Wearable Motion Sensor Fusion for Fall Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI-based systems primarily focus on detecting falls after they have occurred, rather than predicting them by analyzing continuous user behavior and quantifying physiological changes.
Innovation Solution
A system utilizing multiple wearable motion sensors that include a computational method and machine learning models to detect and predict falls by analyzing user data through a two-dimensional transform and coherence analysis, classifying activities into clusters to identify potential falls.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple wearable motion sensors are used to detect falls, then measurement precision and reliability are improved, but device complexity increases
Solution Approach 1:
The system divides the fall detection task into separate processing streams for each sensor, with each sensor independently analyzing its own motion data through clustering algorithms. This segmentation allows multiple sensors to be used without proportionally increasing overall system complexity, as each sensor processes data independently rather than requiring complex inter-sensor coordination.
Solution Approach 2:
The system combines results from multiple independent sensor analyses through a voting mechanism, where each sensor's classification is aggregated to reach a final fall detection decision. This merging approach leverages the improved measurement precision of multiple sensors while using a simple voting rule rather than complex integration algorithms to maintain manageable system complexity.
2Measurement precision
If machine learning models are trained on user-specific data, then measurement precision is improved, but loss of time for data collection and training increases
Solution Approach 1:
The system performs preliminary unsupervised clustering analysis on user motion data during a training phase to automatically identify activity-specific clusters without requiring manual labeling. This preliminary action creates a personalized activity model that improves subsequent fall detection precision while minimizing the time investment required, as the clustering algorithm automatically adapts to individual user patterns without extensive manual data collection or expert annotation.
Data Source
AI summary
In an example system, multi-sensor motion data that reflects the motion of a user over a period of time is received as recorded by a plurality of motion sensors on wearable devices, specific changes are determined in the data by, firstly, quantifying it using a two-dimensional data transform; secondly, extracting an anomalous area; and thirdly, calculating a number of the anomaly's properties, and the results are input into a machine learning model to detect that the user fell. The machine learning model processes the anomaly's properties and evaluates the current state of the user's activity by classifying the properties against a state space previously calculated by analyzing historical activities of daily living. Depending on a two-dimensional transform implemented, the machine learning model detects when the user falls, as well as potentially allows predicting that the user will suffer a fall in advance of the actual event, aiming at solving the stroke prediction problem.