. The computer device can be a terminal, and its internal structural diagram can be as Figure 7 shown. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a weight prediction method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0146] Those skilled in the art can understand that Figure 7 the structure shown in

[0147] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0148] In one embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0149] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0150] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0151] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in this application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., and are not limited thereto. The processors involved in the embodiments provided in this application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., and are not limited thereto.

[0152] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0153] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for predicting body weight, characterized in that, the method includes: obtaining a medical image of a detection object; determining a density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is the numerical value of a parameter that can characterize the density information of the detection object; determining a mapped body weight of the detection object according to the density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value; performing a correction process on the mapped body weight through a pre-determined correction model to obtain the target body weight of the detection object.

2. The method according to claim 1, characterized in that, the determining the mapped body weight of the detection object according to the density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value includes: determining the density value of each voxel of the detection object according to the density characterization parameter value and the mapping relationship; predicting the weight of each voxel according to the density value of each voxel and the size of each voxel; obtaining the mapped body weight of the detection object based on the weight of each voxel.

3. The method according to claim 2, characterized in that, the obtaining the mapped body weight of the detection object based on the weight of each voxel includes: performing a summation process on the weight of each voxel to obtain the mapped body weight of the detection object.

4. The method according to claim 1, characterized in that, the medical image includes a computed tomography image; the determining the density characterization parameter value of the detection object according to the medical image includes: extracting a local image corresponding to the detection object in the medical image; extracting the gray value of each voxel in the local image to obtain the gray value of each voxel in the local image; converting the gray value of each voxel into a computed tomography value and determining the computed tomography value as the density characterization parameter value of the detection object.

5. The method according to claim 1, characterized in that, the medical image includes a magnetic resonance image; the determining the density characterization parameter value of the detection object according to the medical image includes: extracting a local image corresponding to the detection object in the medical image; obtaining the signal intensity value of each voxel in the local image; determining a magnetic resonance value based on the signal intensity value of each voxel and determining the magnetic resonance value as the density characterization parameter value of the detection object.

6. The method according to claim 1, characterized in that, the mapping relationship between the density characterization parameter value and the density value is determined by the following method: obtaining the density characterization parameter value and the density value of different tissues of a sample object; performing a fitting process on the density characterization parameter value and the density value of different tissues to obtain the mapping relationship between the density characterization parameter value and the density value.

7. The method according to claim 1, characterized in that, the correction model is determined by the following method: obtaining a sample data set; the sample data set includes multiple groups of body weight data, and each group of body weight data includes a sample true body weight and a corresponding sample mapped body weight; The initial correction model is trained according to the sample true weight and the sample mapped weight in each group of weight data in the sample dataset to obtain the correction model.

8. A body weight prediction device, characterized in that the device includes: an image acquisition module configured to acquire a medical image of a detection object; a parameter value determination module configured to determine a density characterization parameter value of the detection object according to the medical image; the density characterization parameter value is a numerical value of a parameter that can characterize the density information of the detection object; a body weight determination module configured to determine a mapped body weight of the detection object according to the density characterization parameter value and a pre-determined mapping relationship between the density characterization parameter value and the density value; a body weight correction module configured to perform a correction process on the mapped body weight through a pre-determined correction model to obtain a target body weight of the detection object.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that when the processor executes the computer program, the steps of the body weight prediction method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the body weight prediction method according to any one of claims 1 to 7 are implemented.

Images