5 results about "Cell movement" patented technology

The application discloses an animal cell adaptive tracking method and system based on deep learning, and belongs to the technical field of cell image analysis; a YOLOv10 deep learning model is combined with a Btrack adaptive tracking algorithm, animal cell samples are first dyed and microimage collection is carried out, a cell detection model is trained after image preprocessing, data labeling and augmentation, the model is converted into an ONNX format to realize efficient inference; then, Kalman filtering prediction and three-round Hungarian algorithm matching are carried out to complete cell adaptive tracking, the problems of cell occlusion, division, overlap and loss of tracking can be effectively handled, and stable and continuous cell motion trajectories are generated; the application can be deployed without GPU, has strong generalization ability, high detection and tracking precision, is suitable for scenes such as microecological research, clinical sample testing, drug effect evaluation and cell behavior analysis, and can significantly improve the automation level and stability of animal cell detection, counting and dynamic tracking.

The application discloses a monitoring method and system for three-dimensional phenotype parameters of stoma cells of living plants, and comprises the following steps: S1, a plant confocal three-dimensional microscopic imaging and photosynthetic parameter measuring device is used, a self-made photosynthetic leaf chamber is used to realize cooperative monitoring of a down-positioned confocal microscope and a portable photosynthetic instrument, morphological changes of the same stoma under different environments are monitored, and three-dimensional Z-stack images and physiological parameters of a stoma complex of the living plants are synchronously obtained; and S2, a fluorescence image is automatically segmented and three-dimensional surface reconstruction is performed by using a granularity-based active contour model, so that three-dimensional phenotype parameters and pore characteristics of the stoma complex are obtained.

The application discloses a multi-agent based spaceomics tissue dynamics reconstruction and simulation perturbation method, and belongs to the technical field of computational biology and bioinformatics. The method models each cell as an agent, constructs continuous dynamic trajectories from the starting slice to the ending slice through unbalanced optimal transport and neural ordinary differential equations, uses a multi-agent reinforcement learning framework to make each agent decide the fate of proliferation, apoptosis, differentiation or migration according to the local microenvironment state, and combines a reward function to optimize the generation of missing intermediate tissue states; then, the cosine similarity is used to quantify the alignment degree of ligand-receptor communication signals and cell movement to identify key regulatory pairs, support computer simulation perturbation experiments on genes, communication pathways or cell populations. The application can reconstruct three-dimensional / four-dimensional tissue dynamics with high fidelity from sparse slices, analyze microenvironment regulation mechanisms and predict perturbation effects, and is suitable for spaceomics data analysis in the fields of developmental biology, oncology and the like.

The invention discloses a biological resonance enabling sheet, which comprises a biological resonance chip, a functional layer and a packaging sleeve, and is characterized in that the functional layer is arranged on the biological resonance chip; the packaging sleeve is arranged on the outermost side in a sleeving mode, the functional layer is formed by mixing silica gel and stone needle powder, and air holes are formed in the side face, facing the functional layer, of the packaging sleeve. According to the biological resonance energizing sheet, the resonance effect of the biological resonance chip is combined with the effect of the stone needle powder, so that the resonance effect is enhanced, and cell movement is promoted.

The disclosed technology creates a set of technologies that formulaically and / or positionally link cell(s) to spreadsheet functions or capabilities that flex the number of cells they instantiate. Thereby automatically altering the cell formula and / or the cell position reflecting those linked / connected flex range changes. Automatically ensuring no overlap of linked / connected cells and automatically adjusting linked / connected formulas for the referenced flex changes.