Pelvic fracture CT image segmentation method based on dynamic spatial feature enhancement module

Abstract

The invention provides a pelvic fracture CT image segmentation method based on a dynamic spatial feature enhancement module, relates to the technical field of medical image segmentation, and solves the technical problem of low accuracy of result data of pelvic fracture CT image segmentation at present. The method comprises the following steps: extracting pelvic bones from pelvic CT scanning data by using a specified anatomical segmentation network; image segmentation is carried out on bone fragments in each pelvic region in the pelvic bone by using a fracture segmentation network comprising a dynamic spatial feature enhancement module; the fracture segmentation network captures global background data and local detail data in a pelvis region in an image segmentation process and keeps an object boundary by using a dynamic spatial feature enhancement module through a multi-path structure, an attention mechanism and a residual learning mode; key fracture fragments of the target skeleton in the pelvis CT scanning data are obtained through the image segmentation process of the fracture segmentation network.

Description

Technical Field

[0001] This application relates to the technical field of medical image segmentation, and in particular, to a method for segmenting pelvic fracture CT images based on a dynamic spatial feature enhancement module. Background Art

[0002] Pelvic fracture segmentation is a key link in the diagnosis and treatment planning of pelvic trauma. Accurately segmenting the pelvic fracture site in the medical images (usually CT images) of fracture patients can provide detailed cross-sectional information for clinicians, helping them accurately evaluate the fracture type, fracture displacement degree, and fracture fragment position, thereby improving the accuracy of pre-surgical planning and post-surgical evaluation. However, the pelvic anatomical structure is complex, the fracture types are diverse, and there are significant individual differences among different patients, making it difficult for automated segmentation. In addition, pelvic fractures are usually accompanied by other soft tissue injuries or bleeding, and there may be tomographic artifacts, noise, and metal implants in the CT images, which will pose higher requirements for the robustness of the segmentation algorithm.

[0003] Currently, most traditional pelvic segmentation methods rely on manual or semi-automatic strategies based on thresholding, region growing, or morphological operations. By adjusting the threshold and selecting seed points, adaptive threshold segmentation and region growing methods are used to extract the bone region. Subsequently, the fracture surface is manually depicted by outlining the fragments in a three-dimensional view or modifying the mask slice by slice. Although these methods have certain feasibility in relatively regular anatomical structures, a large number of interactive operations by professional doctors are still required when facing complex fracture morphologies, resulting in the accuracy of the segmentation results being easily affected by subjective factors, making the accuracy of the current pelvic fracture CT image segmentation result data relatively low. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for segmenting pelvic fracture CT images based on a dynamic spatial feature enhancement module to solve the technical problem of relatively low accuracy of the current pelvic fracture CT image segmentation result data.

[0005] In a first aspect, this application provides a method for segmenting pelvic fracture CT images based on a dynamic spatial feature enhancement module, and the method includes:

[0006] Obtain pelvic CT scan data to be segmented and processed;

[0007] Extract pelvic bones from the pelvic CT scan data by using a specified anatomical segmentation network; wherein, the specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, and the cascaded 3D nn-Unet architecture is generated by pre-training with a normal pelvic CT image dataset and adjusting the fractured pelvic dataset.

[0008] Use a fracture segmentation network containing a Dynamic Spatial Feature Enhancement Block (DSFE-Block) to perform image segmentation on the bone fragments in each pelvic region of the pelvic bone, and use the Dynamic Spatial Feature Enhancement Block to enable the fracture segmentation network to capture the global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning. Obtain the key fracture fragments of the target bone in the pelvic CT scan data through the image segmentation process of the fracture segmentation network; wherein, the multi-path structure corresponds to multiple parallel paths that respectively process the initial input tensor input to the Dynamic Spatial Feature Enhancement Block, and the multiple parallel paths include a local feature path, a parallel pooling path, and an attention path.

[0009] In a possible implementation, the initial input tensor is represented by X ∈ R B×C×D×H×W where B represents the batch size, C represents the number of channels, and D, H, and W represent the spatial dimensions;

[0010] The local feature path is used to process the initial input tensor through a 3×3×3 convolutional layer, a batch normalization BN layer, and a ReLU activation function using the following formula:

[0011] X1 = ReLU(BN(Conv3D(X)));

[0012] where X1 represents the processing result of the local feature path on the initial input tensor, X represents the initial input tensor, ReLU represents the activation function, BN represents the batch normalization layer function, and Conv3D represents an additional 3×3×3 convolutional layer.

[0013] In a possible implementation, the parallel pooling path is used to perform max pooling and average pooling on the local feature path processing result through the following formula, and concatenate the max pooling processing result and the average pooling processing result along the channel dimension:

[0014] X pool = Concat(MaxPool3D(X1), AvgPool3D(X1));

[0015] where X1 represents the local feature path processing result, Convcat represents the concatenation processing function for two convolutional operations, MaxPool3D represents the max pooling convolutional layer function with a pooling kernel size of 3×3×3, AvgPool3D represents the average pooling convolutional layer function with a pooling kernel size of 3×3×3, X PoolDenotes the combined feature after concatenating the max pooling processing result and the average pooling processing result;

[0016] The parallel pooling path is also used to further process the combined feature by the following formula using the activation function, the batch normalization layer function, and the additional 3×3×3 convolutional layer:

[0017] X2 = ReLU(BN(Conv3D(X pool )));

[0018] where X2 represents the result data after further processing the combined feature, ReLU represents the activation function, BN represents the batch normalization layer function, X Pool represents the combined feature, and Conv3D represents the additional 3×3×3 convolutional layer.

[0019] In a possible implementation, the attention path is used to calculate the attention coefficient by the following formula using the additional 3×3×3 convolutional layer and the Sigmoid activation function:

[0020] X att = σ(Conv3D(GAP(X)));

[0021] where σ represents the Sigmoid activation function, Conv3D represents the additional 3×3×3 convolutional layer, X represents the initial input tensor, Xatt represents the attention coefficient, and GAP represents the global average pooling function.

[0022] In a possible implementation, after obtaining the attention coefficient, the attention coefficient is applied to the result data by element-wise multiplication by the following formula to obtain the weighted feature:

[0023] X weighted = X2 ⊙ X att ;

[0024] where ⊙ represents element-wise multiplication, X2 represents the result data after further processing the combined feature, X att represents the attention coefficient, and X weighted represents the weighted feature;

[0025] The execution process of the residual learning method includes combining the weighted feature with the initial input tensor by residual connection by the following formula to retain the original information:

[0026] Y = X + X weighted ;

[0027] where X weighteddenotes the weighted feature, X denotes the initial input tensor, and Y denotes the final result after combining the weighted feature with the initial input tensor.

[0028] In a possible implementation, the loss function of the fracture segmentation network is a mixed DC_and_BCE_loss, which simultaneously optimizes the voxel-level classification accuracy and the region-level segmentation quality through the mixed DC_and_BCE_loss; the mixed DC_and_BCE_loss is used to ignore the loss calculation of specific regions through a masking mechanism to enhance the applicability for processing specified complex segmentation tasks; wherein, the specified complex segmentation tasks include cases where some regions are ignored.

[0029] In a possible implementation, the loss function of the fracture segmentation network includes a binary cross-entropy BCE loss function and a Soft Dice loss function; the BCE loss function is used to measure the classification difference between the predicted segmentation result and the target segmentation region at the voxel level; the Soft Dice loss function is used to evaluate the overlap degree between the predicted segmentation result and the target segmentation region through the following formula to optimize the matching quality of the overall region:

[0030]

[0031] wherein, denotes the binary cross-entropy loss function, which is used for voxel-level classification loss; denotes the Soft Dice loss function, which is used to measure the overlap of the segmentation region; α and β are two weight hyperparameters, which are respectively used to control the contribution ratio of the BCE loss and the Dice loss to the total loss.

[0032] In a second aspect, the present application provides a pelvic fracture CT image segmentation device based on a dynamic spatial feature enhancement module, including:

[0033] An acquisition module, which is used to acquire pelvic CT scan data to be segmented and processed;

[0034] An extraction module, which is used to extract pelvic bones from the pelvic CT scan data by using a specified anatomical segmentation network; wherein, the specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, and the cascaded 3D nn-Unet architecture is generated by pre-training with a normal pelvic CT image dataset and adjusting the fracture pelvic dataset.

[0035] A segmentation module is configured to perform image segmentation on bone fragments in each pelvic region of the pelvic bone by using a fracture segmentation network including a dynamic spatial feature enhancement module, and use the dynamic spatial feature enhancement module to enable the fracture segmentation network to capture global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning. Key fracture fragments of the target bone in the pelvic CT scan data are obtained through the image segmentation process of the fracture segmentation network. The multi-path structure corresponds to multiple parallel paths that respectively process an initial input tensor input to the dynamic spatial feature enhancement module, and the multiple parallel paths include a local feature path, a parallel pooling path, and an attention path.

[0036] In a third aspect, the present application further provides an electronic device, including a memory and a processor. A computer program that can run on the processor is stored in the memory. When the processor executes the computer program, the method described in the first aspect above is implemented.

[0037] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and run by a processor, the computer-executable instructions cause the processor to run the method described in the first aspect above.

[0038] The present application brings the following beneficial effects:

[0039] A pelvic fracture CT image segmentation method based on a dynamic spatial feature enhancement module provided by this application can obtain pelvic CT scan data to be segmented; use a specified anatomical segmentation network to extract pelvic bones from the pelvic CT scan data; wherein, the specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, and the cascaded 3D nn-Unet architecture is generated by pre-training with a normal pelvic CT image dataset and adjusting a fractured pelvic dataset; use a fracture segmentation network containing a dynamic spatial feature enhancement module to perform image segmentation on bone fragments in each pelvic region of the pelvic bones, and use the dynamic spatial feature enhancement module to capture global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning. Through the image segmentation process of the fracture segmentation network, key fracture fragments of the target bones in the pelvic CT scan data are obtained; wherein, the multi-path structure corresponds to multiple parallel paths that respectively process the initial input tensor input to the dynamic spatial feature enhancement module, and the multiple parallel paths include a local feature path, a parallel pooling path, and an attention path. In this solution, through the multi-path design, attention mechanism, and residual learning of the dynamic spatial feature enhancement module, the fracture segmentation network can effectively capture fine details and retain object boundaries, achieving a more accurate segmentation effect, improving the data accuracy of the pelvic fracture CT image segmentation result, and solving the technical problem of the low data accuracy of the current pelvic fracture CT image segmentation result.

[0040] To make the above objects, features, and advantages of this application more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] To more clearly illustrate the specific embodiments of this application or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of this application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0042] Figure 1 It is a schematic flowchart of the pelvic fracture CT image segmentation method based on the dynamic spatial feature enhancement module provided by the embodiment of this application;

[0043] Figure 2 It is another schematic flowchart of the pelvic fracture CT image segmentation method based on the dynamic spatial feature enhancement module provided by the embodiment of this application;

[0044] Figure 3 In the pelvic fracture CT image segmentation method based on the dynamic spatial feature enhancement module provided by the embodiments of the present application, it is an example of the internal architecture of the dynamic spatial feature enhancement module DSFE-Block;

[0045] Figure 4 It is a schematic structural diagram of a pelvic fracture CT image segmentation device based on the dynamic spatial feature enhancement module provided by the embodiments of the present application;

[0046] Figure 5 It shows a schematic structural diagram of an electronic device provided by the embodiments of the present application. Detailed implementation manners

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0048] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes other steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products, or devices.

[0049] With the rapid development of medical image processing and artificial intelligence technologies, more and more research has begun to attempt to improve the accuracy of pelvic fracture segmentation through automatic segmentation algorithms based on deep learning. Typical methods include using convolutional neural networks (CNNs) or hybrid network structures to automatically segment the pelvic region and detect fractures. Currently, deep convolutional neural networks (CNNs) can be used in combination with multi-directional block images for pelvic fracture detection. By processing CT images from multiple perspectives, the network can capture the features of the pelvic region more comprehensively, thereby improving the accuracy of fracture detection. Although this method provides a fast and accurate fracture diagnosis tool for clinical use, its output cannot provide a fully automated solution for subsequent operations.

[0050] In the pelvic fracture segmentation task, due to the complex pelvic anatomical structure, diverse fracture region morphologies, and high uncertainty, there are many problems with current pelvic fracture segmentation methods. For example, insufficient local and global feature representation: The morphology and position of the pelvic fracture region have significant uncertainty, and its distribution may cover a large range of the pelvic region, while containing small and complex local features such as bone fractures, fragments, or irregular boundaries. Traditional segmentation methods have problems with insufficient local features when capturing local details: lacking the ability to accurately capture local features such as fine cracks and edges, resulting in inaccurate segmentation of small fractures. Thus, the accuracy of the current result data for pelvic fracture CT image segmentation is relatively low.

[0051] In addition, the spatial and channel dependencies of traditional segmentation methods are weakened. The fracture region may show discontinuity in the spatial dimension (for example, bone fractures may be distributed at different positions in the pelvis), while there may be implicit associations in the channel dimension (for example, different channels of multimodal medical images or the feature channels of the segmentation model). Conventional segmentation models generally ignore the following aspects when extracting and fusing features: First, the discontinuity of spatial features: The spatial distribution of the fracture region is complex and may be discontinuous, and existing models are difficult to capture these non-linear features through a single scale or simple convolution. Second, the lack of interaction between channels: The associated features between different channels (such as different window widths and window levels of CT images) are not fully exploited, resulting in a reduction in the integrity of feature representation. Therefore, the accuracy of the current result data for pelvic fracture CT image segmentation is relatively low.

[0052] Based on this, the embodiments of the present application provide a method for pelvic fracture CT image segmentation based on a dynamic spatial feature enhancement module, which can solve the technical problem of relatively low accuracy of the current result data for pelvic fracture CT image segmentation.

[0053] The embodiments of the present invention will be further introduced below with reference to the accompanying drawings.

[0054] Figure 1 It is a schematic flow chart of a method for pelvic fracture CT image segmentation based on a dynamic spatial feature enhancement module provided by the embodiments of the present application. As Figure 1 shown, the method includes:

[0055] Step S110, obtaining pelvic CT scan data to be segmented.

[0056] As a possible implementation, in order to automatically segment the main fragments of the target bone from the CT scan, first obtain the pelvic CT scan data to be segmented. The pelvic CT scan data herein is the pelvic fracture CT image (pelvic fracture CT picture).

[0057] Step S120: Extract pelvic bones from pelvic CT scan data using a specified anatomical segmentation network.

[0058] The specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, which is generated by pre-training with a normal pelvic CT image dataset and adjusting the fractured pelvic dataset.

[0059] As Figure 2 shown, the method provided by the embodiment of the present application includes two stages. In the first stage, first, a pelvic bone is extracted from a CT scan using an anatomical segmentation network (specified anatomical segmentation network), and the anatomical segmentation network adopts a cascaded 3D nn-Unet architecture, which is pre-trained on a healthy pelvic CT image dataset (normal pelvic CT image dataset) and further fine-tuned on a fractured pelvic dataset.

[0060] Step S130: Use a fracture segmentation network containing a dynamic spatial feature enhancement module to perform image segmentation on bone fragments in each pelvic region of the pelvic bones, and use the dynamic spatial feature enhancement module to capture global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning. The key fracture fragments of the target bones in the pelvic CT scan data are obtained through the image segmentation process of the fracture segmentation network.

[0061] As Figure 2 shown, in the second stage, then apply a fracture segmentation network (FractureSeg network) to segment the bone fragments in each pelvic region. As an example, the fracture segmentation network consists of four key components: an encoder, a decoder, a bottleneck layer (Vision-LSTM), and a dynamic spatial feature enhancement module (DSFE-Block). Among them, the architecture of the DSFE-Block consists of three parallel paths: a local feature path, a parallel pooling path, and an attention path, and the three parallel paths process the input tensor respectively. It can also be understood that the multi-path structure corresponds to multiple parallel paths that process the initial input tensor input to the dynamic spatial feature enhancement module respectively, and the multiple parallel paths include a local feature path, a parallel pooling path, and an attention path.

[0062] Due to the diverse sizes and shapes of fracture fragments, capturing local details and global context is crucial for the segmentation of pelvic fractures. In the embodiments of the present application, a dynamic spatial feature enhancement module is used in the skip connections of the fracture segmentation network. Through the multi-path design, attention mechanism, and residual learning of the dynamic spatial feature enhancement module, the fracture segmentation network can effectively capture fine details and retain object boundaries, achieving a more accurate segmentation effect, improving the data accuracy of the pelvic fracture CT image segmentation results, and solving the technical problem of the relatively low data accuracy of the current pelvic fracture CT image segmentation results.

[0063] Regarding the discontinuity of spatial features, the spatial distribution of the fracture area is complex and discontinuous, and the dynamic spatial feature enhancement module captures non-linear spatial features through multi-path feature extraction; regarding the insufficient interaction between channels, fracture segmentation requires fully exploring the associations between different channels (such as medical image multi-modal data or feature channels), and the dynamic spatial feature enhancement module models the interaction between channels through the attention mechanism, that is, for the problem of weakened spatial and channel dependencies, the dynamic spatial feature enhancement module (DSFE-Block) effectively solves the problems of discontinuity in the spatial dimension, insufficient interaction in the channel dimension, and insufficient cross-spatial and channel comprehensiveness in the fracture area segmentation. The problems of discontinuity in the spatial dimension, insufficient interaction in the channel dimension, and insufficient cross-spatial and channel comprehensiveness in the fracture area segmentation are solved by the dynamic spatial feature enhancement module.

[0064] The fracture segmentation network provided by the embodiments of the present application significantly improves the effect of 3D pelvic fracture CT segmentation by effectively integrating the bottleneck layer and the dynamic spatial feature enhancement module. It not only overcomes the problem of feature loss in the conventional network when processing complex fracture areas, effectively retains key anatomical information, but also can capture local details and global semantic information of the pelvic fracture area at the same time, achieving an effective balance between detail performance and integrity, and significantly improving the performance and stability of the fracture segmentation task. The solution provided by the embodiments of the present application provides important technical support for fracture diagnosis and treatment planning in medical image processing and has broad clinical application prospects.

[0065] The above steps are introduced in detail below.

[0066] In some embodiments, the initial input tensor is represented by X ∈ R B×C×D×H×W , where B represents the batch size, C represents the number of channels, and D, H, and W represent the spatial dimensions; the local feature path is used to process the initial input tensor through a 3×3×3 convolutional layer, a batch normalization BN layer, and a ReLU activation function according to the following formula:

[0067] X1 = ReLU(BN(Conv3D(X)));

[0068] Among them, X1 represents the processing result of the local feature path on the initial input tensor, X represents the initial input tensor, ReLU represents the activation function, BN represents the batch normalization layer function, and Conv3D represents an additional 3×3×3 convolutional layer.

[0069] Regarding the discontinuity of the spatial features, the spatial distribution of the fracture area is complex and discontinuous. In the embodiments of the present application, the dynamic spatial feature enhancement module captures non-linear spatial features through multi-path feature extraction, such as Figure 3 As shown, in the local feature path, based on the initial input tensor X, a convolutional operation (3×3×3 Conv) is adopted, combined with batch normalization (BN) and ReLU activation, which strengthens the feature expression in the local area, so as to capture the local details of the fracture area more efficiently and accurately.

[0070] In some embodiments, the parallel pooling path is used to perform max-pooling processing and average-pooling processing on the processing result of the local feature path through the following formula, and concatenate the max-pooling processing result and the average-pooling processing result along the channel dimension:

[0071] X pool = Concat(MaxPool3D(X1), AvgPool3D(X1));

[0072] Among them, X1 represents the processing result of the local feature path, Convcat represents the concatenation processing function for two convolutional operations, MaxPool3D represents the max-pooling convolutional layer function with a pooling kernel size of 3×3×3, AvgPool3D represents the average-pooling convolutional layer function with a pooling kernel size of 3×3×3, and X Pool represents the combined feature after concatenating the max-pooling processing result and the average-pooling processing result;

[0073] The parallel pooling path is also used to further process the combined feature through the activation function, the batch normalization layer function, and an additional 3×3×3 convolutional layer through the following formula:

[0074] X2 = ReLU(BN(Conv3D(X pool )));

[0075] Among them, X2 represents the result data after further processing the combined feature, ReLU represents the activation function, BN represents the batch normalization layer function, X Pool represents the combined feature, and Conv3D represents an additional 3×3×3 convolutional layer.

[0076] In the parallel pooling path, such as Figure 3As shown, through the combined operation of max pooling (3×3×3 MaxPool) and average pooling (3×3×3 AvgPool), the multi-scale information of the spatial distribution is retained, and different pooling features are combined through a concatenation operation. Through this multi-scale mechanism, the discontinuity of the spatial distribution in the fracture area can be more comprehensively addressed, and the model of the fracture segmentation network can be enhanced in its perception of complex shapes.

[0077] In some embodiments, the attention path is used to calculate the attention coefficient by means of an additional 3×3×3 convolutional layer and a Sigmoid activation function according to the following formula:

[0078] X att = σ(Conv3D(GAP(X)));

[0079] where σ represents the Sigmoid activation function, Conv3D represents the additional 3×3×3 convolutional layer, X represents the initial input tensor, Xatt represents the attention coefficient, and GAP represents the global average pooling function.

[0080] To enhance the importance of specific features, the dynamic spatial feature enhancement module in the embodiments of the present application generates a context-aware attention map by performing a global average pooling (GAP) operation, as Figure 3 shown, and calculates the attention coefficient using a 3×3×3 convolutional layer (3×3×3 Conv) and a Sigmoid activation function. In view of the insufficient interaction between channels, the fracture segmentation network needs to fully explore the associations between different channels (such as multi-modal medical image data or feature channels). The dynamic spatial feature enhancement module realizes the modeling of the interaction between channels through the attention mechanism, that is, the global average pooling (GAP) globally aggregates the input features to generate a context-aware attention map, so as to more efficiently and accurately extract the correlations between channels.

[0081] In some embodiments, after obtaining the attention coefficient, the attention coefficient is applied to the result data through element-wise multiplication according to the following formula to obtain the weighted feature:

[0082] X weighted = X2 ⊙ X att ;

[0083] where ⊙ represents element-wise multiplication, X2 represents the result data after further processing of the combined features, X att represents the attention coefficient, and X weighted represents the weighted feature.

[0084] For feature fusion and residual learning, the attention coefficient Xatt is applied to X2 through element-wise multiplication to highlight the most important features. To address the insufficient interaction between channels, after using convolution to further process the output of GAP to generate the attention coefficient, it is applied to the pooled features to dynamically adjust the feature weights of each channel and strengthen the implicit association between channels.

[0085] The execution process of the above residual learning method includes combining the weighted features with the initial input tensor using residual connection through the following formula to retain the original information:

[0086] Y = X + X weighted ;

[0087] where, X weighted represents the weighted features, X represents the initial input tensor, and Y represents the final result after combining the weighted features with the initial input tensor.

[0088] To address the lack of comprehensiveness across space and channels, the dynamic spatial feature enhancement module improves the efficiency of feature fusion through cross-space and cross-channel collaborative modeling. The multi-path architecture of the dynamic spatial feature enhancement module realizes the joint expression of spatial information (local details, multi-scale features) and channel information (contextual attention, channel weights). In the residual connection, as Figure 3 shown, the dynamic spatial feature enhancement module adds the initial input X to the weighted features, achieving feature enhancement while retaining the integrity of the original information. This residual mechanism compensates for the limitations of a single feature path, enabling the collaborative modeling of spatial dependencies and channel implicit relationships, and improving the model working efficiency of the fracture segmentation network.

[0089] In the embodiments of the present application, through the above multi-scale feature modeling, spatial-channel dependency integration, and the modeling ability of remote context information, the segmentation efficiency of the fracture segmentation network for fracture fragments in pelvic CT scans can be improved.

[0090] In some embodiments, the loss function of the fracture segmentation network is a mixed DC_and_BCE_loss, which simultaneously optimizes the voxel-level classification accuracy and the region-level segmentation quality through the mixed DC_and_BCE_loss; the mixed DC_and_BCE_loss is used to ignore the loss calculation of specific regions through a masking mechanism to enhance the applicability for processing specified complex segmentation tasks; where the specified complex segmentation tasks include cases where some regions are ignored.

[0091] The loss function of the FractureSeg network adopts a hybrid DC_and_BCE_loss, which can optimize the voxel-level classification accuracy and the region-level segmentation quality simultaneously. In addition, DC_and_BCE_loss supports ignoring the loss calculation of specific regions through a masking mechanism, thereby enhancing its applicability in dealing with complex segmentation tasks (such as in the case where some regions are ignored). By combining the advantages of BCE and Dice losses, both the local voxel classification accuracy and the overall segmentation quality can be optimized.

[0092] In some embodiments, the loss function of the fracture segmentation network includes a binary cross-entropy (BCE) loss function and a Soft Dice loss function; the BCE loss function is used to measure the classification difference between the predicted segmentation result and the target segmentation region at the voxel level; the Soft Dice loss function is used to evaluate the overlap degree between the predicted segmentation result and the target segmentation region through the following formula to optimize the matching quality of the overall region:

[0093]

[0094] where represents the binary cross-entropy loss function, which is used for voxel-level classification loss; represents the Soft Dice loss function, which is used to measure the overlap of the segmentation region; α and β are two weight hyperparameters, which are used to control the contribution ratio of the BCE loss and the Dice loss to the total loss respectively.

[0095] Exemplarily, the loss function consists of two parts: the first part is the binary cross-entropy (BCE) loss, which is used to measure the classification difference between the predicted segmentation result and the target segmentation region at the voxel level; the second part is the Soft Dice loss, which is used to evaluate the overlap degree between the predicted segmentation result and the target segmentation region to optimize the matching quality of the overall pelvic region.

[0096] Figure 4 A structural schematic diagram of a pelvic fracture CT image segmentation device based on a dynamic spatial feature enhancement module is provided. As Figure 4 shown, the pelvic fracture CT image segmentation device 400 based on the dynamic spatial feature enhancement module includes:

[0097] An acquisition module 401, which is used to acquire pelvic CT scan data to be segmented and processed;

[0098] An extraction module 402 is configured to extract pelvic bones from the pelvic CT scan data by using a specified anatomical segmentation network; wherein, the specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, and the cascaded 3D nn-Unet architecture is generated by pre-training with a normal pelvic CT image dataset and adjusting a fractured pelvic dataset.

[0099] A segmentation module 403 is configured to perform image segmentation on bone fragments in each pelvic region of the pelvic bones by using a fracture segmentation network including a dynamic spatial feature enhancement module, and use the dynamic spatial feature enhancement module to capture global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning. Key fracture fragments of the target bones in the pelvic CT scan data are obtained through the image segmentation process of the fracture segmentation network; wherein, the multi-path structure corresponds to multiple parallel paths for respectively processing an initial input tensor input to the dynamic spatial feature enhancement module, and the multiple parallel paths include a local feature path, a parallel pooling path, and an attention path.

[0100] The pelvic fracture CT image segmentation device based on a dynamic spatial feature enhancement module provided by an embodiment of the present application has the same technical features as the pelvic fracture CT image segmentation method based on a dynamic spatial feature enhancement module provided by the above embodiment, so it can also solve the same technical problems and achieve the same technical effects.

[0101] An electronic device provided by an embodiment of the present application, as Figure 5 shown, the electronic device 500 includes a processor 502 and a memory 501. A computer program that can run on the processor is stored in the memory. When the processor executes the computer program, the steps of the method provided by the above embodiment are implemented.

[0102] See Figure 5 , the electronic device further includes: a bus 503 and a communication interface 504. The processor 502, the communication interface 504, and the memory 501 are connected through the bus 503; the processor 502 is configured to execute an executable module stored in the memory 501, such as a computer program.

[0103] Among them, the memory 501 may include a high-speed random access memory (Random Access Memory, referred to as RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 504 (which can be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0104] The bus 503 can be an ISA bus, a PCI bus, an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, Figure 5 only a bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0105] Among them, the memory 501 is used to store programs. After receiving an execution instruction, the processor 502 executes the programs. The methods executed by the devices defined by the processes disclosed in any embodiment of the present application can be applied to the processor 502 or implemented by the processor 502.

[0106] The processor 502 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor 502 or instructions in the form of software. The above-mentioned processor 502 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed and completed by the hardware decoding processor, or executed and completed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 501, and the processor 502 reads the information in the memory 501 and combines its hardware to complete the steps of the above method.

[0107] Corresponding to the above pelvic fracture CT image segmentation method based on the dynamic spatial feature enhancement module, an embodiment of the present application further provides a computer-readable storage medium, and the computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and run by a processor, the computer-executable instructions cause the processor to run the steps of the above pelvic fracture CT image segmentation method based on the dynamic spatial feature enhancement module.

[0108] The pelvic fracture CT image segmentation device based on the dynamic spatial feature enhancement module provided by the embodiments of the present application may be specific hardware on the device or software or firmware installed on the device, etc. For the device provided by the embodiments of the present application, its implementation principle and the technical effects produced are the same as those of the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding content in the foregoing method embodiments. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can all refer to the corresponding processes in the above method embodiments, and will not be repeated here.

[0109] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0110] For another example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of the devices, methods, and computer program products according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order from that marked in the accompanying drawings. For example, two consecutive blocks can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0111] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0112] In addition, the functional units in the embodiments provided in the present application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0113] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the pelvic fracture CT image segmentation method based on the dynamic spatial feature enhancement module described in various embodiments of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM for short), random access memories (RAM for short), magnetic disks, or optical discs that can store program codes.

[0114] It should be noted that: similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0115] Finally, it should be noted that: the above-mentioned embodiments are only specific implementation manners of this application, used to illustrate the technical solution of this application, rather than limiting it. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed in this application can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of this application. All should be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A pelvic fracture CT image segmentation method based on a dynamic spatial feature enhancement module, characterized in that: The method comprises: Acquire pelvic CT scan data to be segmented; Extracting pelvic bones from the pelvic CT scan data using a specified anatomical segmentation network; wherein the specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, and the cascaded 3D nn-Unet architecture is generated by pre-training with a normal pelvic CT image dataset and adjusting a fractured pelvic dataset; A fracture segmentation network including a dynamic spatial feature enhancement module is used to perform image segmentation on bone fragments in each pelvic region of the pelvic skeleton, and the dynamic spatial feature enhancement module is used to enable the fracture segmentation network to capture global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning. The key fracture fragments of the target bone in the pelvic CT scan data are obtained through the image segmentation process of the fracture segmentation network; wherein the multi-path structure corresponds to a plurality of parallel paths that respectively process the initial input tensors input to the dynamic spatial feature enhancement module, and the plurality of parallel paths include a local feature path, a parallel pooling path, and an attention path.

2. The method according to claim 1, characterized in that The initial input tensor is represented by X∈R B×C×D×H×W Indicates, where B represents the batch size, C represents the number of channels, and D, H and W represent the spatial dimensions; The local feature path is used to process the initial input tensor using a 3×3×3 convolutional layer, a batch normalization BN layer, and a ReLU activation function through the following formula: X1 = ReLU(BN(Conv3D(X))); Among them, X1 represents the local feature path processing result after the local feature path processes the initial input tensor, X represents the initial input tensor, ReLU represents the activation function, BN represents the batch normalization layer function, and Conv3D represents an additional 3×3×3 convolutional layer.

3. The method according to claim 2, characterized in that The parallel pooling path is used to perform maximum pooling and average pooling on the local feature path processing results through the following formula, and to concatenate the maximum pooling processing results and the average pooling processing results along the channel dimension: X pool =Concat(MaxPool3D(X1),AvgPool3D(X1)); Wherein, X1 represents the processing result of the local feature path, Convcat represents the concatenation processing function for two convolution operations, MaxPool3D represents the maximum pooling convolution layer function with a pooling kernel size of 3×3×3, AvgPool3D represents the average pooling convolution layer function with a pooling kernel size of 3×3×3, X Pool represents the combined features after splicing the maximum pooling processing result and the average pooling processing result; The parallel pooling path is also used to further process the merged features using the activation function, the batch normalization layer function, and the additional 3×3×3 convolutional layer through the following formula: X2=ReLU(BN(Conv3D(X pool ))); Wherein, X2 represents the result data after further processing of the merged features, ReLU represents the activation function, BN represents the batch normalization layer function, and X Pool represents the merged features, and Conv3D represents the additional 3×3×3 convolutional layer.

4. The method according to claim 3, characterized in that The attention path is used to calculate the attention coefficient using the additional 3×3×3 convolutional layer and the Sigmoid activation function by the following formula: X att =σ(Conv3D(GAP(X)))= Where σ represents the Sigmoid activation function, Conv3D represents the additional 3×3×3 convolutional layer, X represents the initial input tensor, and X att represents the attention coefficient, and GAP represents the global average pooling function.

5. The method according to claim 4, characterized in that After obtaining the attention coefficient, the attention coefficient is applied to the result data using element-wise multiplication through the following formula to obtain the weighted feature: X weighted =X2⊙X att ; Wherein, ⊙ represents element-level multiplication, X2 represents the result data after further processing the combined features, and X att represents the attention coefficient, X weighted represents weighted features; The residual learning method includes combining the weighted features with the initial input tensor using a residual connection to retain the original information through the following formula: Y=X+X weighted ; Among them, X weighted represents a weighted feature, X represents the initial input tensor, and Y represents the final result after combining the weighted feature with the initial input tensor.

6. The method according to claim 1, characterized in that The loss function of the fracture segmentation network is a hybrid DC_and_BCE_loss, which simultaneously optimizes the voxel-level classification accuracy and the region-level segmentation quality; the hybrid DC_and_BCE_loss is used to ignore the loss calculation of specific areas through a mask mechanism to enhance the applicability of processing specified complex segmentation tasks; wherein the specified complex segmentation task includes the situation where some areas are ignored.

7. The method according to claim 6, characterized in that The loss function of the fracture segmentation network includes a binary cross entropy BCE loss function and a Soft Dice loss function; the BCE loss function is used to measure the classification difference between the predicted segmentation result and the target segmentation area at the voxel level; the Soft Dice loss function is used to evaluate the overlap between the predicted segmentation result and the target segmentation area through the following formula to optimize the matching quality of the overall area: in, represents the binary cross entropy loss function, which is used for voxel-level classification loss; represents the SoftDice loss function, which is used to measure the overlap of the segmented regions; α and β are two weight hyperparameters, which are used to control the contribution ratio of BCE loss and Dice loss to the total loss respectively.

8. A pelvic fracture CT image segmentation device based on a dynamic spatial feature enhancement module, characterized in that: include: An acquisition module, used for acquiring pelvic CT scan data to be segmented; An extraction module for extracting pelvic bones from the pelvic CT scan data using a specified anatomical segmentation network; wherein the specified anatomical segmentation network is an optimized cascaded 3D nn-Unet architecture, and the cascaded 3D nn-Unet architecture is generated by pre-training with a normal pelvic CT image dataset and adjusting a fractured pelvic dataset; A segmentation module is used to perform image segmentation on bone fragments in each pelvic region of the pelvic skeleton using a fracture segmentation network including a dynamic spatial feature enhancement module, and to use the dynamic spatial feature enhancement module to enable the fracture segmentation network to capture global background data and local detail data in the pelvic region and retain object boundaries during the image segmentation process through a multi-path structure, an attention mechanism, and residual learning, and to obtain key fracture fragments of the target bone in the pelvic CT scan data through the image segmentation process of the fracture segmentation network; wherein the multi-path structure corresponds to a plurality of parallel paths for respectively processing the initial input tensors input to the dynamic spatial feature enhancement module, and the plurality of parallel paths include a local feature path, a parallel pooling path, and an attention path.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to execute the method according to any one of claims 1 to 7.

Images