Prediction method and device of blood pressure data, electronic equipment and storage medium
By employing a blood pressure prediction method based on the Transformer model and utilizing cascaded encoder-decoder networks to process PPG and ABP time-series data, the method addresses the insufficient accuracy of PPG signal-based blood pressure parameter prediction in existing technologies and achieves more efficient ABP time-series data prediction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGDONG OPPO MOBILE TELECOMMUNICATIONS CORP LTD
- Filing Date
- 2023-11-20
- Publication Date
- 2026-06-26
AI Technical Summary
Existing technologies are not accurate enough when using PPG signals to predict blood pressure parameters. Conventional machine learning methods require a lot of feature engineering processing, and deep learning methods are not designed for time-series data, resulting in information loss and gradient vanishing.
A blood pressure prediction model based on the Transformer model is adopted. By acquiring PPG and ABP time series data, prediction is performed using a cascaded encoder-decoder network and output layer, which solves the problem of parallel processing of time series data and improves prediction accuracy.
It enables better prediction of ABP time series data based on PPG time series data, effectively solving the problems of information loss and gradient vanishing, and improving the prediction accuracy and efficiency of blood pressure parameters.
Smart Images
Figure CN117398081B_ABST