Non-Invasive Sleep Activity Index via Audio-Video Stream Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional polysomnography techniques require numerous sensors, which disrupt patients' natural sleep patterns and are unsuitable for newborns or young children, making it difficult to accurately capture physiological variables.
Innovation Solution
A method that calculates an activity index using an audio/video stream, combining data streams such as motion and sound to create a normalized index representative of an individual's level of activity, without the need for complex hardware or additional sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If numerous sensors are applied to the patient's body for polysomnography measurements, then physiological variables can be measured, but the patient cannot achieve natural sleep conditions
Solution Approach 1:
The invention extracts the useful measurement function from traditional contact sensors and relocates it to a camera-based system. The camera captures physiological information (breathing movements, eye movements, body position) remotely without physical contact, thereby eliminating the disruptive effect of numerous sensors on natural sleep while preserving measurement capabilities
Solution Approach 2:
The invention replaces the mechanical sensor system (electrodes, straps, and other contact-based devices) with an optical measurement system using a camera. This substitution eliminates the need for physical attachment to the patient's body, allowing natural sleep conditions to be maintained while still capturing physiological data through visual observation of body movements and facial expressions
2Measurement precision
If traditional sensors are used for monitoring newborns or young children, then physiological variables can be captured, but the measurement conditions become more complex and require continuous recording with limited sensors
Solution Approach 1:
The invention creates a universal measurement system that functions across multiple patient populations (adults, children, newborns) and multiple measurement scenarios (clinical settings, home environments, day and night). The camera-based system provides multi-functional capability to capture various physiological indicators through a single device, simplifying the measurement process for vulnerable populations like newborns who require continuous monitoring in diverse conditions
3Device complexity
If an audio/video stream is used for sleep analysis, then the system can be simple and non-invasive, but the ability to capture detailed physiological variables is limited
Solution Approach 1:
The invention segments the audio/video stream into multiple independent analysis channels, each extracting specific physiological information. The video component is divided into analysis of breathing movements, eye movements, body position, and other discrete physiological indicators. This segmentation allows the simple camera-based system to capture detailed physiological variables by processing different aspects of the visual data through specialized algorithms
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for calculating an activity index of an individual in a sleep phase, which is implemented via an index calculating device. According to the invention, such a method includes: - a step (10) of obtaining from an initial audio/video stream (FlvO) at least two streams (x1, x2, x3), each including a plurality of time stamped samples; - a step (20) of formatting said at least two streams (x1, x2, x3), outputting at least two formatted streams (x1f, x2f, x3f); a step (30) of combining said at least two formatted streams, outputting a combined stream (xc); - a step (40) of calculating, from said combined stream (xc), at least one index representative of an activity level of an individual (IrNA).