Sleep Motion Monitoring for Early Alzheimer’s Transition Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing Alzheimer's disease (AD) are applied too late in the prodromal stage of amnestic mild cognitive impairment (MCI) and do not leverage sleep patterns for early detection, lacking effective automatic systems to monitor and analyze sleep data for AD progression.
Innovation Solution
A wearable motion tracking device with a high-frequency inertial sensor and computing module uses continuous wavelet transform and a convolutional neural network to analyze sleep patterns, detecting transitions from healthy to AD-related states by generating probability reports and warnings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional diagnostic methods (psychological interviews, lumbar puncture) are used to detect Alzheimer's disease, then diagnostic accuracy can be achieved, but the detection occurs too late in the disease progression to enable preventive intervention
Solution Approach 1:
The system performs preliminary detection of Alzheimer's disease by analyzing sleep patterns during the prodromal stage (MCI phase), before traditional diagnostic methods would be applied. The inertial sensor continuously monitors sleep behaviors and the neural network detects early signs of cognitive deterioration, enabling intervention before full-blown AD develops.
2Measurement precision
If no automated sleep monitoring system is implemented, then device complexity remains low, but the ability to detect prodromal AD through sleep pattern analysis is lost
Solution Approach 1:
The system uses an autonomous inertial sensor that automatically collects, processes, and analyzes sleep data without requiring manual intervention. The embedded neural network performs real-time classification of sleep patterns, and the system self-generates alerts when prodromal AD is detected, eliminating the need for complex external monitoring infrastructure.
3Measurement precision
If continuous high-frequency monitoring of sleep patterns is implemented, then early detection of AD progression is enabled, but energy consumption and device resource usage increase
Solution Approach 1:
The inertial sensor operates continuously at high frequency throughout the night to capture complete sleep cycle data, enabling accurate detection of prodromal AD markers. The system maintains continuous monitoring without interruption, processing data in real-time through the neural network to ensure no early signs are missed.
Data Source
Figure 1
AI summary
The invention relates to a system for detecting the prodromal development of Alzheimer's disease from sleep patterns of a user (2), said system comprising: at least one inertial sensor (1) adapted to continuously measure an acceleration pattern (3) during a period of time, and wherein the acceleration pattern (3) comprises a set of data associated to the user's (2) movements during sleep; wherein the inertial sensor (1) comprises means for transmitting the data associated to the acceleration patterns (3) to a computing module (4). Advantageously, the computing module (4) is configured with a neural network (7) comprising a reference dataset obtained from the acceleration patterns (3) belonging to both healthy and patients of Alzheimer's disease, wherein the computing module (4) is further configured to obtain, by means of the neural network (7), a probability (8) of transitioning to a state characterised by Alzheimer's disease by comparison between the user's (2) acceleration pattern (3) and the reference dataset.