5results about How to "Forecast stability" patented technology

The invention belongs to the technical field of biological information processing, and particularly relates to a method and system for constructing a risk prediction model from Kawasaki disease to huge coronary artery tumor. The invention provides a Kawasaki disease MGCAA risk prediction model construction method and system, and the method comprises the steps: building a prediction model based on six conventional clinical variables through employing a random forest (RF ranger) algorithm; an SHAP explanation mechanism is introduced, so that the contribution of each variable to a prediction result can be visually displayed in global and individual levels, and the transparency and clinical interpretability of the model are improved; through external verification and intercept-only recalibration, the reliability and the applicability of the model in different groups of people can be ensured; according to the method, an online webpage tool (Shiny App) is deployed, and a doctor can input clinical data of a patient in real time and immediately obtain individualized risk prediction and explanation, so that clinical early recognition of an MGCAA high-risk child patient is realized, and formulation of an individualized treatment scheme and early intervention measures is assisted.

This invention discloses a method for calculating the three-phase phase equilibrium of oil, gas, and water at the nanoscale, relating to the field of shale oil reservoir development technology. The method includes the following steps: receiving input parameters; constructing a total pressure model of the confined fluid at the nanoscale based on the input parameters; constructing a three-phase pressure equilibrium model among the oil, gas, and water phases at the nanoscale based on the total pressure model; constructing a phase stability test model, and determining whether phase separation occurs in the system under given temperature and pressure conditions based on the tangent plane distance function; when the phase stability test model determines that phase separation has occurred, calculating the phase fraction and composition of the oil-gas-water three phases by iteratively solving for the conditions satisfying the three-phase pressure equilibrium model and the equality of the fugacity of each phase component, and outputting the results. This invention can accurately and stably predict the fluid phase state in the nanopores of shale oil reservoirs, providing key technical support for its component simulation and efficient development.

This invention discloses an intelligent planning method and system for spinal pedicle screw placement based on medical imaging. The method first acquires spinal medical images, target vertebral body segmentation results, and screw planning annotation data. Based on the target vertebral body segmentation results, the target vertebral region is extracted and divided into left and right pedicle sub-blocks along the left-right direction. The right pedicle sub-block is mirror-mapped, and its screw start and end coordinates are simultaneously transformed to mirror coordinates, thereby unifying the samples from both sides into a standard sample space with the same orientation. This reduces the interference of anatomical orientation differences on model training and improves sample utilization efficiency. Then, a three-dimensional deep learning model is trained using the standard sample set, simultaneously predicting screw start, end, and diameter-related parameters. The screw diameter is output using ordinal regression, and the predicted diameter value is solved through ordered thresholds and cumulative probabilities, making the prediction results more consistent with the ordered attributes of clinical specifications. For the case to be planned, the model output results are restored to the original coordinate system according to the corresponding relationship to obtain the placement path and diameter parameters of both pedicles. This invention can simultaneously improve the positioning accuracy of the placement path, directional consistency, and screw specification prediction stability, and is suitable for preoperative planning and assisted surgical systems for the spine.

The application provides a protein-metal ion binding site recognition method, device and equipment, and belongs to the field of protein detection. The method comprises the following steps: extracting features of known metal ion binding proteins to obtain a plurality of sample evolutionary information features; for an unknown protein sequence, candidate distant homologous metal ion binding proteins are determined through multiple sequence comparison and cosine similarity screening, and a training set is constructed; a composite framework of a bidirectional long short-term memory network and a full connection neural network is adopted, input features include evolutionary information features and physicochemical property features, and output is probability values of different metal ion binding sites; the composite framework is trained through the training set to obtain a site prediction model; and the prediction model is used to predict the binding sites of the unknown protein sequence. Stable and efficient prediction of metal ion binding sites is realized.

The invention belongs to the technical field of 3D scene reconstruction, and particularly relates to a 3D scene reconstruction method and system based on illumination adaptive plane representation, and the method comprises the steps: completing scene multi-view image collection and equipment parameter recording, extracting multi-scale image features, carrying out the targeted enhancement processing of a weak light image, and integrating all image features to provide high-quality data support; based on the scene range of the enhanced image, initializing a multi-layer grid structure, predicting the plane height through stable calculation, quantifying the plane perception effectiveness and reliability and storing related data, and providing structured plane representation for subsequent reconstruction; calling early-stage reconstruction data and the integrity matrix to identify a defect area, constructing a state input tensor fusing plane key indexes, and screening out an optimal plane set for defect supplementation through a double-depth Q network; and extracting features and position data of the optimal plane, projecting the data to a 3D space, and completing semantic decoding in combination with scene prior.