122 results about "Tissue segmentation" patented technology

The present invention relates to the technical field of stomatology. Provided is an auxiliary dental implant generation method based on a diffusion model. The method in the present invention comprises: acquiring oral CBCT image data of historical patients, preprocessing the oral CBCT image data of the historical patients to obtain a CBCT image dataset, using the CBCT image dataset to train a multi-task segmentation network, and using the segmentation network to obtain an intraoral tissue segmentation result; using the intraoral tissue segmentation result to train detection networks from the three dimensions of a cross-sectional plane, a coronal plane and a sagittal plane, respectively; using the detection networks to obtain detection results in the three directions of the cross-sectional plane, the coronal plane and the sagittal plane; fusing the detection results in the three directions of the cross-sectional plane, the coronal plane and the sagittal plane, and using a majority voting algorithm to construct a three-dimensional bounding box, so as to acquire an edentulous area; and using the intraoral segmentation result and the edentulous area as prompt information to guide, by means of an iterative process, a network to generate a post-implantation effect. The implantation effect obtained by the present invention is highly accurate, thereby providing a more precise auxiliary tool for stomatology.

The invention discloses a tooth intelligent segmentation and parameterized bionic reconstruction method fusing multi-modal data, and belongs to the field of artificial intelligence and computer-aided biomedical engineering. The method comprises a CBCT intelligent segmentation module, an intraoral scanning intelligent segmentation module, an internal tissue segmentation module, a point cloud generation registration module and a parameterized bionic reconstruction module. The point cloud generation registration module further comprises a point cloud generation sub-module and a point cloud registration sub-module; the parameterized bionic reconstruction module further comprises a dentin reconstruction sub-module, a cancellous bone reconstruction sub-module and a gingival reconstruction sub-module. The invention aims to realize tooth semantic segmentation, point cloud reconstruction and parameterized bionic modeling of an organization structure under cross-modal data fusion, and provides high-precision bionic internal structure information for teeth in combination with an artificial intelligence model and parameterized modeling.

The invention discloses an image segmentation method and system, and relates to the technical field of image segmentation. According to the image segmentation method, to-be-segmented laryngoscope medical image data is obtained and input into a pre-trained image segmentation model for preliminary segmentation processing, and a laryngoscope structure segmentation image is generated and comprises a plurality of laryngoscope segmentation areas and corresponding segmentation data sets; performing comprehensive analysis to obtain a structure segmentation perception index of each throat segmentation region, performing analysis to obtain a plurality of low-confidence throat regions, and inputting the low-confidence throat regions into an image restoration model in combination with a segmentation data set of a set annular adjacent throat segmentation region for analysis to obtain an enhanced segmentation data set of each enhanced throat region; according to the method, the laryngoscope structure segmentation image is updated through the enhanced segmentation data set of each enhanced laryngoscope region to obtain the structured laryngoscope tissue segmentation image, so that the enhanced regions have the consistency of pixel levels, and the segmentation precision and clinical availability are effectively guaranteed.

The invention discloses a collaborative optimization method and device for three-dimensional tissue segmentation and registration of a brain nerve image, and the method comprises the steps: S01, constructing a segmentation and registration collaborative model which comprises a shared feature encoder, a segmentation path and a registration path, the segmentation path is used for generating a segmentation probability distribution diagram, and the registration path is used for generating a deformation field; s02, acquiring a training set of the brain three-dimensional magnetic resonance image pair; s03, performing cooperative training on the segmentation and registration cooperative model according to a multi-task cooperative loss function, the loss function including segmentation loss, registration loss and a cooperative regularization term, and the cooperative regularization term modulating a deformation field gradient penalty term by using a multi-scale boundary weight map and a tissue-specific mechanical weight; and S04, receiving an image pair to be registered in real time, and inputting the image pair to be registered into the trained segmentation registration collaborative model to obtain a registration result. According to the method, the calculation efficiency can be remarkably improved while the segmentation and registration precision is ensured.

The invention belongs to the technical field of intelligent diagnosis, and particularly relates to an AI combined MRI and clinical JIA diagnosis system and a storage medium. According to the system disclosed by the invention, the early auxiliary diagnosis of the juvenile idiopathic arthritis is carried out by combining artificial intelligence with multi-dimensional and multi-modal information of multi-sequence MRI images of knee joints of children and various clinical information. The main technology of the method is child knee joint tissue segmentation based on deep learning, a multi-dimensional feature extraction strategy based on a segmentation result, and disease classification based on multi-modal feature integration and deep learning. By integrating the multi-dimensional features of the multi-sequence MRI images and fusing different modal features such as image information and clinical information, an auxiliary diagnosis result with high accuracy can be provided. The technology provided by the invention is beneficial to the realization of early diagnosis and early treatment of juvenile idiopathic arthritis, and has a very good application prospect.

The invention discloses a nerve regulation and control intervention system based on time domain interference electrical stimulation, and the system comprises a magnetic resonance image registration module which receives and registers an imported magnetic resonance image, and obtains a to-be-processed MRI image; the head model tissue segmentation module is used for performing head model tissue segmentation on the to-be-processed MRI image; the stimulation simulation module is used for performing stimulation simulation based on the tissue segmentation structure to obtain a front lead field matrix; the intervention coordinate selection module is used for determining a target brain region corresponding to a target spot brain region coordinate position needing stimulation intervention and putting the target brain region into a running sequence; the electrode position arrangement optimization module is used for verifying, optimizing and finely adjusting the electric field intensity of the target brain region under different electrode arrangement schemes to obtain an electrode configuration parameter result and displaying the electrode configuration parameter result; the parameter importing module is used for receiving the selected electrode configuration parameters and imported preset electrical stimulation parameters; and the electrical stimulation module is used for performing electrical stimulation on the target brain region of the target object. According to the application, the accuracy and the effectiveness of nerve regulation and control treatment can be realized.

The invention discloses a weak supervision pathological image tissue segmentation method based on text prompt learning. The method comprises the steps of feature extraction and initial class activation graph generation; using an MCRM module to optimize the initial class activation graph to obtain a refined class activation graph; and aggregating the plurality of refined class activation graphs to form a fused pseudo mask, taking the fused pseudo mask as a supervision signal, training a segmentation model, and after the training is completed, segmenting the new pathological image tissue by using the segmentation model. According to the method, a text prompt learning mechanism is utilized to focus the model on learning high-discrimination features, so that the influence of tissue co-occurrence is reduced. An initial class activation graph is optimized through a multi-mode class activation graph refining module, and the integrity of boundary segmentation is enhanced. Meanwhile, pseudo masks from different network layers are fused to train a segmentation model, and semantic segmentation of the pathological image is realized. According to the method, high-annotation data dependence is effectively relieved, and the generalization ability of the model is improved, so that application in the field of artificial intelligence-assisted medical treatment is promoted.

The invention discloses a PET / MR (positron emission tomography / magnetic resonance) all-in-one machine-based dynamic attenuation correction and respiratory and cardiac dual motion correction method. The method comprises the following steps of: generating a dynamic mu graph of a respiratory cycle; mR cardiac function movie sequence scanning of the cardiac cycle; accurately segmenting the cardiac muscle; 6, performing tissue segmentation on the dynamic mu graph; clustering the PET data; pET data space-time registration is carried out; merging PET data of the same time phase; performing frame-by-frame dynamic attenuation correction; respiratory movement correction; correcting a partial volume effect; carrying out myocardial MR image space standardization on different time phases of the cardiac cycle; and carrying out spatial standardization on the myocardial PET image of different time phases in the cardiac cycle. According to the method, the mu graph of the respiration and cardiac cycles is reconstructed through time-phase sharing and is accurately registered with the PET data, so that the high-quality attenuation correction of the heart PET image is realized and the motion correction is completed at the same time.

The invention discloses a high-precision and automatic dental crown generation method. The method comprises seven steps of oral cavity three-dimensional data acquisition, full-mouth tooth semantic segmentation, upper and lower jaw point cloud precise registration, abutment and associated gingiva precise segmentation and trimming, abutment surrounding environment multi-tissue segmentation, neck-edge line secondary precise detection, and dental crown generation and multi-dimensional refinement. The method comprises the steps of oral cavity three-dimensional data acquisition, full-mouth tooth semantic segmentation, upper and lower jaw point cloud precise registration, abutment and associated gingiva precise segmentation and trimming, abutment surrounding environment multi-tissue segmentation, neck-edge line secondary precise detection and dental crown generation. Through combination of deep learning and a point cloud algorithm, high-precision segmentation of teeth and gingiva and precise reduction of an occlusion relationship are realized, an oral anatomical feature library and a clinical repair standard are fused, and through multi-dimensional fine adjustment optimization, a dental crown 3D model which is high in fitting degree, harmonious in occlusion and capable of meeting clinical requirements is generated. The method is full-process automatic, greatly improves efficiency, reduces operation threshold, reduces material waste and diagnosis and treatment cost, remarkably improves repair success rate, and is suitable for various oral repair scenes.

The invention discloses a method, device and equipment for generating a head model, a storage medium and a program product. The method is characterized by comprising the following steps of: receiving and verifying original three-dimensional T1 weighted image data and an original three-dimensional grid model; performing tissue segmentation based on the verified three-dimensional T1 weighted image data and the original three-dimensional T1 weighted image data to obtain a cerebral grey matter mask, a cerebral white matter mask, a scalp mask and a skull mask; performing topological structure repair on the brain grey matter mask and the brain white matter mask to obtain a brain mask; respectively converting the scalp mask, the skull mask and the brain mask into a scalp mesh model, a skull mesh model and a brain mesh model; calculating and applying a spatial transformation matrix from the verified three-dimensional grid model to the scalp grid model, and performing spatial registration and fusion on the scalp grid model, the skull grid model and the brain grid model to obtain a head model; the method has the advantages that the efficiency, precision and robustness of head model generation are effectively improved, and the method has good cross-platform deployment capability.

The invention relates to the field of artificial intelligence and medical image processing, in particular to a mammary gland pathology visual model establishing method based on multi-model fusion and combined distillation, which comprises the following steps: performing tissue segmentation and dyeing normalization on a full-slice image; inputting the image blocks into a pre-training teacher model of a plurality of freezing parameters in parallel to extract high-dimensional features, and generating unified enhanced features through a learnable feature fusion network; constructing a student model, and performing end-to-end training by using a joint loss function including feature simulation, logic output distillation and multi-task supervision; and connecting a plurality of task specific prediction heads to the student model, and realizing full-slice-level multi-task diagnosis and treatment prediction through an aggregation strategy. According to the technical scheme, efficient knowledge migration and multi-task cooperation can be achieved, and the accuracy, generalization ability and reasoning efficiency of mammary gland pathology analysis are remarkably improved.

The invention relates to a fetal brain tissue segmentation method based on prior information guidance, and the method comprises the following steps: obtaining a public data set, and selecting a part of data in the public data set as a training set; preprocessing the training set to obtain a grey-white matter volume ratio, an edge feature map and a gestational week feature map; constructing a prior segmentation model GA-Swin, and designing a loss function LTotal to train the model; and finally obtaining a trained fetal brain segmentation model. According to the method, on the basis of following the development law, each tissue of the fetal brain can be precisely segmented, and collaborative optimization of precision interpretability and biological rationality in medical image analysis is achieved.

The invention provides a method for constructing a CT-MRI (Computed Tomography-Magnetic Resonance Imaging) personalized three-dimensional heart model based on multi-modal imaging, which comprises the following steps of: firstly, carrying out myocardial and infarction region segmentation on a CMR-LGE image, and carrying out myocardial and intramyocardial adipose tissue segmentation on a CE-CT image to generate a corresponding Label image; then, a high-precision three-dimensional volume mesh model of the corresponding ventricle is constructed based on the segmentation results of the two types of images; and then, mapping the corresponding Label image with the tissue type to a corresponding finite element model according to the coordinates of the grid, and obtaining a corresponding personalized ventricular model with the tissue type. An interpolation method based on a general ventricular coordinate system is adopted, data of an MR ventricular model with infarct tissue attributes is transferred to a CT ventricular model, and therefore a mixed CT-MRI ventricular model is constructed. And finally, storing the constructed CT-MRI ventricular model data with the integrated tissue attributes as a standard geometric and topological file format.

The invention discloses a sacral plexus and sciatic nerve ultrasound image recognition and puncture guide method and device, and the method comprises the steps: inputting a sacral plexus and sciatic nerve ultrasound image into a target image classification model, and carrying out the classification of the sacral plexus and sciatic nerve ultrasound image, and obtaining the image types of the sacral plexus and sciatic nerve ultrasound image; determining a region category set based on the image categories of the sacral plexus and sciatic nerve ultrasound images; inputting the sacral plexus and sciatic nerve ultrasonic image into a target image segmentation model and carrying out image segmentation on the sacral plexus and sciatic nerve ultrasonic image to obtain a plurality of tissue segmentation regions and region categories of the plurality of tissue segmentation regions on the sacral plexus and sciatic nerve ultrasonic image; and sending prompt information based on the plurality of tissue segmentation regions and the region categories of the plurality of tissue segmentation regions. According to the method, the accuracy of sacral plexus and sciatic nerve ultrasonic image recognition and puncture guiding can be improved.

The invention discloses a pericardium adipose tissue segmentation method, device and equipment based on an MR image and a medium. The method comprises the following steps: acquiring a target heart magnetic resonance image containing pericardium adipose tissue of a target object; segmenting pericardium adipose tissue in the target heart magnetic resonance image based on a tissue segmentation model to obtain target pericardium adipose tissue of the target object; the tissue segmentation model comprises a three-branch cross-domain feature collaborative encoder, a double attention feature fusion module and a dynamic boundary perception decoder, and the three-branch cross-domain feature collaborative encoder performs feature extraction from global, local and frequency domains based on a three-branch structure; the double attention feature fusion module is established based on a space attention mechanism and a parallel channel attention mechanism, and the dynamic boundary perception decoder is used for enhancing the global and boundary perception capability. According to the scheme, the pericardium adipose tissue in the heart MR image can be accurately and efficiently segmented by using the tissue segmentation model based on multi-feature collaboration.

The invention relates to the technical field of medical training, and discloses a lumbar puncture simulation teaching system based on an AR technology, and the system comprises an AR content generation module which collects a lumbar medical image and carries out the tissue segmentation and focus recognition; performing three-dimensional geometric modeling based on the tissue segmentation result and the focus recognition result to generate a lumbar vertebra 3D model; the AR display interaction module develops an interaction function of the lumbar vertebra 3D model by using a Unity platform and AR equipment; interactive operation of the user is monitored, and feedback is conducted in real time; and the operation evaluation module is used for evaluating the interaction operation of the user and assigning an improvement suggestion according to an evaluation result. A three-dimensional visual scene is constructed through the AR technology, a user can observe the needle inserting process from different angles according to the complex anatomical structure of the lumbar vertebra part, and dynamic effects such as cerebrospinal fluid pressure change are simulated; in the lumbar puncture simulation training, the AR system can feed back operation data in real time, such as the needle inserting angle, the puncture strength and the puncture depth, and the user is helped to accurately master operation skills.

The invention discloses a benign and malignant identification and growth prediction system based on pulmonary nodule radiomics, and relates to the technical field of medical image processing. Comprising an acquisition module used for acquiring a pulmonary nodule segmentation mask, a CT value and surrounding tissue segmentation data; the calculation module is used for synchronously calculating a first score representing internal CT value distribution heterogeneity, a second score representing surface complexity and a third score representing blood vessel interaction according to the data; the judgment module is used for triggering a high-risk alarm when the scores of the three items all exceed corresponding threshold values; the prediction module is used for correcting the basic growth model according to the three scores and generating growth prediction data; and the output module outputs an alarm and a prediction result. The system also comprises optimization modules of weight fusion, time sequence processing, growth partitioning and the like. According to the scheme, through multi-dimensional feature fusion and dynamic modeling, more accurate identification of benign and malignant pulmonary nodules and more reliable growth trend prediction are realized, and key support is provided for clinical decision making.

The invention provides an automatic segmentation and region division imaging method and system for ultrasonic velocity estimation, a storage medium and electronic equipment, and the method comprises the steps: dividing an ultrasonic B image into a plurality of tissue regions based on an edge recognition result; obtaining channel data in the ultrasonic imaging process; based on the sound velocity range, performing beam forming according to the channel data; based on the reconstructed image at each wave velocity, according to each tissue area, performing focusing quality evaluation; determining the optimal sound velocity of each tissue area based on the focusing quality evaluation result of each tissue area; based on the optimal sound velocity of each tissue area, performing color coding mapping on each tissue area to obtain a sound velocity color image of each tissue area; and superposing the sound velocity color image of each tissue area with the ultrasonic B image to obtain a final ultrasonic sound velocity image. Through automatic tissue segmentation and region division, in combination with accurate sound velocity estimation and visual sound velocity color coding mapping, the method has a wide application prospect and an important clinical value.

The invention discloses an immediate implantation operation method and system based on augmented reality. The method comprises the steps that CT data of the oral cavity and the head of a patient are acquired and subjected to denoising and normalization processing; the method comprises the following steps: constructing oral digital twins of a patient based on CT data, and completing tissue segmentation, key position marking and gravity center calculation; calculating tooth extraction and implantation paths through operation design software, dynamically predicting stress states of the alveolar bone at different implantation angles and depths based on finite element analysis, and generating an implantation planning scheme; accurately registering the digital twin with the oral cavity of the patient by using multi-point identification and an SLAM (Simultaneous Localization and Mapping) algorithm, and establishing a space mapping relation of an implanting tool; a doctor wears AR glasses to carry out surgery operation, a virtual guiding path and implanting depth and angle information are displayed in real time, and a three-dimensional virtual resistance simulation curved surface is overlaid to prompt bone wall resistance changes; tooth extraction and implantation operations are completed according to AR prompts, stable fusion of the implant and a tooth extraction socket or a bone wall is achieved, and data of the whole operation process are recorded for postoperative analysis.

Segmentation or other classification of digital pathology images with a deep learning model allows for sophisticated spatial features for cancer diagnosis to be extracted in an automated, fast, and accurate manner. A tiered analysis of tissue structure based in part on deep learning methods is provided. First, tissues depicted in a digital pathology image are segmented into cellular compartments (e.g., epithelial and stromal compartments). Second, the heterogeneity in the different cellular compartments are examined based on a clustering algorithm. Tissue can then be characterized in terms of inertia (or other spatial measures or features), which can be used to recognize disease. In some instances, multidimensional inertia (i.e., inertia computed in different cellular compartments or clustered components) can be used as an indicator of disease and its outcome.

The invention discloses an oral cavity image multi-tissue full-automatic segmentation method, which belongs to the technical field of oral cavity image multi-tissue segmentation, combines multi-stage classification with image processing, and improves the accuracy of image recognition, enhances the depth of image understanding and improves the efficiency of image processing through the cooperative operation between the multi-stage classification and the image processing. By means of the multi-stage classifier, images can be classified from local to global and from rough to fine, the classification mode is beneficial to more accurately recognizing target objects or features in the images, in the image processing process, the region of interest can be gradually reduced through multi-stage classification, the target judgment capacity is enhanced in each stage, and the target recognition efficiency is improved. In this way, efficient and accurate target detection is achieved, in some cases, multi-level classification can be combined with other image processing tasks for learning, and the efficiency and accuracy of the model are improved by sharing the feature extractor and the model parameters.

The invention particularly relates to an algorithm based on cervical tissue squamous epithelium lesion diagnosis. The algorithm comprises the following steps of tissue segmentation, region cutting and a classification algorithm. According to the method, multi-step fine processing is carried out on a cervical tissue slice image, for example, in the tissue segmentation stage, gray transformation and Gaussian filtering are adopted to enhance image features, and a cross entropy loss function is used to optimize model parameters; flexible segmentation is carried out according to form contour information during region cutting; in the classification algorithm, image preprocessing and data enhancement are carried out, and a multi-layer neural network is utilized to extract and analyze features, so that squamous epithelium lesion tissues can be identified more accurately, the lesion degree is judged comprehensively, and the diagnosis accuracy is improved.

This invention discloses a kidney pathology image processing method integrating multi-tissue segmentation and quantitative lesion analysis, belonging to the field of medical image processing technology. The method includes the following steps: acquiring and preprocessing PAS-stained whole-slice images of kidney pathology; fine-tuning the segmentation model using an unsupervised domain adaptive strategy to address batch-to-batch staining differences; inputting the preprocessed image into a multi-class semantic segmentation neural network to obtain tissue segmentation results; training the network based on pixel-level annotations, employing a Class-Token mechanism, encoder-decoder architecture, and multi-scale feature fusion, and optimizing the Dice loss and binary cross-entropy loss based on joint weighting of categories and boundaries; performing statistical analysis based on the segmentation results and outputting quantitative analysis results. This invention provides an objective, reproducible, and intelligent auxiliary tool for the accurate assessment and large-scale clinical research of chronic kidney disease.

The present disclosure provides an abnormal tissue growth prediction method, device, electronic equipment and storage medium, the method comprising: obtaining an abnormal tissue growth prediction model; obtaining a historical sample sequence and a target growth duration, the historical sample sequence being arranged in chronological order by M historical samples obtained by examining abnormal tissues; performing a time sequence feature extraction operation; inputting the time sequence features of the historical sample sequence and the Mth historical sample with added noise data into a generative model, outputting post-growth image noise and post-growth abnormal tissue segmentation results; based on a preset denoising formula, using the Mth historical sample with added noise data and post-growth image noise to obtain post-growth image prediction results. In this way, the recurrent neural network for extracting time dimension information is embedded into the generative model for extracting spatial dimension information, improving the performance of the model and making the prediction results of abnormal tissue growth more accurate.

The invention discloses an image segmentation method, device and equipment for pericardium adipose tissue and a medium. The method comprises the following steps: acquiring a heart magnetic resonance image of a target object, wherein the heart magnetic resonance image comprises pericardium adipose tissue; segmenting the pericardium adipose tissue in the heart magnetic resonance image by using a pre-trained tissue segmentation model to obtain a target pericardium adipose tissue of the target object; the tissue segmentation model comprises an encoder, a decoder and a space channel attention module, the encoder comprises a double-branch cooperation module, the decoder comprises a multilayer fusion convolution module, the double-branch cooperation module is used for feature extraction based on a double-branch network, and the space channel attention module is used for feature enhancement in space and channel dimensions. And the multi-layer fusion convolution module is used for feature fusion based on grouped convolution. According to the scheme, semantic information and structural features in the image can be fully mined by using the tissue segmentation model, the segmentation performance in a complex scene is improved, and the pericardium adipose tissue is efficiently and accurately segmented.

The invention relates to the technical field of iconography, in particular to an iconography-based three-dimensional visual model reconstruction method and system, and the method comprises the following steps: carrying out the multi-tissue segmentation in a two-dimensional slice image through employing a Unet-2D algorithm, and obtaining a plurality of tissue segmentation images; voxelization processing is carried out on each tissue segmentation image; distributing a high-contrast color for enhancing the display contrast among the tissues to the three-dimensional volume data of each tissue by combining a multi-objective optimization technology with a Unit-3D algorithm to obtain high-contrast three-dimensional volume data of each tissue; and performing visualization processing on the high-contrast three-dimensional surface model of the liver through a Poisson surface reconstruction technology and a volume rendering technology. According to the method, the three-dimensional reconstruction processing technology is utilized to accurately describe the shape, spatial distribution and the like of the calculus, and through enhanced display processing, when three-dimensional visualization reconstruction is carried out, the hepatic calculus part is highlighted, the calculus is strongly displayed, and the visualization effect is improved.

The application provides a multi-modal tissue segmentation method and a surgical navigation system, and the method comprises the following steps: inputting three-dimensional medical images of multiple modes into corresponding deep learning network models respectively for segmentation to obtain corresponding three-dimensional segmentation results; registering the three-dimensional medical images of each mode to obtain a registration relationship; and according to the registration relationship, combining the three-dimensional medical images of each mode to weight and fuse the three-dimensional segmentation results of each mode according to preset weights of each type of tissue. The application trains a segmentation model for each mode of medical image, each deep learning model has stronger segmentation capability for the medical image of the corresponding mode, and has higher segmentation precision; different types of medical images are respectively designed with preset weights (mode weights) of each type of tissue, and the segmentation results are fused according to the mode weights in the fusion process, so that the tissue structure characteristics are more fully reflected, and the segmentation precision is improved.

The present application relates to the technical field of medical image processing, in particular to a pancreatic lesion CT image positioning method and system based on multi-tissue segmentation, comprising the following steps: acquiring a CT image containing a pancreatic tumor; using a pre-established feature extraction network to perform feature extraction on the CT image to obtain image features; using a region proposal network to generate candidate regions for the image features to obtain pancreatic tumor segmentation candidate regions. The present application uses a global feature extraction network structure and a local feature extraction network structure as a teacher model to guide the training of a lightweight MobileNet network to construct a feature extraction network for pancreatic tumor positioning, which can achieve high-precision positioning performance by aggregating global and local features, that is, it can extract tumor lesions of multiple sizes and improve the lightweight structure to reduce the amount of calculation.

The invention relates to the technical field of medical image processing, in particular to a preoperative image segmentation method and system based on hybrid supervision enhancement, and the method comprises the following steps: constructing an organ tissue segmentation model by using a Unet network, and constructing a vascular tissue segmentation model by using the Unet network; setting a supervision enhancement module, and combining the organ tissue segmentation model and the vascular tissue segmentation model into a teacher model; a student model is constructed by using an FCN network, and a preoperative image segmentation model is obtained through mixed supervised learning of a teacher model and the student model. For preoperative three-dimensional medical image segmentation, a mixed supervision segmentation model is provided, a supervision enhancement mode is adopted when a teacher model is constructed, it can be guaranteed that the teacher model accurately segments each organization, and a lightweight model structure is obtained based on a teacher-student multi-fine-grained feature learning strategy; and a student model capable of precisely segmenting each organization can be realized, so that the segmentation precision and the segmentation efficiency coexist.

The present invention provides a fetal brain tissue segmentation method, device, medium and terminal based on a cycle consistency network. By acquiring target fetal brain image data, inputting the target fetal brain image into a pre-trained cycle consistency network, and outputting a segmentation result of the target fetal brain image corresponding to the gestational week; The present invention adopts a learning framework combining cycle consistency and adversarial learning, and realizes the accurate capture of fine-grained domain-invariant brain structures in thick clinical and thin reconstruction cases of fetal brain MR images.