13 results about "Clinical report" patented technology

The invention provides a deep learning-based oral cavity image recognition method, a storage medium and a deep learning-based oral cavity image recognition system. The method comprises the steps of deploying a federated learning framework and collecting a multi-modal oral cavity image data set; extracting local features to obtain image features, and generating a modal adaptive weight map; a multi-head self-attention mechanism is used for fusing the cross-modal features to generate a fused feature map, and deconvolution up-sampling is carried out to form high-resolution multi-modal feature representation. A tooth segmentation mask is generated based on this representation, and an initial diagnostic report is generated. And aggregating the attention weight of each client through an encryption protocol, and generating interpretable decision support data. And finally, generating a structured clinical report by using a natural language. According to the method, the Grad-CAM thermodynamic diagram is combined with the encrypted and aggregated attention weight, so that the privacy security is guaranteed, the model interpretability is enhanced, the clinical credibility and the diagnosis decision efficiency are improved, and the problems of insufficient diagnosis precision of complex lesions and insufficient utilization of multi-modal information in the prior art are solved.

The invention relates to the technical field of image processing and reproductive hospital inspection, and provides a sperm movement tracking and quality analysis method based on deep learning and optical flow fusion, which comprises the following steps of: identifying sperms through a target detection model, acquiring a detection frame, and screening to obtain detection points; predicting the predicted position of the sperm in the current frame by using a sparse optical flow method based on the historical trajectory, performing data association on the predicted position and a detection point, establishing a matching relationship, updating the trajectory according to the matching relationship, establishing a new trajectory or terminating a lost trajectory, calculating kinematics parameters and sperm concentration based on the complete motion trajectory, and completing activity grading. According to the invention, through fusion of deep learning detection and optical flow prediction, high-precision identification under a complex background and stable tracking under a high-density cross scene are realized, the identity exchange rate is significantly reduced, and the shielding robustness is enhanced; and meanwhile, full-process automation from video input to clinical report is realized, and the accuracy, repeatability and clinical credibility of an analysis result are improved.

The invention discloses a medical image report generation method, system and device and a storage medium, and relates to the technical field of computer vision and natural language processing, and the method comprises the steps: constructing a causal graph model, taking an image as an input variable, taking a final report as an output variable, and taking a region-level pathological state in the image as an intermediary variable; the data imbalance factor is an unobservable hybrid factor; based on a causal graph model, intervening the intermediary variable by using a front door adjustment strategy, and establishing a causal path from visual evidence to report text; based on the intervened intermediary variable, executing a report generation process: a, identifying an abnormal region in the image by using a focus detection model and outputting a corresponding pathological discovery description; and b, inputting the intervened pathological discovery description and the original image into a visual language model to generate a complete clinical report by taking the intervened pathological discovery description and the original image as conditions. And the sensitivity of the model to pathological changes and the clinical reliability of report generation are improved.

A computer-implemented method, computer-system and computer-program product for generating medical training data for training artificial-intelligence (AI) algorithms for supporting clinical reporting and documentation are described. To generate the training data medical image data of a patient comprising medical image data elements are received and a medical findings report is generated, edited and / or received that summarizes individual medical findings. It comprises machine-readable findings-report elements the contents of which comprise semantic features. The contents of the findings report elements are automatically assigned to unique identifiers, wherein each identifier uniquely represents the medical semantic content of exactly one individual medical finding. The medical image data are annotated by linking one or more medical image data elements to the unique identifiers of one or more contents of the findings-report elements, and the annotated received medical image data are stored as training data for AI algorithms for supporting clinical reporting and documentation.

The invention discloses an MRI (Magnetic Resonance Imaging)-based thyroid eye disease activity assessment method and system, and aims to solve the problems that the deep state is difficult to assess and the subjectivity is strong in the existing clinical activity score. The method comprises the following steps: acquiring and preprocessing a magnetic resonance image and generating a standardized three-dimensional image; automatically segmenting an orbit structure through a deep learning network; extracting multi-parameter image features; calculating a disease activity probability score by using a machine learning classifier; and mapping the probability score into an activity grading result and generating an evaluation report. The system comprises an image data acquisition and preprocessing module, an orbit structure automatic segmentation module, a multi-parameter image feature extraction module, an activity evaluation model construction and reasoning module and a clinical report generation module. According to the technical scheme, image evaluation efficiency and repeatability can be remarkably improved, subjective deviation is reduced, active-stage inflammatory edema and inactive-stage fibrosis are accurately and quantitatively distinguished, evaluation specificity is improved, and support is provided for individualized treatment.

The present disclosure relates to predicting a survival rate of a patient with cancer following a treatment. In some embodiments, one or more processors: apply a large language model to a clinical report to extract a plurality of clinical features; acquire a pre-treatment image and a post-treatment image of the patient with cancer; apply a segmentation algorithm on an annotated volume of interest (VOI) of the pre-treatment image to obtain a first segmented VOI; apply the segmentation algorithm on the annotated VOI of the post-treatment image to obtain a second segmented VOI; determine a plurality of radiomics features and a plurality of deep learning features from the first segmented VOI and the second segmented VOI; and apply a machine learning model to the plurality of clinical features, the plurality of radiomics features, and the plurality of deep learning features to predict the survival rate of the patient with cancer.

The embodiment of the invention discloses a three-dimensional myocardial scar visualization system. According to one specific embodiment, the system comprises a data processing end configured to perform heart structure and scar segmentation on target heart magnetic resonance data to obtain a binary tag graph for the heart and the myocardium; constructing an individualized ventricular network and a scar three-dimensional surface model based on the binary label graph; generating an initial myocardial scar three-dimensional model; generating lead implantation track information; generating a myocardial scar three-dimensional model according to the lead implantation track information; and the model application end is configured to perform model rendering on a target application page according to the model parameters corresponding to the myocardial scar three-dimensional model sent by the data processing end so as to provide an interactive operation function and generate a clinical report. According to the embodiment, accurate and visual three-dimensional visual decision support can be provided for preoperative planning and intraoperative navigation of heart diseases, so that the accuracy of surgical planning and the clinical working efficiency are remarkably improved.

A structured ultrasound report generation method based on a large language model, belonging to the field of medical image processing, firstly utilizes a multi-agent data construction framework to automatically synthesize large-scale, high-quality "verbal diagnostic clues-structured report" data pairs from unpaired ultrasound report text. Next, a template-enhanced supervised fine-tuning technique is employed to align verbal clues with institutional standard templates, using a pre-set template library to standardize the report's structure and terminology, bridging the domain knowledge gap of the model. Finally, a defect-oriented preference optimization strategy allows the model to learn from its own generated defective reports, continuously improving its diagnostic accuracy and reliability. This invention can convert fragmented diagnostic information verbally uttered by doctors during examinations into a complete, standardized, and high-quality structured clinical report in real time and automatically, significantly improving the efficiency and standardization of report generation.

The invention discloses a multi-modal prompt fine tuning method based on a thyroid ultrasound image combined with target detection, which comprises the following steps: 1) realizing precise positioning of thyroid nodules by introducing a target detection module, and obtaining high-level visual features by combining a pre-training image feature extraction technology; 2) performing semantic mining on clinical report text input by adopting a soft prompt fine tuning enhanced text processing technology, and realizing efficient field adaptation under limited ultrasonic image-text data; and 3) constructing a multi-modal fusion mechanism through a multi-head attention network, integrating visual features and semantic embedding vectors, and generating a risk score and an abnormal possibility index of the thyroid nodule through a risk assessment module. The nodule feature evaluation method adaptive to the ultrasonic image analysis scene is constructed through modular design, the technical pain points of the ultrasonic image can be effectively processed, multi-modal information is fully mined, and the dependence on a large amount of labeled data is reduced while the stable quantitative feature evaluation capability is provided.

The application discloses a method and system for locating an epileptogenic focus of epilepsy based on magnetoencephalography and a medium, and relates to the technical field of artificial intelligence, and comprises the following steps: registering a structural image of a patient after preprocessing with a magnetoencephalogram of the patient, dividing a cerebral cortex region into a grid based on the preprocessed structural image and the registration result, and obtaining a forward model of the patient; preprocessing the magnetoencephalogram and detecting a spike wave, and obtaining a spike wave time point sequence; calculating source coordinates and source directions of the spike wave time points based on the forward model and the spike wave time point sequence, classifying the spike wave time point sequence based on the source coordinates and the source directions, and obtaining a clustering result; calculating source coordinates and source directions of a class center of each class in the clustering result by using a magnetic dipole algorithm; and generating a clinical report of the patient; the locating method solves the problem that epilepsy diagnosis is greatly different and prone to errors due to differences in the levels of different doctors, and can quickly determine the location of an epileptogenic focus.

The invention relates to the technical field of medical diagnosis and artificial intelligence, in particular to a urinary tract disease prediction system based on multi-modal chromosome abnormality and clinical data, which integrates demographic statistics, clinical symptoms, laboratory detection, molecular biology of nine key chromosome sites and multi-dimensional data of images, and performs pre-processing and single-modal feature extraction to obtain a prediction result of the urinary tract disease. Fusion features are generated in a targeted mode through a task self-adaption fusion module, and then urinary tract epithelial cancer or neoplastic lesion positive prediction, positive sample TNM staging and pathological grading prediction and focus origin positioning are achieved through a multi-task model. According to the system, an edge cloud collaborative architecture is adopted, the computing power requirements of different medical institutions are met, the feature contribution degree is determined through an SHAP method, a visual clinical report is generated, the problems that in the prior art, multi-modal data integration is insufficient, and non-invasive staging and grading are lacked are solved, prediction reliability and clinical adaptability are improved, and support is provided for clinical auxiliary decision making.

The present application relates to the field of intelligent medical detection, and provides a nasopharyngoscope multi-task analysis and report generation method and system based on a multi-modal large model, which comprises preprocessing the collected original images, automatically identifying and cropping the effective field of view area, performing image scaling, constructing a multi-task instruction set, adopting a base multi-modal large model, adjusting the parameters of the base multi-modal large model by using the multi-task instruction set, inputting the scaled images into the multi-modal large model, sequentially outputting the model content according to the task set order to encapsulate the context text, inputting the context text together with the scaled images into the multi-modal large model again, and generating a clinical report. The present application protects a serial processing method of classification-detection-generation, that is, the structured output of a previous task is used as a constraint condition of a subsequent generation task, the spatial orientation of a sign is locked through cascading reasoning, and the anatomical position error rate of an analysis report is effectively reduced.

The present disclosure relates to a method and device for generating clinical record data for recording medical treatment. The method includes receiving medical data in which medical treatment, performed in advance, is recorded; recording information, included in the medical data, in a layer corresponding to an item related to the medical data from among a plurality of layers classified according to a plurality of items; and generating a clinical report based on the plurality of layers.