Object sketching method, apparatus, and electronic device

Abstract

The present disclosure provides an object delineation method, comprising: obtaining target image data, the target image data at least representing an image generated based on a target ray on a target object, the target object comprising a plurality of sub-objects; based on the target image data, generating first information and second information, the first information representing target material distribution information in the target object, and the second information representing boundary information of each sub-object in the target object; and delineating an image of at least one sub-object according to the first information and the second information.

Description

Technical Field

[0001] This disclosure relates to the medical field and the field of deep learning, and more specifically, to an object delineation device and electronic device. Background Technology

[0002] Existing medical imaging technologies are widely used in clinical diagnosis and treatment, especially in the field of tumor radiotherapy, where accurate delineation of human organs and lesion areas is crucial for treatment planning. Currently, manual delineation remains the mainstream method; however, this process is time-consuming and labor-intensive, and highly dependent on the physician's experience, leading to significant differences in the subjectivity and consistency of the delineation results. Therefore, deep learning-based automatic delineation technology has gradually become a research hotspot, capable of rapidly processing massive amounts of image data, thereby improving efficiency and accuracy. Nevertheless, existing technologies still exhibit significant limitations when faced with blurred organ boundaries, complex lesion morphologies, and poor image quality, failing to fully meet clinical needs. In particular, automatic segmentation technology based on multimodal medical imaging data such as CT and MRI typically requires MRI-to-CT image registration, and the errors introduced during registration significantly affect segmentation accuracy. However, due to the large errors introduced during CT and MRI image registration, existing methods still lack sufficient segmentation accuracy and robustness when dealing with blurred organ boundaries, complex lesion morphologies, and image quality affected by noise. Summary of the Invention

[0003] In view of this, the present disclosure provides an object drawing method and an electronic device.

[0004] One aspect of this disclosure provides an object delineation method, comprising: acquiring target image data, the target image data at least characterizing an image generated on a target object based on a target ray, the target object comprising multiple sub-objects; generating first information and second information based on the target image data, the first information characterizing target material distribution information in the target object, the second information characterizing boundary information of each sub-object in the target object; and delineating an image of at least one sub-object based on the first information and the second information.

[0005] According to embodiments of this disclosure, the target substance distribution information includes the distribution state information of the target substance and / or the density information of the target substance.

[0006] According to embodiments of this disclosure, the target substance includes at least one of carbon and oxygen.

[0007] According to embodiments of this disclosure, generating first information includes: determining at least one test feature parameter of a target object based on target image data; determining at least one test material type of the target object based on the test feature parameter; determining material information of at least one target substance in the target object based on the material distribution characteristics corresponding to the test material type and the various test feature parameters of the target object; and generating first information based on the material information corresponding to at least one substance.

[0008] According to embodiments of this disclosure, drawing an image of at least one sub-object based on first information and second information includes: fusing the first information and second information to obtain target information; inputting the target information into a target model to generate an image of at least one sub-object of the target.

[0009] According to embodiments of this disclosure, the target model includes an upsampling module, a downsampling module, and an edge linear attention module; the upsampling module is used to extract deep features from the target information; the downsampling module is used to extract semantic information from the deep features and generate images of sub-objects; the edge linear attention module is used to generate boundary information based on the feature maps generated by the upsampling module and the downsampling module and transmit it to the upsampling module, so that the upsampling module generates images of sub-objects based on the boundary information.

[0010] According to embodiments of this disclosure, the training process of the target model includes: acquiring sample first information and sample second information, wherein the sample first information represents the sample material distribution state information in the sample object, and the sample second information represents the sample boundary information of each sub-object in the target object; and training the target model based on the sample first information, the sample second information, and the target loss function.

[0011] According to embodiments of this disclosure, the target loss function includes at least a first loss function, which is used to calculate the error between the image boundary of the sub-object and the sample second information in the output of the model.

[0012] Another aspect of this disclosure provides an object delineation apparatus, comprising: a first acquisition module for acquiring target image data, the target image data at least characterizing an image generated on a target object based on a target ray, the target object comprising multiple sub-objects; a first generation module for generating first information and second information based on the target image data, the first information characterizing target material distribution information in the target object, the second information characterizing boundary information of each sub-object in the target object; and a delineation module for delineating an image of at least one sub-object based on the first information and the second information.

[0013] Another aspect of this disclosure provides an electronic device including: at least one processor; and a memory connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the object drawing method of any of the foregoing embodiments.

[0014] Another aspect of this disclosure provides a computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform an object drawing method according to any of the foregoing embodiments.

[0015] Another aspect of this disclosure provides a computer program product, including a computer program / instructions, characterized in that the computer program / instructions, when executed by a processor, implement the operation of the object drawing method of any of the foregoing embodiments.

[0016] According to embodiments of this disclosure, the object delineation method provided by this disclosure has at least one of the following beneficial effects: by combining the distribution information of the target substance with the boundary information of sub-objects, and automatically delineating sub-objects in the target image based on the fusion of different types of features, the accuracy and robustness of the delineation can be significantly improved. Specifically, the distribution information of the target substance can reflect the physical or chemical composition of different tissues, helping to distinguish tissue regions with similar morphology but different compositions; the boundary information enhances the clarity of the sub-object outline, especially when tissue boundaries are blurred or the image signal-to-noise ratio is low, it still has strong recognition capabilities. By fusing these two types of information and inputting them into the target model, accurate segmentation of complex structures can be achieved, effectively overcoming the problems of misjudgment and missed detection caused by single features in existing methods. Attached Figure Description

[0017] The above and other objects, features and advantages of this disclosure will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0018] Figure 1 A flowchart illustrating an object delineation method according to an embodiment of the present disclosure is shown schematically.

[0019] Figure 2 A flowchart illustrating the generation of first information in an object delineation method according to an embodiment of the present disclosure is shown schematically.

[0020] Figure 3 A flowchart illustrating the drawing of a sub-object in an object drawing method according to an embodiment of the present disclosure is shown schematically.

[0021] Figure 4 This schematically illustrates a structural diagram of a target model in an object delineation method according to an embodiment of the present disclosure;

[0022] Figure 5The diagram illustrates a flowchart of the training process for a target model in an object delineation method according to an embodiment of the present disclosure.

[0023] Figure 6 A block diagram schematically illustrates an object delineation apparatus according to an embodiment of the present disclosure; and

[0024] Figure 7 A block diagram of an electronic device suitable for implementing the methods described above, according to embodiments of the present disclosure, is illustrated schematically. Detailed Implementation

[0025] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0026] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0027] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0028] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0029] In the embodiments disclosed herein, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of data (e.g., including but not limited to user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0030] Embodiments of this disclosure provide an object delineation method, comprising: acquiring target image data, the target image data at least characterizing an image generated on a target object based on a target ray, the target object comprising multiple sub-objects; generating first information and second information based on the target image data, the first information characterizing target material distribution information in the target object, the second information characterizing boundary information of each sub-object in the target object; and delineating an image of at least one sub-object according to the first information and the second information.

[0031] Figure 1 A flowchart illustrating an object delineation method according to an embodiment of the present disclosure is shown schematically.

[0032] like Figure 1 As shown, the object drawing method can include at least operations S110~S130.

[0033] In operation S110, target image data is acquired. The target image data at least represents the image generated based on the target ray on the target object, and the target object contains multiple sub-objects.

[0034] Target image data can be image data acquired through medical imaging equipment (such as CT scanners, MRI equipment, etc.). The target rays include, but are not limited to, X-rays, gamma rays, or other forms of electromagnetic waves, which can be used to visualize the internal structure of the target object. The target object is typically the human body or other objects with a certain tissue structure.

[0035] The target object comprises multiple sub-objects, which can be different organs inside the human body (such as the heart, liver, lungs, etc.) or other specific parts of an object that need to be analyzed. Each sub-object has different physical and chemical properties, and their morphology and location typically exhibit different characteristics in the image.

[0036] Specifically, target image data can be generated by acquiring ray data from multiple projection angles and reconstructing it into three-dimensional image data, or by further analyzing images composed of stacked two-dimensional images.

[0037] For example, the target image data could be image data from a CT scan, where each pixel value represents density information at a specific location, or image data from an MRI scan, where each pixel value represents water molecule density information in different tissues.

[0038] In operation S120, based on the target image data, first information and second information are generated. The first information represents the distribution information of target substances in the target object, and the second information represents the boundary information of each sub-object in the target object.

[0039] Specifically, the first information is generated through elemental decomposition or other material analysis methods on the target image data. These methods can extract the distribution of different tissues or substances. This first information can reflect the density and compositional distribution of various substances in the target object, and even include physical or chemical elemental information of different tissues or substances. For example, through elemental decomposition technology, information related to different material components can be separated from CT image data, thereby obtaining more accurate material distribution characteristics of various organs.

[0040] The second piece of information can be extracted through boundary detection algorithms, which can clearly define the outer contours of each sub-object within the target object. Especially in areas with blurred boundaries or poor image quality, boundary detection technology can still provide effective boundary information. Common boundary detection methods include the Sobel operator and Canny edge detection, which can detect the edges of the target object based on image gradients or pixel value changes.

[0041] In operation S130, based on the first information and the second information, an image of at least one sub-object is drawn. Specifically, based on the generated first and second information, image segmentation technology is used to extract the image region of the target sub-object by fusing material distribution state information and boundary information.

[0042] According to embodiments of this disclosure, by combining the distribution information of the target substance with the boundary information of sub-objects, sub-objects in the target image are automatically delineated based on the fusion of different types of features, which can significantly improve the accuracy and robustness of the delineation. The distribution information of the target substance reflects the physical or chemical composition of different tissues, helping to distinguish tissue regions with similar morphology but different compositions; the boundary information enhances the clarity of the sub-object outlines, exhibiting strong recognition capabilities, especially when tissue boundaries are blurred or the image signal-to-noise ratio is low. By fusing these two types of information and inputting them into the target model, accurate segmentation of complex structures can be achieved, effectively overcoming the problems of misjudgment and missed detection caused by single features in existing methods.

[0043] According to embodiments of this disclosure, the target substance distribution information includes the distribution state information of the target substance and / or the density information of the target substance.

[0044] The target substance can be a physical or chemical component that has characterization significance in the target object, such as common human tissue constituent elements such as carbon, oxygen, hydrogen, and calcium, or a complex substance that can be identified by radiation response characteristics, such as soft tissue, bone tissue, adipose tissue, and lesion tissue.

[0045] Distribution state information can be used to describe the spatial location of the target substance in the target image, that is, the regional presence of the substance within the target object. It can take the form of grayscale images, pseudo-color images, or label images in two-dimensional or three-dimensional formats, and is typically obtained through element decomposition algorithms or physically based image reconstruction techniques. This type of information can reflect the component boundaries between different tissues, helping to enhance the distinguishability between sub-objects.

[0046] Density information is a numerical representation of the physical concentration, attenuation coefficient, or equivalent density of a target substance per unit space, typically calculated from CT values ​​(Hounsfield Units) or MRI signal intensity. This information can further refine the material characteristics of tissues, helping to improve the model's ability to distinguish tissue types.

[0047] According to embodiments of this disclosure, in addition to distribution state information and density information, the target material distribution information may also include relative proportion information, spectral response characteristic information, ionizing radiation absorption coefficient information, or magnetic resonance relaxation characteristic information. Relative proportion information describes the proportion of a specific material in a local region and can be used for tissue heterogeneity analysis; spectral response characteristic information reflects the attenuation differences of materials under multiple energy conditions; ionizing radiation absorption coefficient information quantifies the absorption capacity of materials for X-rays of different energies; magnetic resonance relaxation characteristic information can be used to distinguish the T1 and T2 relaxation time characteristics of different types of soft tissues. All of this information can be used as fusion feature inputs to a neural network model to improve the overall performance and accuracy of the automatic delineation algorithm.

[0048] Figure 2 The flowchart illustrating the generation of first information in an object delineation method according to an embodiment of the present disclosure is shown schematically.

[0049] like Figure 2 As shown, based on the aforementioned embodiments, S120 may include operations S210 to S240.

[0050] In operation S210, at least one test feature parameter of the target object is determined based on the target image data.

[0051] Specifically, the target image data can be image data obtained through medical imaging equipment (such as CT, MRI, or dual-energy CT). In this process, the target object is first imaged using different energy rays or scanning modes to obtain image information reflecting different physical or chemical characteristics of the target object. Test characteristic parameters include, but are not limited to, the HU value, pixel grayscale value, image density, relative electron density, and effective atomic number (Zeff) of the target object. These parameters can reflect the physical properties of the target object under different imaging modes. For example, the HU value is commonly used in CT images to represent the attenuation degree of different parts of the image and can be used to infer the density of a substance; while relative electron density and effective atomic number (Zeff) can provide the relative properties of different substances in the image, aiding in further classification and substance identification. For example, in dual-energy CT images, the HU values ​​obtained using two different energy sources at two different energy spectra can be used to calculate the relative electron density and effective atomic number of the target object, providing necessary test characteristic parameters for subsequent analysis of the target object.

[0052] In operation S220, at least one test material type of the target object is determined based on the test characteristic parameters.

[0053] Specifically, based on test feature parameters extracted from target image data, such as relative electron density and effective atomic number Zeff, the target object can be classified as a material. Test material type refers to classifying at least a portion of the target object into different categories based on its physical properties, elemental composition, and how these properties are represented in the image data. For example, by comparing the effective atomic number Zeff of the target object with a predetermined threshold, different tissue types can be distinguished. When Zeff is greater than a certain preset threshold, the corresponding part of the target object can be identified as bone tissue; if Zeff is less than the threshold, it can be identified as soft tissue. This process helps to establish accurate classification models for different types of tissues or substances, ensuring efficient analysis of image data. For example, in dual-energy CT scans, bone tissue typically contains a large amount of calcium, resulting in a high effective atomic number, while soft tissue is mainly composed of carbon, oxygen, and hydrogen, resulting in a lower Zeff. Based on this characteristic, test feature parameters can be used to quickly and accurately identify bone tissue and soft tissue.

[0054] In operation S230, based on the material distribution characteristics corresponding to the test material type and the various test characteristic parameters of the target object, the material distribution state information of at least one target substance in the target object is determined. Specifically, after determining the various material types of the target object, the material distribution characteristics of that type of material can be further analyzed. For example, the elemental composition of bone tissue includes elements such as calcium, phosphorus, carbon, oxygen, and nitrogen, and their distribution characteristics in different regions exhibit significant regularity. By analyzing the test characteristic parameters of the target object and combining them with the known material element distribution patterns, the specific physical or chemical composition and distribution of each substance in the target object can be inferred.

[0055] Material distribution state information refers to the spatial distribution characteristics of a target substance within a target object. This information reflects the probability of presence, relative concentration, or mass percentage of a specific substance at different locations. Typically, this information can be constructed as a two-dimensional image matrix or a three-dimensional voxel matrix, where the value of each pixel or voxel corresponds to the relative content or density level of the substance at that location, thereby achieving a visual representation of the substance in space.

[0056] For example, after determining that the test material type of the target object includes bone tissue, based on the elemental composition of the bone tissue (such as high concentrations of calcium and phosphorus, and moderate concentrations of carbon and oxygen) and the test characteristic parameters, the spatial distribution of calcium or carbon elements in the bone tissue can be further inferred, and then a distribution image of the element can be constructed. This image constitutes the material distribution state information of the target substance, which can be used for subsequent image processing, automatic delineation, or feature extraction tasks.

[0057] In operation S240, first information is generated based on the material information corresponding to at least one substance. Specifically, first information can be generated based on the distribution state information, density information, and elemental composition of the target substance. This information can be further used as input for image segmentation or organ delineation algorithms. The first information may include data such as the spatial distribution characteristics, elemental composition, and physical density of the target substance, providing more accurate basic data for subsequent automated processing or diagnosis. For example, in the analysis of bone tissue, the first information generated based on the distribution and physical properties of the target substance (such as calcium, phosphorus, etc.) can be a set of three-dimensional matrices, where each element represents the relative density or mass distribution of calcium in a certain region of the bone tissue. In this way, the first information provides a more detailed and accurate physical basis for subsequent delineation and segmentation.

[0058] Figure 3 The flowchart illustrating the drawing of a sub-object in an object drawing method according to an embodiment of the present disclosure is shown schematically.

[0059] like Figure 3 As shown, based on the aforementioned embodiments, S130 may include operations S310 to S350.

[0060] In operation S310, the first information and the second information are fused to obtain the target information.

[0061] Specifically, the first information represents the distribution characteristics of the target material within the target object, while the second information represents the boundary contour information of each sub-object. These two information possess different semantic feature dimensions. To fully utilize their complementarity in material composition and morphological boundaries, a feature fusion operation is used to fuse the first and second information in the spatial or feature domains, forming target information containing rich structural and component features. This fusion operation can include pixel-level fusion, feature-level concatenation, and attention-guided weighted fusion to enhance the subsequent model's ability to express complex structures. The first and second information can be concatenated into a multi-dimensional input tensor along the channel dimension, or they can be input into different encoding branches and fused in an intermediate layer, with a unified output of the fused target information. This target information will simultaneously reflect the boundary morphology and material distribution state of the sub-objects.

[0062] In operation S320, target information is input into the target model to generate an image of at least one sub-object of the target. Specifically, the target model can be a trained deep learning neural network model that can receive the fused target information, automatically extract multi-level information such as semantic features, spatial features, and boundary features, and generate a prediction result to delineate the target sub-object. The generated sub-object image can be a label image, mask image, or pseudo-color image, identifying the precise region of the corresponding sub-object in the target image. For example, the target model can adopt a network architecture with an encoder-decoder structure, extracting multi-scale features in the encoding stage and combining shallow boundary information for refined prediction in the decoding stage, thereby outputting a more accurate sub-object image result, that is, it can delineate organs with higher accuracy.

[0063] Figure 4 A schematic diagram illustrating the structure of a target model in an object delineation method according to an embodiment of the present disclosure is shown.

[0064] like Figure 4 As shown, the target model includes an upsampling module, a downsampling module, and an edge linear attention module. The downsampling module is used to extract deep features from the target information. The upsampling module is used to extract semantic information from the deep features and generate images of sub-objects. The edge linear attention module is used to generate boundary information based on the feature maps generated by the upsampling and downsampling modules and transmit it to the upsampling module so that the upsampling module can generate images of sub-objects based on the boundary information.

[0065] like Figure 4As shown, the target model adopts an encoder-decoder structure. The left side represents the encoder section. Image data is first input to the Edge Encoder to extract edge-related features. Then, the Mamba Encoder further transforms the input data into a compressed feature representation (deep features), forming the downsampling module. In the downsampling module, feature maps (Encoder Feature Maps) are extracted layer by layer, and through multi-level downsampling operations indicated by arrows, the spatial resolution is continuously compressed and deep features are extracted.

[0066] After downsampling, the feature map is input to the decoder path for step-by-step reconstruction, corresponding to the upsampling module in the diagram. This module restores the low-resolution feature map to high-resolution features through the upsampling operation indicated by the arrow. Each downsampling layer generates a corresponding output (Decoder Feature Map), which is transmitted step-by-step until a structurally complete and accurately distributed sub-object image is generated. The upsampling module not only performs spatial reconstruction tasks but also fuses semantic and boundary information to generate a structurally complete and accurately distributed sub-object image.

[0067] Between the upsampling and downsampling paths, multiple edge linear attention modules are set up to establish boundary enhancement connections between layers of different scales, represented by cubes in the figure. This module simultaneously receives feature maps from the downsampling module (encoder path) and the upsampling module (decoder path), and generates boundary information (edge ​​feature map) based on them.

[0068] like Figure 4 As shown in the upper right corner, firstly, the feature map from the encoder path (the yellow feature block in the upper left of the figure) is input into a Softmax module to generate an attention distribution for the encoder features; at the same time, the feature map from the decoder path is also input into another independent Softmax module to generate an attention distribution on the decoder side.

[0069] Subsequently, the feature map from the decoder path and the attention weights generated by the encoder path undergo the first matrix multiplication operation (Matrix Multiply, the first X module in the figure) to form a preliminary fused feature. Then, this preliminary fused feature will undergo a second matrix multiplication operation (Matrix Multiply, the second X module in the figure) with the feature map from the encoder path to further model the deep correlation between contexts and generate an intermediate edge feature map containing edge representations.

[0070] The intermediate edge feature map enters the feature fusion module (Concat, module C in the figure), where it is stitched and fused with the feature map of the decoder path to form a structurally enhanced edge-aware feature map (the blue-gray fused feature block in the lower right corner of the figure). This map serves as one of the inputs to the subsequent upsampling module, used to generate sub-object images with clear boundaries.

[0071] Figure 5 The diagram illustrates a flowchart of the training process for a target model in an object delineation method according to an embodiment of the present disclosure.

[0072] like Figure 5 As shown, based on the aforementioned embodiments, the training process of the target model may include operations S510~S520.

[0073] In operation S510, sample first information and sample second information are obtained. Sample first information represents the distribution state information of sample substances in the sample object, and sample second information represents the sample boundary information of each sub-object in the target object.

[0074] Specifically, the first information of the sample is the material distribution information extracted based on the physical characteristics of the sample object after performing medical image scanning (such as CT, MRI, or dual-energy CT). This material distribution information can include the spatial distribution characteristics of specific substances (such as calcium, carbon, and adipose tissue) within the sample object, typically represented as a two-dimensional or three-dimensional matrix, where each element corresponds to the relative concentration or mass percentage of the specific substance within the sample object. The second information is the boundary contour information extracted from the image data of the target object using edge detection algorithms or pre-annotation methods. This information characterizes the spatial boundary positions of various sub-objects (such as organs, bones, and lesions) within the target object and can be presented as a mask or edge label map. For example, for a bone sample object, the first information could be a three-dimensional distribution matrix of calcium within the bone, reflecting the concentration changes of calcium in different regions; the second information could be a mask map of the bone edges, marking the boundary between the bone and surrounding soft tissue, providing boundary supervision information for subsequent model training.

[0075] In operating the S520, a target model is trained based on the first sample information, the second sample information, and the target loss function. The target model can be a deep learning network with an encoder-decoder structure. The training process uses the first and second sample information as inputs, fusing features from both to optimize model parameters. During training, the first and second sample information are first preprocessed, for example, through feature concatenation or weighted fusion, integrating material distribution state information and boundary information into a comprehensive target input feature. This target input feature is then fed into the target model, which generates prediction results through forward propagation. The target loss function measures the difference between the model's prediction and the true label, and updates the model parameters through backpropagation, gradually converging the model to its optimal state. The training process can employ batch gradient descent or its variants (such as the Adam optimizer) to accelerate convergence, while setting appropriate learning rates and regularization strategies (such as Dropout) to prevent overfitting. For example, in medical image segmentation tasks, the first information of the sample (calcium element distribution matrix) and the second information of the sample (bone edge mask map) are input into the target model. The model learns the relationship between material distribution and boundary features through multiple rounds of iteration, and finally generates sub-object images that can accurately delineate the bone region.

[0076] According to embodiments of this disclosure, the target loss function includes at least a first loss function, which is used to calculate the error between the content output by the edge linear attention module and the sample second information in the output of the target model.

[0077] Specifically, the edge linear attention module is mainly used to capture the boundary features of each sub-object within the target object and output edge feature maps to represent its spatial boundary structure. The first loss function supervises the edge feature maps output by this module against the true edge label maps in the sample's second information, thereby optimizing the attention mechanism's response to edge regions. Through this supervision mechanism, the model can explicitly learn the boundary-related attention distribution during training, making the generated feature maps closer to the actual edge regions, thus improving the structural integrity and edge alignment accuracy of the overall segmentation result.

[0078] For example, suppose the edge feature map output by the edge linear attention module is... ,in, This represents the edge response intensity at the i-th pixel; the edge label map in the second information of the sample is... ,in, This indicates whether the i-th pixel is an edge pixel, and N is the total number of pixels.

[0079] The first loss function can be the binary cross entropy (BCE) loss function L.edge Its expression can be as follows:

[0080]

[0081] In some embodiments, the loss function can be extended to an edge-weighted cross-entropy loss, assigning higher weights to regions near true boundary pixels to improve the model's sensitivity to blurred boundaries or fine-grained structures. This loss function can be optimized in parallel with the target model's main segmentation loss, thereby achieving an end-to-end boundary-guided training framework.

[0082] It should be noted that the above loss function is only an example. Those skilled in the art can design and select other loss functions that conform to the scheme provided in this disclosure, which should be included within the scope of protection of this disclosure.

[0083] According to embodiments of this disclosure, by combining element decomposition and boundary detection techniques from CT images and utilizing a deep learning model for multi-feature fusion, the accuracy and robustness of organ delineation are significantly improved. Compared to traditional methods, this invention uses the element decomposition results of CT images to reflect the boundary information of organs, avoiding errors caused by multimodal image registration. Simultaneously, it obtains accurate boundary features through an edge extraction algorithm, which are input into the network along with the original grayscale information from CT images, achieving end-to-end feature extraction and fusion. A Mamba encoder is used to efficiently capture multi-level image information, and a BoundaryLinearAttention module is innovatively introduced to achieve efficient fusion of multi-scale features through a linear attention mechanism, further enhancing the ability to represent boundary details. Furthermore, the introduction of a boundary-assisted loss function strengthens the model's ability to identify boundary regions, resulting in higher accuracy and adaptability in complex structures and low-contrast regions. This provides more reliable and efficient technical support for tumor radiotherapy planning, significantly reducing clinical workload and improving treatment outcomes.

[0084] Figure 6 A block diagram of an object drawing apparatus according to an embodiment of the present disclosure is shown schematically.

[0085] like Figure 6 As shown, the object drawing device 600 may include a first acquisition module 610, a first generation module 620, and a drawing module 630.

[0086] The first acquisition module 610 is used to acquire target image data, which at least represents an image generated based on the target ray of the target object, and the target object contains multiple sub-objects. In some embodiments, the first acquisition module 610 can be used to perform operation S110 in the object delineation method described above, which will not be elaborated here.

[0087] The first generation module 620 is used to generate first information and second information based on the target image data. The first information represents the distribution information of target substances in the target object, and the second information represents the boundary information of each sub-object in the target object. In some embodiments, the first generation module 620 can be used to perform operation S120 in the object delineation method described above, which will not be elaborated here.

[0088] The outlining module 630 is used to outline the image of at least one sub-object based on the first information and the second information. In some embodiments, the outlining module 630 can be used to perform operation S130 in the object outlining method described above, which will not be elaborated here.

[0089] Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure, or at least part of the functions of any one or more of them, can be implemented in one module. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be implemented by dividing them into multiple modules. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, one or more of the modules, submodules, units, and subunits according to embodiments of the present disclosure can be at least partially implemented as computer program modules, which, when run, can perform corresponding functions.

[0090] For example, any plurality of the first acquisition module 610, the first generation module 620, and the delineation module 630 can be combined into one module / unit / subunit, or any one of these modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functionality of one or more of these modules / units / subunits can be combined with at least part of the functionality of other modules / units / subunits and implemented in one module / unit / subunit. According to embodiments of this disclosure, at least one of the first acquisition module 610, the first generation module 620, and the delineation module 630 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging the circuitry, or implemented in hardware or firmware, or in any one of software, hardware, and firmware implementations or a suitable combination of any of these. Alternatively, at least one of the first acquisition module 610, the first generation module 620, and the outlining module 630 may be implemented at least partially as a computer program module, which can perform corresponding functions when the computer program module is run.

[0091] It should be noted that the data processing system part in the embodiments of this disclosure corresponds to the data processing method part in the embodiments of this disclosure. The specific description of the data processing system part is referred to in the data processing method part, and will not be repeated here.

[0092] Figure 7 A block diagram of an electronic device suitable for implementing the methods described above, according to embodiments of the present disclosure, is illustrated schematically. Figure 7 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0093] like Figure 7 As shown, an electronic device 700 according to an embodiment of the present disclosure includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 702 or a program loaded from a storage portion 708 into a random access memory (RAM) 703. The processor 701 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 701 may also include onboard memory for caching purposes. The processor 701 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0094] RAM 703 stores various programs and data required for the operation of electronic device 700. Processor 701, ROM 702, and RAM 703 are interconnected via bus 704. Processor 701 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 702 and / or RAM 703. It should be noted that the programs may also be stored in one or more memories other than ROM 702 and RAM 703. Processor 701 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in said one or more memories.

[0095] According to embodiments of this disclosure, the electronic device 700 may further include an input / output (I / O) interface 705, which is also connected to a bus 704. The electronic device 700 may also include one or more of the following components connected to the input / output (I / O) interface 705: an input section 706 including a keyboard, mouse, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the input / output (I / O) interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed.

[0096] According to embodiments of this disclosure, the method flow according to embodiments of this disclosure can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 709, and / or installed from removable medium 711. When the computer program is executed by processor 701, it performs the functions defined in the system of embodiments of this disclosure. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0097] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0098] According to embodiments of this disclosure, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0099] For example, according to embodiments of this disclosure, a computer-readable storage medium may include the ROM 702 and / or RAM 703 described above and / or one or more memories other than ROM 702 and RAM 703.

[0100] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this disclosure. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the control methods provided in the embodiments of this disclosure.

[0101] When the computer program is executed by the processor 701, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0102] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices or magnetic storage devices. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via communication section 709, and / or installed from removable medium 711. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof. According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code may be executed entirely on a user computing device, partially on a user device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing devices can be connected to user computing devices via any type of network, including local area networks (LANs) or wide area networks (WANs), or they can be connected to external computing devices (e.g., via the Internet using an Internet service provider).

[0103] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of the present disclosure can be combined and / or combined in various ways, even if such combinations are not explicitly described in the present disclosure. In particular, the features described in the various embodiments of this disclosure may be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0104] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A method for drawing an object, comprising: Acquire target image data, wherein the target image data at least characterizes an image generated on a target object based on a target ray, and the target object contains multiple sub-objects, wherein the sub-objects are organs; Based on the target image data, first information and second information are generated. The first information represents the distribution information of target substances in the target object, and the second information represents the boundary information of each sub-object in the target object. The first information is generated by element decomposition of the target image data. Based on the first information and the second information, draw an image of at least one of the sub-objects; The target substance distribution information includes the distribution state information of the target substance and / or the density information of the target substance; The target substance includes at least one of the elements carbon, oxygen, hydrogen, and calcium in the target object.

2. The method according to claim 1, wherein, The generation of the first information includes: Based on the target image data, at least one test feature parameter of the target object is determined; Based on the test characteristic parameters, at least one test material type for the target object is determined; Based on the material distribution characteristics corresponding to the test material type and the various test characteristic parameters of the target object, determine the material information of at least one of the target substances in the target object; First information is generated based on the substance information corresponding to at least one of the substances.

3. The method according to claim 1, wherein, Based on the first information and the second information, delineate the image of at least one of the sub-objects, including: The target information is obtained by fusing the first information and the second information. The target information is input into the target model to generate an image of at least one of the target's sub-objects.

4. The method according to claim 3, wherein, The target model includes an upsampling module, a downsampling module, and an edge linear attention module; The downsampling module is used to extract deep features from the target information; The upsampling module is used to extract semantic information from deep features and generate an image of the sub-object. The edge linear attention module is used to generate boundary information based on the feature maps generated by the upsampling module and the downsampling module and transmit it to the upsampling module, so that the upsampling module generates the image of the sub-object based on the boundary information.

5. The method according to claim 3, wherein, The training process of the target model includes: Obtain sample first information and sample second information, wherein the sample first information represents the sample material distribution state information in the sample object, and the sample second information represents the sample boundary information of each sub-object in the target object; The target model is trained based on the first information of the sample, the second information of the sample, and the target loss function.

6. The method according to claim 5, wherein, The target loss function includes at least a first loss function, which is used to calculate the error between the image boundary of the sub-object and the second information of the sample in the output of the model.

7. An object outlining device, comprising: The first acquisition module is used to acquire target image data, wherein the target image data at least characterizes an image generated on a target object based on a target ray, and the target object includes multiple sub-objects, wherein the sub-objects are organs. The first generation module is used to generate first information and second information based on the target image data. The first information represents the distribution information of target substances in the target object, and the second information represents the boundary information of each sub-object in the target object. The first information is generated by element decomposition of the target image data. as well as A delineation module is used to delineate the image of at least one of the sub-objects based on the first information and the second information; The target substance distribution information includes the distribution state information of the target substance and / or the density information of the target substance; The target substance includes at least one of the elements carbon, oxygen, hydrogen, and calcium in the target object.

8. 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 to enable the at least one processor to perform the method of any one of claims 1 to 6.

Images