15 results about "Subtype classification" patented technology

The application discloses a dynamic prototype multi-model medical image classification method based on feature credibility evaluation, and relates to the technical field of medical image processing.The method effectively solves the problems in traditional medical image subtype classification, such as lack of high-quality labeled data, high uncertainty of pseudo-labels, and difficulty of static prototypes in adapting to dynamic changes of lesions, and the like.Through construction of a multi-structure feature extraction network and completion of hierarchical feature fusion, the method combines feature cross learning, local attention modeling and expanded convolution to supplement context information, and strengthens semantic consistency and structural continuity of lesion region features.Meanwhile, the method constructs a spatial similarity graph through cosine similarity, and models and fuses a spatial uncertainty graph with any and cognitive uncertainty based on a Dirichlet distribution, obtains a reliable evidence graph through exponential fusion, and generates a pseudo-label with sample-level confidence, so that effective supervision information is accurately screened from a feature level, cumulative deviation of false pseudo-labels is greatly reduced, and stability of a semi-supervised learning process is improved.

The application relates to the field of biomedical technology, and discloses a diagnostic agent for identifying melanoma molecular subtype typing and application, wherein the diagnostic agent comprises a first antibody specifically combined with a SOX10 protein and a second antibody specifically combined with an EGR1 protein; the diagnostic agent can be applied to preparation of a diagnostic product for evaluating the prognostic effect of a melanoma patient, preparation of a diagnostic product for predicting the treatment sensitivity of melanoma to a BRAF inhibitor, and preparation of a diagnostic product for guiding an individualized treatment scheme of melanoma. The diagnostic agent for melanoma molecular subtype typing can be directly transformed into clinical practice, and can assist in realizing real individualized treatment.

The present application relates to a kind of melancholic depression plasma biomarker kit detection method, for detecting the plasma biomarker of melancholic depression.The present application provides a kind of application method by detecting plasma adiponectin (Adiponectin) protein as the application method of melancholic depression biomarker, the present application method is used for the subtype classification diagnosis of melancholic depression for non-diagnostic purposes, improve the sensitivity and specificity of melancholic depression screening.

The application discloses a kind of glass wart measurement and subtype classification method based on physician mark mask, belong to medical image processing and ophthalmic artificial intelligence auxiliary diagnosis technical field.This method takes the binary mask marked by physician as core input, in turn through image reading and abnormal compatibility processing, mask pretreatment and lesion contour extraction, morphological parameter calculation and physical unit conversion, automatic classification based on AREDS standard Subtype, finally complete visual annotation, structured report generation and batch processing.The application realizes Chinese path compatibility by binary stream decoding, extracts lesion contour using OpenCV related interface, calculates area, maximum diameter, height, roundness and other parameters, completes the conversion of pixels and actual physical units combined with calibratable scale, strictly follows AREDS standard to divide subtype according to maximum diameter of lesion, and supports batch processing and abnormal protection.The application avoids the error of automatic segmentation algorithm, improves the measurement accuracy and result consistency, realizes full-process automation, adapts to clinical diagnosis and treatment and scientific research demand, and provides reliable tool for early quantitative evaluation of age-related macular degeneration.

This invention provides a method and system for cancer subtype identification based on multi-omics data, belonging to the field of bioinformatics processing technology. The method includes: constructing a DILCORE model that integrates a multi-branch variational autoencoder, contrastive learning, and cross-view attention mechanisms. The multi-branch variational autoencoder decomposes omics observation data into common components for subtype classification and view-specific noise components, achieving noise suppression; the InfoNCE contrastive loss is introduced to bring common representations of the same sample closer across different views, enhancing cross-omics consistency; the cross-view residual self-attention mechanism is used to adaptively weight and fuse common vectors; finally, a self-supervised clustering fine-tuning optimization strategy is introduced to jointly improve representation quality and clustering performance in the latent space. This achieves deep and effective integration of multi-omics data, significantly improving the accuracy of cancer subtype identification and providing a powerful tool for personalized cancer treatment and prognostic assessment.

Described herein are gene signatures providing prognostic, diagnostic, treatment and molecular subtype classifications of cancers through genomic and epigenomic profiling, Methods and compositions for determining cancer and subtypes, including breast cancer are described and specific and sensitive detection of biomarkers of interest is provided. Such biomarkers are indicative of disease pathogenesis, which provides opportunity for selection of treatment, including treatment regimes directed at identifying candidates for responsiveness and overcoming resistance mechanisms.

This invention provides a graph-guided multimodal decoupling-based Alzheimer's disease (AD) subtype contrastive clustering method, solving the technical problems of traditional methods failing to fully utilize the neighborhood structure between patients and lacking the ability to distinguish between consistent and unique features among different pathological subtypes. The technical solution includes the following steps: S10, constructing a neighborhood graph structure for each modality of Alzheimer's disease (AD) patients and extracting structural enhancement features for each modality using a graph convolutional network; S20, using an autoencoder and feature decoupling module; S30, generating pseudo-labels by passing the representations through a clustering layer; S40, constructing a hard sample contrast perception mechanism based on the consistency differences and confidence levels of predictions for each modality and performing dynamic weighted learning; S50, minimizing all loss functions and predicting the subtype clustering result for each AD. This invention fully utilizes multimodal complementary information, avoids modal noise interference, and improves the accuracy of AD subtype classification.

This invention discloses a computational pathology image analysis method based on task-specific microenvironment reconstruction, comprising: task-specific semantic anchor construction, which utilizes a large language model to generate morphological descriptions and project them onto the visual space to construct a task-specific semantic anchor set; semantically guided local tumor microenvironment reconstruction, which combines self-attention and semantic anchors to screen key regions and identify key morphological anchors, and achieves structure-aware reconstruction through cross-attention; memory-driven global feature enhancement, which utilizes a dynamic memory bank to fuse intra-sample topological associations and cross-sample pathological priors to generate task-specific representations; and finally, inputting the data into a multi-instance learning aggregator for diagnostic prediction. This invention, through semantic guidance and memory enhancement, effectively solves the problems of semantic misalignment between general features and clinical tasks, as well as the loss of tumor microenvironment topology, significantly improving the accuracy and robustness of computational pathology image analysis. It can be widely applied to tasks such as cancer subtype classification and grading assessment.

This invention belongs to the fields of medical image processing and computer-aided diagnostic technology, and discloses a method and system for molecular subtyping of breast cancer based on DCE-MRI habitat omics analysis. By segmenting and preprocessing breast DCE-MRI images, a dual-path collaborative habitat segmentation module is used to analyze tumor heterogeneity from two complementary dimensions: pharmacokinetics and dynamic enhancement patterns. The segmented habitat features undergo deep interaction and adaptive weighting through a multi-level fusion module to generate a highly discriminative joint feature representation, which is finally output as a molecular subtype diagnosis by a classifier. This network enables deep interaction and complementary enhancement of the two feature paths, improving feature representation capabilities. In the testing phase, for complete modality data, a complete diagnostic process from habitat analysis to subtype classification can be completed. This invention achieves high-precision auxiliary identification of breast cancer molecular subtypes, with advantages such as strong tumor heterogeneity analysis capability, high feature discrimination power, and automated process.

The application discloses a delirium intelligent identification system based on multi-modal features and explainability analysis, and particularly relates to a multi-modal information modeling method and system based on facial local motion features, eye movement behavior parameters and time sequence features in a video sequence, and especially to an intelligent identification device and an implementation method thereof for automatically extracting, analyzing and discriminating behavior and expression time sequence features of a monitored object by using a non-contact acquisition device; the application can be applied to auxiliary evaluation of occurrence identification and subtype classification of delirium with acute attack and fluctuant symptoms in a medical monitoring scene; the scheme constructs a multi-modal fusion network integrated with fine eye movement features, micro-expression time sequence features and video space-time features, and further introduces an integrated voting strategy to realize fine and high-robustness three-classification identification from "whether delirium" to "specific subtype (non-delirium, agitated delirium and inhibited delirium)".

The application discloses a kind of endometrial carcinoma molecular typing methods based on cascaded multiple instance learning, method is by the binary mask of tumor area generated to endometrial carcinoma whole section image preprocessing, to extract the image block in region of interest and construct instance set by dyeing normalization processing;Using PatchVMamba instance encoder, the feature of each image block in instance set is extracted to obtain the instance feature vector of each image block, and the instance feature bag of each whole section image is formed;Using attention mechanism, the instance feature bag is weighted and aggregated to obtain the slice-level feature representation;The slice-level feature is input into the three binary classifiers arranged in cascade to obtain the probability value corresponding to the three subtypes of endometrial carcinoma respectively, and a unified threshold is used for cascade judgment of the subtype classification of endometrial carcinoma to output the final typing result.The cascaded structure of the application makes the consistency of the typing result and the diagnosis of pathological experts high, and realizes accurate typing.

The present application relates to the technical field of cell feature extraction, and specifically discloses a white blood cell multi-dimensional feature extraction and classification identification method and system. The method comprises: analyzing the shape and extension state of red blood cells to obtain an eccentricity vector of a global push piece; dividing white blood cells into non-polluted targets and potential polluted targets based on the eccentricity vector; dividing into upwind areas and downwind areas, and extracting a side pollution index through optical density comparison; constructing a dynamic physiological tolerance for calibration to obtain a net feature reconstruction sequence; integrating the original features of non-polluted targets and the net feature reconstruction sequence into a non-interference feature group, and outputting a classification identification result. The system comprises a contour collection module, a target separation module, a side analysis module, a feature reconstruction module, and a feature shaping module. The present application is conducive to reducing physical feature pollution caused by one-way diffusion of nuclear substances during the push piece process, and improving the accuracy of white blood cell subtype classification identification by the system.

Embodiments described herein provide a neural network-based cancer detection and subtyping tool for predicting the presence of tumors, their primary tissue, and their subtypes using small RNA sequencing (smRNA-seq) data (e.g., oncRNA count data). More specifically, this AI-based cancer detection and subtyping tool uses variational Bayesian estimation and semi-supervised training to adjust for batch effects and to learn low-dimensional distributions to account for the biological variability of the data. Methods for determining possible subtypes (groups) in a cancer sample are also provided.

This application relates to a method and system for classifying Parkinson's disease subtypes, belonging to the field of artificial intelligence. The method includes the following steps: S1, collecting and processing TUG gait data from Parkinson's patients and healthy controls using millimeter-wave radar to obtain gait parameters and Doppler spectra; S2, dynamically gating and fusing the gait parameters and Doppler spectra to obtain multimodal features; S3, employing a two-stage progressive training strategy: in the first stage, only the PD classifier is optimized and the subtype classifier is frozen, outputting the PD classification result based on the multimodal features; in the second stage, only the subtype classifier is optimized and the PD classifier is frozen, outputting the subtype classification result based on the multimodal features. This application solves the problem of unstable training and unsatisfactory training results caused by the significant heterogeneity of multimodal data.

Described herein are gene signatures providing prognostic, diagnostic, treatment and molecular subtype classifications of cancers through genomic and epigenomic profiling, Methods and compositions for determining cancer and subtypes, including breast cancer are described and specific and sensitive detection of biomarkers of interest is provided. Such biomarkers are indicative of disease pathogenesis, which provides opportunity for selection of treatment, including treatment regimes directed at identifying candidates for responsiveness and overcoming resistance mechanisms.