This invention discloses a
dynamic assessment and intervention method for
patient fall risk based on multi-
source data fusion. It constructs a complete
system for managing
fall risk in
hospitalized patients, encompassing basic data collection, personalized
dynamic assessment, multi-dimensional real-time monitoring, individualized intervention, and cross-departmental closed-loop management. By connecting to hospital EMR, HIS systems, and IoT devices, it automatically collects multi-
source data on patient
baseline data, medication, physiological data, and
environmental data. After cleaning and fusion, features are extracted. A three-layer indicator
system and LSTM temporal neural network are used to dynamically assess risk. Basic and personalized intervention plans are matched according to
risk level, and tasks are pushed through multiple channels while tracking effects. Relying on a cloud-edge collaborative cross-departmental platform, it achieves end-to-end
data linkage, early warning push, and plan optimization, adapting to the needs of the
entire population and across departments, improving assessment accuracy and intervention execution rate, shortening early warning
lag time, freeing up
nursing manpower, and forming a continuously improving closed-loop management mechanism.