Incontinence Event Detection Using Property Vector Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing incontinence detection systems fail to accurately distinguish between urinary and faecal incontinence events and often provide unnecessary alerts, leading to resource wastage and inefficiency in care institutions, as they struggle with erroneous signal processing and inability to determine event volume or significance.
Innovation Solution
A method involving the processing of sensor data in segments using Reference Property Vectors to identify wetness events, employing algorithms like K-nearest neighbours, support vector machines, and logistic regression to determine segment types and event characteristics, such as volume and duration, through the use of a moving window approach and relevancy ratios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing incontinence detection systems are used to detect wetness events, then staff are alerted to the occurrence of events, but the systems cannot distinguish between urinary and faecal incontinence events and provide erroneous results
Solution Approach 1:
The sensor signal is divided into segments representing different phases of incontinence events (pre-event, event, post-event). Each segment is analyzed separately using reference property vectors to classify event types, improving both detection accuracy and signal processing reliability by treating different event phases independently rather than as a continuous unanalyzed signal.
Solution Approach 2:
The system transforms raw sensor signals into property vectors that capture key characteristics of wetness events. By changing the parameter representation from raw sensor data to structured property vectors with multiple features, the system achieves more reliable event classification and distinction between urinary and faecal incontinence.
2Reliability
If existing detection systems alert carers to all detected events, then carers are notified of potential issues, but unnecessary alerts are generated leading to resource wastage
Solution Approach 1:
The system calculates event volume and significance parameters by analyzing property vectors. Alerts are generated based on these calculated parameters rather than all detected events, allowing the system to filter out insignificant events and reduce unnecessary resource utilization while maintaining reliable alert accuracy for genuine incontinence events.
3Reliability
If manual checks are conducted regularly to monitor incontinence events, then staff can detect events, but this places a significant burden on care institution resources and interrupts individual activities
Solution Approach 1:
The detection system operates autonomously using sensor data and automated property vector analysis to identify and classify incontinence events. The system serves itself by automatically processing signals, classifying events, and generating alerts without requiring manual intervention, thereby improving detection reliability while maintaining care institution productivity and avoiding interruptions to individual activities.
4Loss of information
If existing systems provide basic event detection, then simple wetness detection is achieved, but they fail to provide useful information about individual events such as volume or significance
Solution Approach 1:
The system transforms raw sensor signals into property vectors that encode multiple event characteristics including volume and significance. By changing from simple detection to multi-parameter property vector analysis, the system recovers complete event information without proportionally increasing processing complexity, as the property vector framework efficiently captures multiple event attributes simultaneously.
Data Source
AI summary
A method for analyzing incoming data, comprising the steps of processing the incoming data in segments to output a sequence of segment types by extracting one or more properties of an incoming data segment and forming an Unknown Property Vector for each segment of data in the incoming data, and processing the sequence of segment types to identify events in the incoming data. The sequence of segment types is determined, for each segment, by reference to a set of Reference Property Vectors that are relevant to the Unknown Property Vector. This may involve application of first and/or second and/or further functions to identify at least a first subset of Reference Property Vectors that are relevant to the Unknown Property Vector. Alternatively, a logistic regression algorithm, derived using clustering or classification methods for identifying candidate vectors, may be used.