528 results about "Tumor region" patented technology

The invention relates to the technical field of medical image analysis, in particular to a tumor partition recognition method and system based on multi-scale image analysis, and the method comprises the following steps: based on an input medical tumor partition image, defining the gray value distribution range of pixel blocks, dividing pixel block groups according to a fixed size, and calculating the neighborhood gray values of the pixel blocks, and determining a distribution change set through accumulation of gray difference values of the pixel block groups, and generating a multi-scale gray distribution diagram. According to the invention, by calculating the gray value distribution and the gray difference value of the pixel block group, enhancing the image detail identification of the tumor area, improving the tumor boundary analysis capability, analyzing the texture feature difference of the boundary area, quantifying the gray gradient difference value and the connection characteristic, improving the geometric characteristic analysis precision, and combining the space geometric expansion data and the texture characteristic, the tumor boundary analysis precision is improved. Partitioning precision and image analysis accuracy are improved, misrecognition of a boundary fuzzy region is effectively avoided, and a finer tumor heterogeneity research basis is provided.

The invention provides a tumor area double-view detection method and system based on image modalities and views, and relates to the field of medical image analys.The method comprises the steps that multi-modal image data are collected from the body of a patient, and image data of different view angles at the same position are obtained according to time frames; extracting multi-modal features from the image data of different modalities and different view angles, and performing weighted fusion on the extracted multi-modal features to form a fused feature map; inputting the fused feature map to train and predict the 3D-CNN network so as to identify and position a tumor region, and marking the tumor region by using a mask output by the segmentation model; calculating the pixel number of the tumor area in each layer of image according to the marked mask, and multiplying the pixel number by the volume of each pixel to obtain the volume of the tumor area; and the segmentation results of each layer of image are superposed, and a three-dimensional model of a tumor region is generated by using 3D reconstruction, so that the limitation of single-mode and single-view image data in tumor detection is overcome.

The invention relates to a tumor image segmentation method. Performing multi-scale feature extraction on the image by using a convolutional neural network (CNN) to capture different-scale information of the tumor; introducing a self-attention mechanism to enhance the attention capability of the model on the tumor area and the accuracy of feature extraction; performing image segmentation through the improved model based on the U-Net architecture; according to the method, the advantages of a CNN and a VisionTransform are combined, local and global features are extracted by using a hybrid architecture, and position codes are added into the Transform so as to enhance the perception capability of spatial position information; adopting a comparative learning strategy to improve the discrimination capability of the model by comparing positive and negative sample pairs, and performing multi-task learning in combination with a tumor classification task and a segmentation task; designing a combination of a weighted cross entropy loss function and a Dice loss function, and improving the sensitivity to a small tumor region; and performing post-processing by using a conditional random field (CRF) to optimize a segmentation boundary, and eliminating small noise points and filling holes of a segmentation region by applying morphological operation to further optimize a segmentation result.

The invention provides a multi-modal fusion method for breast cancer prognosis prediction based on deep learning. The multi-modal fusion method comprises the following steps: preprocessing a data set, marking a tissue region of a full-slice pathological image, performing slice processing, extracting multi-scale features, re-sampling a medical image, and extracting features of a tumor region; a two-dimensional region fusion module and a three-dimensional region fusion module are respectively selected to carry out multi-scale feature fusion on the preprocessed full-slice pathological image and the preprocessed medical image; performing multi-region feature fusion on the medical image by using a three-dimensional region fusion module; aligning the multi-modal data, performing cross-modal feature fusion by using self-attention pooling, and capturing cross-modal interaction information; and predicting the disease-free lifetime and the total survival of the patient through the loss function. The problems of limitation of single-mode analysis, unsatisfactory prognosis effect and the like are solved, and more accurate breast cancer prognosis prediction is realized.

The invention relates to the technical field of medical image processing, in particular to a multimodal-based digestive tract tumor lesion image segmentation method and segmentation system. The method comprises the following steps: acquiring an alimentary canal tumor lesion image, and extracting an alimentary canal tumor ultrasonic image; evaluating the tumor invasion depth based on the digestive tract tumor ultrasonic image; performing focus three-dimensional visual modeling according to the tumor invasion depth to obtain a digestive tract tumor focus model; extracting a tumor tissue pathological image according to the digestive tract tumor lesion image; performing tumor region segmentation based on the tumor tissue pathological image to obtain digestive tract tumor region data; performing tissue arrangement anomaly detection based on the digestive tract tumor region data to obtain tissue arrangement anomaly data; and calculating a tissue arrangement disorder index according to the tissue arrangement abnormal data and the digestive tract tumor area data. According to the invention, the tumor identification accuracy and the malignant region segmentation precision are improved based on the medical image processing technology.

The invention relates to the technical field of medical image processing, in particular to an efficient detection method, system and equipment for a craniopharyngeal tubuloma lesion, and the method comprises the steps: obtaining a plurality of continuous images, carrying out the threshold segmentation, obtaining a tissue region and a suspected tumor region, obtaining the edge feature value of each image according to the shape feature of the tissue region, and obtaining the edge feature value of each image; the extrusion degree of each image is obtained according to the position difference of adjacent tissue areas corresponding to suspected tumor areas in adjacent images, so that the same area probability of two adjacent suspected disease image groups, a tumor feature expression group and a craniopharyngeal tube tumor image set are obtained; according to the change trend of the extrusion degree of the tumor feature expression group in the craniopharyngeal tube tumor image set, obtaining the probability of the craniopharyngeal tube tumor area, screening the images in the craniopharyngeal tube tumor image set, and completing the detection of the craniopharyngeal tube tumor lesion area. According to the method, the image containing the craniopharyngeal tubuloma focus is obtained by analyzing the multiple MRI images, and the image containing the craniopharyngeal tubuloma focus is obtained.

The invention relates to the field of brain tumor analysis, and discloses a brain tumor image analysis system based on artificial intelligence, comprising: a spatial alignment unit for acquiring original image data of the brain of a subject; performing spatial alignment on the original image data according to a cross-modal registration algorithm to obtain standardized image data; the feature extraction unit is used for performing tumor region initial segmentation on the standardized image data according to a three-dimensional convolutional neural network so as to obtain a coarse segmentation probability graph; and extracting three-dimensional geometric feature parameters of the tumor candidate region according to the coarse segmentation probability graph. According to the method, the original image data is spatially aligned through the cross-modal registration algorithm, and the spatial consistency between different image sources is ensured, so that the image data under different modals can be accurately compared and analyzed, and an accurate spatial reference is provided for subsequent tumor region identification and processing.

The invention discloses an ovarian tumor cross malignancy classification system based on a hybrid radiomics model. On one hand, a backbone network in the Med3D network is used as a depth feature extractor to construct a target network, and multi-modal 3D magnetic resonance image data is used to carry out fine tuning training on the target network to obtain a corresponding ovarian tumor classification network for obtaining depth features and tumor classification results. On the other hand, omics characteristics of the tumor region of interest are extracted, and a tumor classification result corresponding to the omics prediction model is obtained. In addition, the depth features and the omics features are fused, and a corresponding classification result is obtained through a KNN classification algorithm. And finally, obtaining a final ovarian tumor cross malignancy classification result according to weighted average of the three classification results. According to the invention, by integrating various models and algorithms, more comprehensive feature mining and analysis are carried out on the multi-modal MRI image data of the ovarian tumor patient, so that the final ovarian tumor classification effect is improved.

The invention provides a bladder cancer prognosis prediction method and system, and relates to the technical field of intelligent detection, and the method specifically comprises the steps: obtaining and preprocessing a bladder cancer hematoxylin-eosin (HE) staining pathological image to obtain an effective color block, and marking to obtain a tumor and necrosis training set; respectively training the two initial models based on the tumor training set and the necrosis training set to correspondingly obtain a tumor model and a necrosis model; acquiring a to-be-predicted bladder cancer pathological image, and performing tumor and necrosis region identification based on the tumor model and the necrosis model to correspondingly obtain a tumor region and a necrosis region; based on the space coordinate information of the tumor area and the necrotic area, carrying out numerical calculation to obtain a space distribution index of the necrotic area relative to the tumor area; and carrying out survival analysis based on the spatial distribution index to obtain survival prognosis information of the bladder cancer patient. The deep learning technology is applied to the prognosis direction of the bladder cancer, a clinician is assisted in evaluating survival prognosis of a patient, and a reference basis is provided for diagnosis and treatment decisions.

The invention discloses a device for tracking a tumor position in real time in radiotherapy, and solves the problems of delay and inaccuracy in a traditional tumor position prediction method through a multi-model fusion method combining a deep learning model and a mathematical prediction model. Different from a traditional prediction method depending on an in-vitro marker or iconography technology, the tumor position is predicted in real time by adopting a long-short-term memory network, and the change of the tumor position along with respiratory movement is accurately captured by utilizing the strong time sequence modeling capability of the long-short-term memory network. By introducing the mathematical prediction model to compensate the future position of the tumor, the method can effectively solve the prediction error caused by system delay, and ensures that rays can accurately irradiate the tumor area in dynamic tumor tracking radiotherapy, thereby improving the accuracy and treatment effect of radiotherapy.

The invention discloses a cascade multi-scale convolution and modal enhancement brain tumor segmentation method based on a Mama framework, and relates to the technical field of brain tumors. According to the method, an MCME-UNet model integrated with an MCMS module is provided, through a hierarchical feature extraction mechanism and a multi-modal feature fusion strategy optimized by an MEM module, the segmentation precision is remarkably improved while the calculation efficiency is kept, and particularly, an EEM module innovatively applies Sobel operator to be connected with residual errors, so that the tumor boundary continuity and the segmentation accuracy are synchronously improved; moreover, a novel training normal form of a focus Tversky loss function is introduced, so that the problem of class imbalance is effectively solved, the sensitivity of the model to a small tumor region is enhanced, a smoother segmentation boundary can be generated, and the stability and accuracy of a segmentation result are greatly improved.

The invention relates to the technical field of tumor interventional operations, and discloses a visual navigation method based on a tumor interventional operation robot. The method comprises the following steps: acquiring real-time medical image data of a tumor area containing multi-modal imaging information so as to comprehensively present anatomical details; and performing three-dimensional reconstruction on the image data to generate a tumor area three-dimensional anatomical structure model capable of visually displaying a space structure. Key anatomical feature points are extracted based on the model, space coordinates are calculated, a surgical robot intervention path is planned according to the coordinates, and an initial navigation track is generated; and continuously collecting real-time pose data of the robot in an operation, dynamically matching the real-time pose data with the initial navigation trajectory, adjusting motion parameters according to a matching result, and generating a corrected navigation instruction. The method can reflect the intraoperative anatomy condition in real time, dynamically optimize the path, solve the problems that traditional navigation depends on preoperative static images and lacks real-time adjustment, reduce operative complications and improve the treatment effect of patients.

The invention discloses a tumor prognosis prediction method and a tumor prognosis prediction system, which are used for constructing a multi-modal fusion model based on image-pathology to improve the prognosis prediction efficiency of tumors, especially pancreatic cancer, and providing reference information for clinical decision-making. According to the technical scheme, the method comprises the following steps: S1, preprocessing an original tumor enhanced CT image, segmenting a tumor region, extracting radiomics features and depth image features of the tumor region, and establishing a CT image feature set; s2, after feature preprocessing is carried out on the tumor clinical data, clinical features with statistical significance are screened out, and a clinical feature set is established; s3, carrying out Hamp; e, preprocessing the pathological image, segmenting a tissue region, extracting spatial relation features, and generating a pathological spatial feature set; and S4, based on a feature interaction method, carrying out multi-modal fusion on the CT image features, the clinical features and the pathological spatial features, inputting a full-connection neural network, constructing a tumor survival risk prediction model, and outputting a tumor survival risk probability through the tumor survival risk prediction model.

The invention discloses a bioelectrical impedance tumor detection method based on pattern recognition, and the method comprises the following steps: S1, carrying out the electrical impedance measurement of a to-be-detected tumor region through a bioelectrical impedance analysis instrument, and obtaining the current, voltage and impedance values at different depths and frequencies; s2, designing a feature extraction module, and extracting key features in the electrical impedance data of each channel under different frequencies; s3, inputting the extracted key features into a multi-level feature fusion module, and fusing the electrical impedance features under different channels and frequencies to form new fused features; and S4, inputting the fusion features into a self-adaptive classification module, and adjusting model parameters in combination with a self-adaptive optimization method. According to the invention, automatic processing and classification are carried out on data through a pattern recognition algorithm, and the accuracy, sensitivity and reliability of tumor detection can be effectively improved in combination with the feature extraction module, the multi-level feature fusion module and the adaptive classification module.

A medical image processing apparatus and a medical image processing method that can improve extraction accuracy of a tumor region included in a diagnosis target image are provided. The medical image processing apparatus configured to extract a predetermined region from a diagnosis target image includes: an organ extraction unit configured to extract an organ region from the diagnosis target image; and a tumor extraction unit generated by executing machine learning using a known tumor region included in each medical image group as teacher data and using an organ region extracted from the medical image group and the medical image group as input data. The tumor extraction unit is configured to extract a tumor region from the diagnosis target image using the organ region extracted from the diagnosis target image by the organ extraction unit.

The invention belongs to the technical field of medical imaging, and discloses a tumor area rapid identification method and system based on cortical electroencephalogram signals, and the method comprises the steps: carrying out the preprocessing of collected multi-channel ECoG signals through an interpolation filling strategy; extracting a target frequency band ECoG signal of each channel; extracting a feature sequence of each channel in a given time period as a channel feature; performing clustering analysis on each channel feature to obtain a plurality of categories; and reordering according to the feature level of each category of region, and completing tumor region boundary identification by taking the category of which the feature level meets the requirement as a tumor region. According to the invention, rapid detection and effective identification of tumor boundaries can be realized.

The invention discloses a tumor image segmentation method and system of a multi-scale feature fusion network based on boundary enhancement, and relates to the technical field of image segmentation, and the method comprises the steps: obtaining a to-be-segmented tumor image; constructing a tumor image segmentation model based on a pyramid visual converter PVTv2 backbone network; training the tumor image segmentation model through the to-be-segmented tumor image and the known tumor image to obtain an optimal tumor image segmentation model; and acquiring a real-time to-be-segmented tumor image and inputting the real-time to-be-segmented tumor image into the optimal tumor image segmentation model to obtain a tumor image segmentation result. Aiming at the endoscope image segmentation of the kidney tumor, the kidney tumor in the endoscope image is efficiently, stably and automatically segmented through the boundary-enhanced multi-scale feature fusion network, clinical doctors are helped to provide accurate tumor area positioning in endoscopy and surgical operations, and compared with the most advanced method, the method has the advantages that the accuracy is high, and the efficiency is high. According to the method, better segmentation capability and stronger generalization capability are obtained.

Described herein are systems and methods for positioning a radiation source with respect to one or more regions of interest in a coordinate system. Such systems and methods may be used in emission guided radiation therapy (EGRT) for the localized delivery of radiation to one or more patient tumor regions. These systems comprise a gantry movable about a patient area, where a plurality of positron emission detectors, a radiation source are arranged movably on the gantry, and a controller. The controller is configured to identify a coincident positron annihilation emission path and to position the radiation source to apply a radiation beam along the identified emission path. The systems and methods described herein can be used alone or in conjunction with surgery, chemotherapy, and / or brachytherapy for the treatment of tumors.

The invention discloses a tumor growth prediction and segmentation method based on an adaptive multilayer convolutional neural network. The method comprises the following steps: S1, acquiring a medical image data set containing a tumor region; s2, preprocessing the medical image data set; s3, constructing an adaptive multilayer convolutional neural network model; s4, inputting the preprocessed medical image data set into a self-adaptive multi-layer convolutional neural network model; s5, automatically adjusting the size, the step length and the number of layers of a convolution kernel by the convolution layer; s6, performing supervised learning training on the adaptive multilayer convolutional neural network model through a Dice loss function; s7, outputting a tumor boundary; s8, predicting the growth direction and position of the tumor at the future time point and the possible expansion range of the tumor; s9, dynamically adjusting the convolution kernel size and step length of each layer and weight distribution of an attention mechanism according to a feedback mechanism; and S10, outputting a tumor segmentation result and a growth trend prediction result. According to the invention, the precision and robustness of tumor segmentation and growth trend prediction are improved.

The invention discloses a brain multi-modal multi-sequence data registration method and device based on deep learning. The method comprises the steps of performing first iteration processing on a target image and a first moving image to obtain a first deformation field, wherein the first iteration processing comprises image alignment constraint processing of the target image, smooth constraint processing of the first deformation field, and area alignment constraint processing of a tumor area in the image; registering the first moving image to the target image based on the first deformation field to obtain a second moving image; performing second iteration processing on the target image and the second moving image to obtain a second deformation field; and registering the second moving image to the target image based on the second deformation field to obtain a registered moving image. In the process of performing unsupervised training on the deep learning model, after multiple times of iterative processing, model parameters are updated based on image alignment constraint loss with a target image, deformation field smooth constraint loss and region alignment constraint loss of a tumor region in the image.

The invention discloses an HER2 state recognition method and system based on an IHC and HE bimodal image, and relates to the medical image processing and analysis technology, and the method comprises the steps: carrying out the preprocessing of a full-slice image WSI containing IHC and HE, and cutting the WSI into designated image blocks; constructing a multi-task learning framework to classify image blocks obtained by cutting the IHC image, and determining HER2 expression intensity in the tumor region; calculating the area ratio of the image block corresponding to each grade in the tumor region to calculate an IHC HER2 score; taking each image block of the cut HE image as the input of a Prov-GigaPath model, and aggregating the multi-dimensional features of each image block by using an ABMIL model, and taking the aggregated multi-dimensional features as an HE HER2 score; and converting the IHC HER2 score and the HE HER2 score into one-hot codes, and outputting an HER2 judgment result by using a logistic regression model. The HER2 state interpretation result can be quickly output, a pathologist is assisted in diagnosis, and the diagnosis efficiency and accuracy are improved.

The embodiment of the invention provides a tumor radiotherapy reaction prediction method and system based on multi-instance learning. The method comprises the following steps: constructing a radiotherapy image data set; extracting multi-dimensional features from the radiotherapy image data set and combining and splicing the multi-dimensional features to obtain embedded representation of the multi-dimensional features; the multi-dimensional features comprise two-dimensional convolution features and three-dimensional convolution features; a multi-instance learning model based on a Transform is constructed; taking the embedded representation of the multi-dimensional features as model input, and performing iterative training on the multi-instance learning model to obtain a trained multi-instance learning model; and predicting the radiotherapy metaphase reaction of the patient by using the trained multi-instance learning model, analyzing and evaluating the model prediction process in combination with interpretable analysis, and performing model evaluation on the prediction result of the tumor area of the patient. According to the method provided by the invention, the reaction condition after tumor radiotherapy can be predicted, so that a doctor can be intelligently assisted to carry out precise radiotherapy decision on a patient, and the prognosis risk and harm of the patient are reduced.

The invention discloses a tumor ablation puncture operation planning method and system, and relates to the technical field of medical image processing, and the method comprises the steps: obtaining multi-modal image data of a tumor region, carrying out the preprocessing, carrying out the fusion processing of the preprocessed multi-modal image data through an image fusion algorithm, generating a fusion image, and carrying out the planning of a tumor ablation puncture operation based on the fusion image. A tumor area is automatically marked by using an image segmentation technology, a doctor manually corrects a segmentation result, an optimal puncture path is selected through design and simulation according to a fusion image with a tumor mark, a model and a tumor entity position are accurately aligned by using a space positioning technology based on the fusion image and augmented reality equipment, and an AR view is generated. And under the guidance of the AR view, the puncture needle carries out ablation operation according to the optimal puncture path. Through high-quality images and three-dimensional guidance, high precision and visualization of the whole process from image acquisition to operation implementation are achieved, and the safety and success rate of the operation are remarkably improved.

The invention discloses an esophageal cancer neoadjuvant chemotherapy and post-immunization pathology complete remission prediction system and method, and relates to the field of medical image processing, and the system comprises an image processing module which is used for carrying out the standardized processing of an input image, and segmenting a tumor region in the image; the feature extraction module is used for extracting voxel-level radiomics features from the tumor area; the feature screening module is used for screening out a plurality of voxel-level radiomics features with the most predictive capability; and the pathology prediction module is used for generating a comprehensive feature vector according to the segmented input image and the screened voxel-level radiomics features, and predicting the curative effect of the combination of neoadjuvant chemotherapy and immunotherapy according to the comprehensive feature vector. According to the technical scheme provided by the invention, the accuracy of predicting complete pathologic remission after neoadjuvant chemotherapy and immunotherapy of esophageal cancer can be improved, the defect that the heterogeneity in tumors cannot be considered in a traditional imaging omics technology is overcome, and the method has a relatively high application value.

The invention provides a sub-region acquisition method of a target tumor and a boron drug concentration spatial-temporal distribution acquisition method, and the method comprises the steps: obtaining the regional distribution of a boron drug enrichment region, a tumor region and a boron drug transportation region corresponding to the target tumor, so as to divide the region occupied by the target tumor into a first sub-region, a second sub-region and a third sub-region; and constructing a boron drug transport model based on the boron drug metabolism characteristics corresponding to the different sub-regions, so as to obtain boron drug concentration spatial and temporal distribution in the target tumor based on the boron drug transport model in combination with the boron drug initial concentration of the region occupied by the target tumor, according to the method provided by the invention, the metabolism condition of the boron drug in the area occupied by the target tumor can be accurately reflected, so that the spatial and temporal distribution of the boron drug concentration in the radiotherapy process can be accurately and conveniently extracted, and a more objective and accurate data basis is provided for subsequent data processing and / or analysis calculation.

The invention discloses a tumor and target region synchronous segmentation method and system based on visual language guidance. The method comprises the following steps: constructing a multi-modal data set fusing a medical image and a clinical text; a unified encoder is adopted to extract shared visual features, a double-branch decoder is adopted to predict a tumor region (GTV) and a clinical target region (CTV) respectively, and information sharing and task differentiation modeling are achieved; introducing a medical pre-training large language model, and respectively extracting language features related to GTV and CTV; a visual language collaborative attention module is designed, dynamic fusion and alignment of visual languages are carried out, task related semantics are guided and enhanced, and redundant information is inhibited; and outputting a precise synchronous segmentation result of the GTV and the CTV after fusing the multi-modal features. Compared with an existing scheme, the problems that the boundary of a tumor target region in a medical image is fuzzy and multi-modal information is insufficient in utilization are effectively solved, the segmentation precision and robustness are remarkably improved, and reliable support is provided for clinical radiotherapy planning.

The invention discloses a medical image weak supervision variation segmentation method based on single-point labeling, and relates to the technical field of tumor segmentation in medical images. The method comprises the following steps: firstly, randomly taking a point in a target tumor area, and constructing a new anisotropic Riemannian metric according to texture direction information of a whole CT image, local direction information of gray scale at a boundary and heterogeneity of the gray scale; secondly, calculating two distances from all points in the whole CT image to the points taken in the tumor area: geodesic distance and Euclidean distance based on new Riemannian metric; then, the two kinds of point-by-point distance information are embedded into a Chan-Vese model; and finally, proving the existence and uniqueness of a model solution, and solving the model solution by adopting a Douglas-Rachford algorithm to obtain a final tumor segmentation result. According to the method, a more accurate, more effective and more robust segmentation result can be obtained, the method can be used for tumor detection, positioning and segmentation, the workload of manual labeling can be greatly reduced, and the method has high clinical application value.

According to the saliency reference and decoupling esophageal tumor CT image diffusion segmentation method provided by the invention, the accuracy of tumor segmentation is improved by combining a saliency pseudo tag and a diffusion denoising network. The method comprises the following steps: firstly, preprocessing an esophageal tumor CT image, generating a significant pseudo tag, and enhancing the attention of a network to a tumor region; and then, constructing a saliency reference segmentation network and a diffusion denoising network, and gradually carrying out noise adding and denoising to restore the image by adopting a conditional diffusion model. A multi-task learning framework for decoupling segmentation and de-noising tasks is designed, and collaborative optimization between tasks is ensured through consistency constraint. Through joint optimization of cross entropy loss, Dice loss, mean square error and consistency loss, the segmentation performance is improved. Experimental results show that the method can effectively improve the segmentation precision of the esophageal tumor, and has high clinical application potential.

The invention discloses a multi-task collaborative bone tumor CT image segmentation method, system and device, and relates to the technical field of bone tumor CT image segmentation. According to the multi-task collaborative bone tumor CT image segmentation method, a bone tumor CT image segmentation model is adopted to carry out bone tumor segmentation on a target bone tumor CT image; the bone tumor CT image segmentation model is trained and obtained through the following method: obtaining an original bone tumor CT image; according to the original bone tumor CT image, screening through a CT value range to obtain a bone image and a soft tissue image; marking a tumor area and a skeleton area in the bone image and the soft tissue image to form a training sample; constructing a multi-task learning network including a skeleton segmentation task and a tumor segmentation task; and training the multi-task learning network based on at least one group of training samples to obtain a bone tumor CT image segmentation model. According to the invention, the bone tumor can be identified more accurately, and the precision of bone tumor segmentation is improved.

The invention relates to the technical field of medical image processing, in particular to a tumor image classification method and system based on deep learning, and the method comprises the following steps: obtaining a tumor image gray-scale map, positioning an edge, recognizing a closed structure region, dividing tumor region image blocks, calling a standard image block, analyzing the contour and texture offset degree, and carrying out the layering comparison of labels. And reversely marking image blocks, analyzing edges and consistency of a detail layer, a texture layer and a structure layer, constructing a collaborative matrix to carry out multi-task mapping, and outputting a tumor classification set. According to the method, a closed structure area is positioned and extracted through gray boundary change, the initial recognition precision is enhanced, non-structural interference is reduced, a standard image block is mapped to realize morphological comparison, the classification distinguishing force is improved, label collection prediction errors are compared in a layered manner, error saliency image blocks are dynamically labeled, and a more sensitive feedback mechanism is realized; the multi-layer division enhances the hierarchical expression of the features, enhances the multi-scale aggregation capability, and improves the classification accuracy and the structure recognition generalization capability.