50 results about "Tumor heterogeneity" patented technology

The invention discloses a multi-scale brain tumor segmentation method and system based on adaptive KAN, and a storage medium. The method comprises the following steps: preprocessing three-dimensional brain magnetic resonance imaging data; the preprocessed data are input into an encoder, the encoder comprises a plurality of levels, and each level executes convolution operation to extract local features, executes spatial KAN processing to extract spatial features and downsamples a feature map; the output of the encoder is input into a bottleneck layer, and the bottleneck layer captures a multi-scale global context by using a plurality of parallel expansion convolution branches; the output of the bottleneck layer is input into a decoder, the decoder comprises a plurality of stages, and each stage executes transposing a convolution up-sampling feature map, executes cross-scale gating processing to fuse encoder jump connection features and decoder features, and executes spatial KAN processing to optimize features; and the output of the decoder is input into the output module. According to the method, the problems of low calculation efficiency, poor tumor heterogeneity adaptation, insufficient multi-scale context capture and the like in the existing brain tumor segmentation can be effectively solved.

The invention provides a breast cancer recurrence risk prediction method, system and device based on an ultrasonic image, and relates to the field of intelligent medical treatment, the method uses a deep convolutional neural network to perform deep network feature extraction on a breast ultrasonic image, and uses a deep learning semantic segmentation algorithm to perform accurate positioning and automatic segmentation on a breast tumor region of interest, thereby improving the accuracy of breast cancer recurrence risk prediction. Meanwhile, habitat analysis is carried out on the ultrasonic images to extract tumor heterogeneity features, multi-level and multi-mode features such as deep learning features, radiomics features and habitat analysis features are fused, a breast cancer recurrence risk prediction model is constructed, breast cancer recurrence risk prediction is carried out, and a breast cancer recurrence risk assessment result is output. And a quantitative basis is provided for clinical treatment decisions. The accuracy and robustness of recurrence risk prediction are remarkably improved through multi-feature fusion, and standardization and objectification of breast cancer prognosis evaluation are achieved. In addition, the method further has the advantages of being easy and convenient to operate, low in cost, noninvasive, nonradiative, good in repeatability and the like.

The invention provides a full-slice image classification method and system based on multi-branch attention and random instance masks and a medium, and the method comprises the steps: obtaining a full-slice digital pathological image, carrying out the image segmentation, and obtaining a plurality of instances; performing feature extraction based on a feature extraction network to obtain an instance feature sequence; carrying out parallel analysis on the instance feature sequence based on a multi-branch attention mechanism, executing a random Top-K instance mask operation, and carrying out weighted summation on the instance feature sequence based on the attention distribution of each attention branch; aggregating the packet level feature representations of all the attention branches, generating a comprehensive feature representation of the full-slice digital pathological image, and obtaining a classification prediction result; by setting a plurality of parallel attention branches, different branches are promoted to actively learn and capture a plurality of different visual modes existing in the full-slice digital pathological image, so that tumor heterogeneity can be effectively represented, and the generalization ability of full-slice digital pathological image classification is improved.

The invention discloses a rectal cancer postoperative recurrence risk prediction system and method based on multi-modal time sequence data, and relates to the technical field of medical artificial intelligence. The system comprises a data acquisition and preprocessing module, a feature extraction module and a multi-modal feature fusion and modeling module. The method comprises the following steps: constructing a cross-modal data set containing time sequence clinical data, a time sequence MR image and a biopsy digital pathological image; respectively extracting clinical features, radiomics and deep learning features of the MR image, and nucleus morphology and spatial distribution features of the pathological image; and fusing all the features by using a Transform network, and constructing a prediction model. According to the method, macroscopic images, micropathology and dynamic time sequence information are integrated, tumor heterogeneity is comprehensively quantified, the problem that prediction of a single-mode static model is not accurate is solved, the postoperative recurrence risk of the stage III rectal cancer patient can be evaluated more accurately, and clinical treatment decision making is assisted.

The invention discloses an angle-adjustable interventional biopsy needle suitable for deep tumors and a sampling method of the angle-adjustable interventional biopsy needle, and relates to the field of tumor interventional biopsy sampling, the angle-adjustable interventional biopsy needle comprises a guide sheathing canal, the guide sheathing canal is composed of an outer sheath layer on the outermost layer, a braid layer in the middle and a first lining layer on the inner layer, and a bending mechanism is arranged at the bottom of the guide sheathing canal; the bending mechanism comprises an active bending tube, the active bending tube is fixedly connected with the bottom of the guiding sheath tube, a protection pad is arranged at the bottom of the active bending tube, and a second lining layer is arranged on the inner side of the active bending tube. On the premise of single puncture, multi-point and multi-angle sampling of different quadrants and different depths in a tumor can be easily achieved, representative tissue samples can be obtained to the maximum extent through the design, sampling errors caused by tumor heterogeneity are effectively avoided, the occurrence rate of false negative results is remarkably reduced, and the sampling efficiency is improved. And a more reliable pathological basis is provided for clinical diagnosis.

The invention provides a model for predicting lung adenocarcinoma prognosis and immunotherapy response based on copper death related LncRNAs, and belongs to the technical field of bioinformatics. According to the prediction model provided by the invention, the accuracy and the stability of prognosis judgment of a lung adenocarcinoma patient can be remarkably improved, and the defects of a traditional staging system in the aspects of reflecting tumor heterogeneity and individual treatment reaction difference are effectively overcome; the model not only can realize reliable individual risk grading, but also can further evaluate the tumor immune microenvironment state and predict the potential reaction of a patient to immunotherapy, so that a powerful auxiliary tool is provided for clinically formulating an accurate treatment strategy, in particular to the application decision of an immune checkpoint inhibitor; and finally, the method has important practical application value for improving treatment selection and life quality of patients.

The invention discloses an interactive lattice radiotherapy planning method and system based on a functional image, and relates to the technical field of radiotherapy, and the method comprises the steps: obtaining a radiotherapy positioning CT image of a patient and a functional metabolism image after the space registration with the radiotherapy positioning CT image; based on the radiotherapy positioning CT image and a preset safety boundary parameter, determining a candidate area of lattice layout in the tumor target area; for each candidate position in the candidate region, adaptively calculating lattice parameters for the position based on the corresponding SUV value of the position in the functional metabolism image; determining a final lattice center set from the candidate positions through an iterative screening algorithm on the basis of the calculated lattice parameters of the candidate positions and preset geometric constraints and biological constraints, and generating a lattice target region set on the basis of the final lattice centers and the lattice diameters corresponding to the final lattice centers; and outputting the lattice target region set for dose calculation of a radiotherapy planning system. Individualized dosage improvement really based on tumor internal heterogeneity is achieved.

The application belongs to the technical field of biological medicine, and discloses an in-situ vaccine type mRNA-tLNP preparation for solid tumor treatment and preparation and application thereof. The preparation is designed by double targets, so that CAR-T kills more tumor cells to overcome tumor heterogeneity, changes the tumor immune microenvironment through autocrine fusion proteins (anti-PD-1 scFv and TGFbetaRII, IL-15 and Flt3L, CD40L), promotes T cell proliferation, and recruits and activates dendritic cells, cross-presents new antigens in the tumor local part, and produces in-situ vaccination effects; circular RNA encoding the above proteins is prepared, and the circular RNA is wrapped by lipid nanoparticles modified by CD3 antibodies, so as to be directly delivered to the body, and a significant solid tumor treatment effect is produced.

The present application relates to medical image analysis and brain network research technical field, specifically to glioma induced brain network remodeling visualization and quantitative analysis method based on fMRI. The method comprises obtaining patient fMRI and structural MRI data and preprocessing, excluding tumor area by lesion mask registration strategy, reducing quality effect interference; dividing tumor core area, peritumoral abnormal area and normal brain area; registering Yeo-17 network template to individual brain area to realize mapping; defining tumor core area as independent network unit, and 17 normal networks to form a new set; calculating whole brain voxel and network functional connection strength, and determining functional connection voxel according to threshold; quantifying intratumoral function proportion RIFR and peritumoral connection proportion RPTR, and generating visualization atlas. The present application accurately maps individual brain function network, overcomes tumor heterogeneity interference, provides repeatable quantitative index, and provides reliable imaging analysis tool for brain glioma function protection and clinical research.

The invention relates to a method and device for predicting the curative effect of an image heterogeneity region fusion technology based on a graph network, and belongs to the technical field of intelligent medical treatment and medical image analysis. The method comprises the following steps: firstly, dividing a tumor and a region around the tumor in a medical image into a plurality of non-connected sub-regions as graph nodes, and generating a comprehensive feature representation fusing a local image feature and a node type embedding vector for each node; then, defining edges between nodes based on a spatial distance threshold value, and constructing a graph structure including intra-tumor, peritumor and cross-boundary connection; and finally, inputting the graph structure into a graph neural network model for training to obtain a curative effect prediction model. According to the method, the complex spatial topological relation between the tumor heterogeneity areas is explicitly modeled through the graph structure, the core problems of spatial information loss and insufficient heterogeneity modeling in the prior art are solved, and the accuracy and robustness of treatment effect prediction are remarkably improved.

The invention relates to the technical field of medical image analysis and brain network research, in particular to a glioma induced brain network remodeling visualization and quantitative analysis method based on fMRI. The method comprises the steps of obtaining and preprocessing fMRI and structural MRI data of a patient, eliminating a tumor area through a focus shielding registration strategy, and reducing mass effect interference; dividing a tumor core region, a peritumoral abnormal region and a normal brain region; a Yeo-17 network template is registered to an individual brain region to realize mapping; defining a tumor core area as an independent network unit, and forming a new set with 17 normal networks; calculating the connection strength of whole brain voxels and network functions, and judging functional connection voxels according to a threshold value; and quantifying the intratumoral function ratio RIFR and the peritumoral ligation ratio RPTR, and generating a visual map. According to the method, an individual brain function network is accurately mapped, tumor heterogeneity interference is overcome, repeatable quantitative indexes are provided, and a reliable imaging analysis tool is provided for brain glioma function protection and clinical research.

The invention discloses a breast cancer liver metastasis evaluation method and system based on iconomics feature analysis, and the method comprises the steps: extracting an original iconomics feature set through collecting the multi-modal image data of a liver region, obtaining a calibration feature subset through screening, constructing a feature time sequence evolution track, generating a high-dimensional fusion feature matrix through feature fusion, and carrying out the recognition of the high-dimensional fusion feature matrix. A specialized platform is combined to complete tumor heterogeneity quantitative grading, and finally comprehensive evaluation of the liver metastasis state and the progress trend is achieved; the system comprises a plurality of function units which are sequentially connected in series and interact in two directions, and full-process connection from feature collection, screening, modeling and fusion to quantitative evaluation and result output is achieved. According to the method, multi-dimensional static and dynamic characteristic information is comprehensively absorbed, invasive trauma is avoided by utilizing a non-invasive technology path, scientific support is provided for clinical treatment scheme formulation, curative effect monitoring and prognosis judgment, and the accuracy, systematicness and clinical practicability of breast cancer liver metastasis evaluation are remarkably enhanced.

The invention discloses a cross-instrument Raman spectrum data alignment method based on a Cycle-GAN network, and the method comprises the following steps: S1, collecting sample data through different Raman spectrometers, and obtaining a first data set and a second data set which have systematic differences and have no paired samples; s2, carrying out denoising, baseline removal and normalization preprocessing on the data set; s3, constructing a Cycle-GAN network model containing structure-symmetric double generators and double discriminators, the generators being used for data set bidirectional mapping, and the discriminators discriminating the authenticity of spectral data; s4, using the data set to train a network model in a self-supervision form; and S5, inputting the to-be-aligned spectral data into the trained corresponding generator to obtain alignment data consistent with the spectral characteristics of the other data set. According to the method, a self-supervised bidirectional Cycle-GAN network architecture is adopted, the method has an automatic optimizing capability, does not depend on a standard spectrum, is compatible with non-paired training data, can also retain differentiated spectrum characteristics of tumor heterogeneity, and can realize accurate alignment of cross-instrument Raman spectra.

The present application is related to a method for in situ detection of multiple lung cancer-related markers in live circulating malignant cells, and tumor heterogeneity, including accurately detecting at least one lung cancer-related nucleic acid marker in live lung cancer cells by using a lung cancer-targeted nanoprobe loaded with a plurality of molecular beacons to reflect levels of different markers in the lung cancer cells from a single cell level and a heterogeneous state of a tumor of a specific patient. The nanoprobe is a nanoparticle self-assembled from a high polymer material, an electropositive protein, a functional polypeptide and / or a functional aptamer, and a molecular beacon of a lung cancer-related marker. The nanoprobe can be used to target different phenotype circulating malignant cells in whole blood, detect nucleic acid markers in living cells, and reflect levels of different markers in the malignant cells from a single cell level.

The application discloses application of GCLC in enhancing cell uptake of BNCT drugs. Boron neutron capture therapy (BNCT) has become a promising method for treating osteosarcoma, and the treatment effect depends to a great extent on 10 Accumulation and distribution of B compounds in tumors. The application provides a gene marker capable of enhancing tumor cell uptake of BNCT drugs and application thereof, and proves that targeting GCLC can overcome the limitation of tumor heterogeneity on BNCT drug uptake by utilizing tumor-specific vulnerability.

This invention discloses a method for assessing the sensitivity of anticancer drugs based on patient-derived tumor tissue, belonging to the field of biomedical technology. The method includes: obtaining fresh tumor tissue from a patient and preparing it into tissue fragments of approximately 1 mm³; culturing the tissue fragments in vitro in a three-dimensional gel matrix; adding the anticancer drug to be tested to the culture system; digesting the tissue fragments after culture to obtain a single-cell suspension; and assessing drug sensitivity by detecting cell viability using flow cytometry. This invention can maximally preserve tumor heterogeneity and microenvironment, has a short operation cycle, and can quickly and accurately guide personalized clinical medication, especially suitable for solid tumors such as brain tumors.

Methods for determining tumor heterogeneity are described. The methods may comprise, for example, obtaining a medical image associated with the sample from a subject; identifying a plurality of patches from the medical image; identifying a plurality of patch groups, wherein each patch group comprises one or more patches of the plurality of patches and corresponds to a region of interest in the medical image; inputting each patch group into a trained machine learning model to generate a plurality of genomic alteration predictions corresponding to the plurality of patch groups, wherein the plurality of genomic alteration predictions is related to the presence of one or more genomic alterations in each of the input patch groups; and generating the phenotypic tumor heterogeneity score by comparing the plurality of genomic alteration predictions.

The invention relates to the technical field of biology, and discloses a method for discovering a new stomach-tumor target spot and a system establishment method. By integrating single cell second-generation sequencing and third-generation full-length transcriptome technologies and adopting a CopyKat and Infercnv dual algorithm model, the technical limitations that tumor heterogeneity cannot be analyzed by traditional batch sequencing and full-length new transcripts and malignant cells are difficult to accurately identify by conventional single cell sequencing are overcome. The method can systematically and precisely complete the whole process from cell subset analysis, malignant cell identification to brand new membrane protein target discovery, and further improves the reliability and transformation application potential of the discovered target through cross-platform verification and function correlation analysis, and has a wide application prospect. And a more efficient and reliable target spot discovery scheme is provided for gastric cancer targeted therapy and prognosis judgment.

This invention relates to the field of molecular probe technology, and more particularly to a near-infrared fluorescent probe, its preparation method, and its applications. The near-infrared fluorescent probe provided by this invention has the structure shown in Formula I. Through click chemical bridging, it enables convenient enrichment and identification of covalently targeted proteins, thereby simultaneously achieving high-sensitivity tumor imaging and in-depth proteomics analysis. The near-infrared fluorescent probe provided by this invention not only retains and optimizes the high-contrast tumor imaging capabilities of IR-780, but also serves as a highly efficient chemical proteomics tool, systematically revealing its target map and binding mechanism. Ultimately, it forms a platform technology integrating bright pan-tumor labeling and functional proteomics analysis, providing a solution to overcome tumor heterogeneity and advance the rational design of probes.

The application discloses a lung cancer gene mutation prediction method based on unsupervised clustering double-stage attention multi-instance learning, and relates to the technical field of pathological image analysis and gene detection. H&E staining pathological whole section images of non-small cell lung cancer patients and corresponding gene mutation data are collected to construct a data set; the images are preprocessed by using the OTSU method, segmented into blocks and high-dimensional feature vectors are extracted; the block features are grouped into cluster feature sets through unsupervised clustering; a double-stage attention mechanism composed of intra-cluster and inter-cluster is used to hierarchically aggregate and generate global features; finally, a classification model is used to output mutation positive / negative prediction results. The application groups the features through unsupervised clustering and structures the features, combines double-stage attention to strengthen key information, does not need complex manual annotation, adapts to various driver gene mutation prediction requirements, effectively deals with tumor heterogeneity and feature sparsity, improves prediction accuracy and generalization ability, and provides low-cost and efficient targeted therapy preliminary screening technical support for clinics.

Techniques for identifying a tumor heterogeneity of a subject are described. In an example method, sequence read data of a sample obtained from the subject is identified. The sequence read data is indicative of endpoint positions of nucleic acid molecules in the sample. The example method further comprises determining endpoint positions of the nucleic acid molecules, generating input features based on the endpoint positions of the nucleic acid molecules, and classifying, using a classifier, the tumor heterogeneity of the subject based on the input features.

The invention discloses an anti-tumor polypeptide and preparation and application thereof, and relates to the technical field of biological pharmacy, the anti-tumor polypeptide comprises a function-synergistic double-target tumor targeting module, a tumor microenvironment response connexon and a concealed effect module, and the polypeptide is formed through head-tail linear polymerization or side chain-side chain linear polymerization; at least one D-amino acid or non-natural amino acid is introduced into the molecule to improve the plasma stability, and at least one click chemical modification site is reserved; amino acid sequences of the double-target tumor targeting module and the hidden effect module are screened and optimized through an AI algorithm; through three-section design of'targeted recognition-microenvironment activation-specific killing ', 'concealment-activation' strategy and double-target collaborative combination, accurate release of an effect module at a tumor part is realized, off-target toxicity to normal tissues is greatly reduced, and the treatment escape risk caused by tumor heterogeneity is effectively overcome.

The present application relates to the field of tumor biomarkers, in particular to a biomarker based on tumor tissue perfusion imaging features. The present application constructs a biomarker OMP based on tumor tissue perfusion imaging features, which is obtained by segmenting the tumor enhanced image in a "peeling onion" manner and after data analysis with specific parameter setting. The biomarker OMP of the present application is used for immune-related treatment effect prediction, fully embodies the influence of tumor heterogeneity on tumor diagnosis and treatment, and has the advantages of simple operation, reliable diagnosis, safety and non-invasiveness, strong repeatability and the like. It can be dynamically evaluated with patient follow-up imaging data, and as a new marker, it has a wide range of application population and high clinical application value.

The present application relates to a kind of based on image network's image heterogeneity area fusion technology to curative effect prediction method and device, belong to intelligent medical treatment and medical image analysis technical field.The method first divides the tumor and peritumoral region in medical image into multiple non-connected sub-regions as graph node, and generates the comprehensive feature representation that fused local image feature and node type embedding vector for each node;Subsequently, based on the edge between nodes defined by spatial distance threshold, a graph structure containing intratumoral, peritumoral and cross-border connection is constructed;Finally, the graph structure is input into graph neural network model for training, and a curative effect prediction model is obtained.The present application explicitly models the complex spatial topological relationship between tumor heterogeneity regions through graph structure, solves the core problems of spatial information loss and insufficient heterogeneity modeling in the prior art, significantly improves the accuracy and robustness of treatment efficacy prediction.

The invention relates to the technical field of medical image processing. The invention discloses a habitat imaging-based gastric cancer neoadjuvant chemotherapy curative effect prediction method and a habitat imaging-based gastric cancer neoadjuvant chemotherapy curative effect prediction system, and the method comprises the following steps: S1, obtaining pre-treatment CT and clinical pathology information, and marking the maximum tumor area and upper and lower layers; s2, dividing the marked region into habitat sub-regions with different biological characteristics through habitat imaging; and S3, performing prognosis analysis on the habitat subregion, screening independent prognosis features related to survival, and generating a habitat subregion image. And S4, constructing a joint attention model, inputting images of the marked area and the habitat subarea, predicting the curative effect and the total lifetime, and obtaining a survival score. And S5, integrating clinical pathological information and survival scores, performing prognosis analysis, constructing a column graph, and evaluating correlation. After a CT image label before treatment is obtained, habitat imaging is used for dividing subareas, related images are generated, the model is input in a combined mode, tumor heterogeneity is revealed, and the curative effect is predicted.

The application discloses a rectal cancer postoperative recurrence risk prediction system and method based on multi-modal time series data, and relates to the technical field of medical artificial intelligence. The system comprises a data acquisition and preprocessing module, a feature extraction module and a multi-modal feature fusion and modeling module. The method comprises the following steps: constructing a cross-modal data set comprising time series clinical data, time series MR images and biopsy digital pathology images; extracting clinical features, imageomics and deep learning features of MR images, and cell nucleus morphology and spatial distribution features of pathology images respectively; and fusing all the features by using a Transform network to construct a prediction model. The application integrates macroscopic images, microscopic pathology and dynamic time series information, comprehensively quantifies tumor heterogeneity, solves the problem of inaccurate prediction of a single modal static model, and can more accurately evaluate the postoperative recurrence risk of a stage III rectal cancer patient, thereby assisting clinical treatment decision-making.

This invention discloses the application of a biomarker in assessing the risk of prostate cancer recurrence, belonging to the field of tumor molecular diagnostics and bioinformatics technology. The biomarker includes at least one of the TACR2 and LPIN3 genes, and RCCD1, PRSS27, and NAP1L5, with the core detection focusing on the methylation-transcriptional combined characteristics of the TACR2 gene. This invention obtains patient samples, uses RT-qPCR to detect the expression levels of multiple genes, and methylation-specific PCR to detect the methylation level of the TACR2 promoter region. A risk scoring model is constructed using LASSO regression, and patient prognosis is stratified based on the score. This invention solves the problems of traditional assessment indicators being affected by tumor heterogeneity and having limited predictive accuracy. It is the first to use TACR2-related multi-omics characteristics for prostate cancer recurrence assessment. The multi-gene combined model improves predictive accuracy and can be developed into a standardized testing kit, providing a scientific basis for clinical recurrence risk assessment and personalized treatment, with significant industrial and clinical application value.

This invention discloses a radiotracer targeting pan-KRAS mutant proteins, its preparation method, and its applications. The tracer uses small molecule ligands that specifically recognize multiple KRAS mutant proteins as targeting modules, and is coupled to diagnostic radionuclides through an optimized linkage system. This tracer is independent of specific mutation sites and can broadly and specifically bind to multiple high-frequency KRAS mutant subtypes, including G12C, G12D, and G12V. Based on PET / SPECT imaging, this tracer enables rapid, non-invasive, systemic, visualized, and quantitative assessment of KRAS mutant protein expression load in living tumors, overcoming the invasiveness, spatiotemporal limitations, and tumor heterogeneity challenges of traditional biopsies. It provides a novel molecular imaging tool for accurate companion diagnosis of KRAS-mutant tumors, screening of patients benefiting from pan-KRAS inhibitors, and monitoring efficacy, possessing significant clinical translational value.

Methods for using tumor heterogeneity score as a biomarker to guide treatment and / or patient monitoring decisions are described. The methods may comprise, for example, acquiring knowledge of a driver mutation in a sample obtained from a subject; identifying one or more variants present in the sample; determining a tumor heterogeneity score for the sample based on the one or more variants; and comparing the tumor heterogeneity score for the sample to a threshold tumor heterogeneity score, where if the tumor heterogeneity score is less than or equal to the threshold tumor heterogeneity score, the subject is identified for treatment with a first anti¬cancer agent, and where if the tumor heterogeneity score is greater than the threshold tumor heterogeneity score, the subject is identified for treatment with a second anti-cancer agent.

The application relates to the field of biomedical detection technology, in particular to a bladder cancer postoperative recurrence monitoring method based on a ctDNA methylation spectrum, which comprises the following steps: collecting postoperative patient cell-free plasma and extracting circulating tumor DNA; performing bisulfite conversion treatment on the DNA; performing targeted amplification using a multiplex PCR primer group designed for a group of predetermined genomic methylation regions related to recurrence, constructing a sequencing library; performing high-throughput sequencing on the library, obtaining methylation level data of a plurality of CpG sites to form a sample methylation spectrum data matrix; the core of the application is that tumor heterogeneity is overcome through multi-marker combination detection, and a machine learning model is used to integrate multidimensional data to realize precise risk stratification, which has the advantages of non-invasiveness, high sensitivity, high specificity and quantifiable output, and provides an effective tool for individualized follow-up management of bladder cancer postoperation.