Method for providing fall risk assessment by using deep learning algorithm and device thereof
A CNN-LSTM hybrid deep neural network with IMUs preprocesses data to classify fall risks accurately, addressing the need for portable and real-time fall risk assessment, improving safety for the elderly and patients.
WO2026141823A1PCT designated stage Publication Date: 2026-07-02SOONCHUNYANG UNIV IND ACAD COOP FOUND
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- SOONCHUNYANG UNIV IND ACAD COOP FOUND
- Filing Date
- 2025-07-29
- Publication Date
- 2026-07-02
Smart Images
Figure KR2025011275_02072026_PF_FP_ABST
Abstract
The present invention relates to a method for providing fall risk assessment by using a deep learning algorithm and a device therefor, the method being performed by a fall risk assessment device and comprising the steps of: loading data from at least one user terminal; pre-processing the loaded data; training a deep learning algorithm on the basis of the pre-processed data; inputting input data to the trained deep learning algorithm; calculating fall risk assessment data on the basis of output data output from the trained deep learning algorithm; and providing the calculated fall risk assessment data to the user terminal, wherein the fall risk assessment data is data classified into two categories of high risk and low risk with respect to a user's fall, and the step of loading data from the user terminal is loading data collected from two IMU sensors disposed on the user's body legs.
Need to check novelty before this filing date? Find Prior Art