Learning Device for Noise-Resistant Vital Sign Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing devices that measure electrocardiographic waveforms, such as those using electrodes on a seat and steering wheel, often suffer from noise interference due to vibration and body movement, leading to decreased accuracy in vital data acquisition.
Innovation Solution
A learning device and method that utilize first sensor data as learning data, combined with second sensor data less affected by noise, and third sensor data indicating noise influence, to generate a trained model that outputs vital data with reduced noise, using supervised learning techniques like neural networks and support vector machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If first sensor data is acquired through a first method (e.g., electrodes on seat and steering wheel), then the burden on the subject is reduced and ease of operation is improved, but noise is likely to occur due to vibration and body movement, decreasing measurement precision
Solution Approach 1:
The patent introduces a learning model as an intermediary that processes the noisy first sensor data. The model is trained using paired data from both the first method (noisy but easy) and second method (clean but burdensome), enabling it to filter noise and produce accurate vital data while maintaining the ease of the first acquisition method
Solution Approach 2:
The patent transforms the raw sensor data parameters through a learned transformation. By training the model to map from the noisy parameter space of the first method to the clean parameter space of the second method, it changes the parameters in a way that removes noise while preserving the underlying vital information
2Productivity
If electrodes are placed on the seat and steering wheel for continuous monitoring, then productivity is improved through automated vital data collection, but noise from vibration and movement increases, worsening measurement precision
Solution Approach 1:
The learning model serves as an intermediary processing layer that enables automated data collection from the first sensor method while compensating for its noise issues. The model is trained to recognize and filter vibration and movement artifacts, allowing continuous monitoring without sacrificing accuracy
Solution Approach 2:
The patent replaces the direct mechanical measurement approach (electrodes physically contact subject) with a computational approach (learning model processes sensor signals). This substitution allows the system to maintain the mechanical simplicity and automation of the first method while achieving the precision normally requiring the second method
Data Source
AI summary
To more efficiently acquire vital data which is less affected by noise.Provided is a learning device including a learning unit that performs learning related to an output of vital data indicating vital signs of a subject by using first sensor data acquired from the subject through a first method as learning data and using training data based on second sensor data acquired from the subject in the same period as a period of acquisition of the first sensor data through a second method which is less affected by noise than the first method, wherein the learning unit performs learning further on the basis of third sensor data which is acquired in the same period as the period of acquisition of the first sensor data and the second sensor data and available as an index indicating a magnitude of influence of the noise occurring in the first sensor data.