11 results about "Low dose ct" patented technology

The invention provides a low-dose computed tomography image reconstruction method and system. The method comprises the steps that a first conventional-dose CT image and a first low-dose CT image are input into an image output model for training; inputting the second conventional dose CT image into the training image output model to simulate degradation; storing the second conventional dose CT image and the second low dose CT image which are matched with each other into a training image set, and inputting the images in the training image set into the image reconstruction model for hierarchical training; the to-be-reconstructed low-dose CT image is input into the training image reconstruction model for image denoising reconstruction so as to obtain a reconstructed conventional-dose CT image, the performance of the image obtained by the method in noise level and structural definition can be matched with or even better than that of a real conventional-dose CT image, and thus the risk that a patient is exposed to radiation is reduced.

The invention discloses a low-dose CT denoising method and device based on wavelet transform and a state space model, and relates to the technical field of medical image processing. The method comprises the following steps: acquiring a low-dose CT image and a corresponding full-dose CT image, and generating a data set; dividing the data set into a training set and a test set according to a preset proportion; constructing and training a CT denoising network, taking the training set as an input parameter, taking the predicted image as an output parameter, and taking the minimum error absolute value of the full-dose CT image and the predicted image as a loss function; and inputting the test set into the trained CT denoising network in the step S3 to obtain a prediction image. According to the low-dose CT denoising method, global dependence can be efficiently established, and details can be accurately reserved.

The application discloses a texture feature guided texture preserving low dose CT image denoising method, and belongs to the field of medical image processing.The application specifically discloses a multi-scale deep residual attention network model with texture feature guidance, which is applied to low dose CT imaging.The main network model comprises four sub-models, one is a multi-scale initial denoising network model for denoising low dose CT, and the other network is used for extracting texture details after the initial denoising network, and the two network parts work cooperatively.The extracted texture details and the initial low dose CT are fused through a multi-scale image and texture feature fusion network model, and then enter a multi-scale main denoising network for further denoising of the low dose CT, which is beneficial to the main denoising network to learn more unobvious details.The low dose CT image denoising method disclosed by the application efficiently removes the noise and stripe artifacts in the low dose CT image, and meanwhile, the structural information and texture feature detail information in the image are preserved.

The invention relates to an ultra-low dose CT image reconstruction method based on an image domain end. The method comprises the following steps: acquiring high and low dose paired CT images of a plurality of patients, and constructing a training sample set by taking an ultra-low dose CT image in the paired CT images as an input sample and taking a low dose CT image as a label; preliminarily constructing a CT image reconstruction unit and a model training unit, and training the preliminarily constructed CT image reconstruction unit based on the training sample set and the model training unit; the model training unit is used for receiving the reconstructed image and the corresponding low-dose CT image, and performing iterative training on the CT image reconstruction unit based on generator composite loss and discriminator loss fused by various loss functions to obtain a converged CT image reconstruction unit; and inputting an ultra-low dose CT image to be processed into the trained CT image reconstruction unit to obtain a reconstructed CT image. According to the invention, the problems that a CT image reconstruction method in the prior art depends on high-end CT scanning equipment and the image reconstruction quality is poor are solved.

The invention discloses a method, a device, equipment and a medium for constructing pairing data of conventional and low-dose CT (Computed Tomography). The method comprises the following steps of: preprocessing non-paired conventional-dose and low-dose CT images; converting the conventional dose CT image into a sinogram domain to generate a pseudo conventional dose CT sinogram; physical noise is added to the pseudo-conventional dose CT sinogram for degradation, and an original noise sinogram is generated; the construction generator is used for generating a pseudo low-dose CT image after inputting the original noise sinogram; the discriminator based on the high-frequency information extractor discriminates the authenticity of the false low-dose CT image; calculating the joint loss of the adversarial loss and the physical content fidelity loss of the pseudo low-dose CT image, and updating the generator; inputting the conventional dose CT image into a generator to obtain a degraded synthetic low dose CT image; the noise characteristic of the output pseudo low-dose CT image is highly vivid and natural, and the organizational structure content of the image is kept consistent with that of a conventional-dose CT image.

Systems and methods for transforming three-dimensional computed tomography (CT) data into two dimensional images are provided. Such a method is provided including retrieving three-dimensional CT imaging data, where the three-dimensional CT imaging data comprises projection data acquired from a plurality of angles about a central axis. Once the three-dimensional CT imaging data is retrieved, the imaging data is processed as a three-dimensional image and the method proceeds to generate a two-dimensional image by tracing rays from a simulated radiation source outside of the three-dimensional image. The two-dimensional image is then presented to a user as a simulated X-Ray.

The invention discloses a low-dose CT lightweight multi-organ segmentation model. The model comprises an input layer, an encoder module, a cross-layer feature multiplexing module, a decoder module and an output layer which are connected in sequence, the input layer is used for inputting low-dose CT image data; the encoder module is embedded into the cascaded DenseASPP module at the feature extraction end of the backbone network; the cross-layer feature multiplexing module is used for constructing high-resolution and high-semantic compatible feature representation; the decoder module is used for feature refinement; the output layer is used for outputting a multi-organ segmentation result; and the composite loss function calculation module adopts an LDCTloss composite loss function, and the LDCTloss composite loss function is fused with dynamic edge perception and multi-scale confidence weighting. The cascaded DenseASPP module enables a receptive field to be continuously covered by densely connecting cavity convolution with different expansion rates, and effectively eliminates a scale blind area of traditional ASPP. And the cross-layer feature multiplexing module constructs high-resolution-high-semantic compatible feature representation, so that deep semantic drift is greatly inhibited.

The invention discloses an ultra-low dose CT imaging method and application of the ultra-low dose CT imaging method in puncture path guidance, the low dose CT imaging method obtains an artifact-free global image through three steps, and the global image retains key information of a region of interest ROI of an internal scanning image and recovers detail features of a non-region of interest ROI. The scanning view in the operation process is reduced through the internal scanning technology, the radiation dose can be greatly reduced, meanwhile, the convolutional neural network is introduced to extract prior information of primary full scanning to repair a subsequent internal scanning image to be processed, cup-shaped artifacts and metal artifacts are effectively inhibited, the image quality is improved, and the image quality is improved. And detail features of a non-ROI (Region of Interest) in the prior image are recovered, and clearer and more accurate positioning basis is provided for an operator. The radiation risk is reduced, and meanwhile the accuracy of clinical puncture operation is improved.

The invention provides a CT image super-resolution reconstruction method for a sepsis inflammation area, and relates to the field of image processing, and the method specifically comprises the steps: collecting a standard dose CT image of the sepsis inflammation area, and processing the standard dose CT image into a corresponding low dose CT image; performing nonlinear mapping on the low-dose CT image to obtain a tension response value, calculating a tension response product of adjacent pixels to generate a resonance response item, constructing a structure regulation factor based on local window normalization gray variance, and outputting a structure resonance feature map in combination with a nonlinear enhancement item; carrying out multi-order derivative weighting, norm integration and direction offset aggregation on the structure resonance characteristic pattern to obtain a disturbance response characteristic pattern; local contrast mapping is constructed through the disturbance response feature map, an edge response enhancement map is obtained through exponential fractional modulation, the edge response enhancement map and the disturbance response feature map are subjected to weighted fusion, and an edge fusion feature map is output through nonlinear activation; and inputting the edge fusion feature map into an image reconstruction module to improve the resolution so as to complete super-resolution reconstruction.

The invention provides a CT image training data construction method and system based on multi-scale similarity matching. The method comprises the following steps: acquiring a normal dose CT image and a low dose CT image which are not paired; defining a scale number and segmentation points, and performing multi-scale discretization processing on the CT image; wherein a subsection point set is determined by adopting a trainable mapping function based on an unsupervised mode; carrying out pixel-by-pixel subtraction on the discretized CT image under each scale to generate a normalized difference image; calculating a mean value of the difference graph to obtain a single-scale similarity; wherein similarity loss and difference loss are constructed based on a comparative learning thought, and trainable mapping function parameters are optimized; multi-scale similarity scores are weighted and fused, and image pairs with the similarity larger than a threshold value are screened to serve as training data. Automatic discrete point learning is introduced based on a comparative learning thought, and an optimal discrete point division rule is learned from small samples in an unsupervised manner, so that the method adapts to actual distribution characteristics of an image, and a high-quality data set is provided for training a de-noising model.

The application discloses an edge enhancement-based low-dose CT image denoising system and method, and belongs to the field of medical image processing. The system comprises a feature extraction module, an encoder, a decoder and an image reconstruction module. In the encoder and the decoder, each encoding layer and each decoding layer comprises a multi-domain feature module. The multi-domain feature module comprises a local feature modeling unit, a state space modeling unit and a multi-domain feature fusion unit. The local feature modeling unit is used for performing convolution operation on the input feature to obtain a local feature. The state space modeling unit is used for performing feature extraction on the input feature through a state space model to obtain a global feature, and after extracting an edge feature, the two are fused into an edge enhancement global feature. The multi-domain feature fusion unit is used for fusing the local feature and the edge enhancement global feature. The application can improve the noise suppression effect of the low-dose CT image, and significantly improve the structure consistency and edge retention capability of the denoised image.