ECG Authentication Using QRS Wave Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ECG-based authentication methods face challenges in achieving high-level accuracy due to variations in ECG readings caused by different physiological states, age, and health conditions, leading to low discrimination features and accuracy in authentication.
Innovation Solution
A method that synchronously trains a dictionary and a classifier using group sparse coding to generate feature vectors from ECG signals, allowing for high-level discrimination and accurate authentication by dividing ECGs into signal segments, extracting features, and performing dimension reduction processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional biometric authentication methods are used, then authentication accuracy is improved, but security is weakened due to forged or falsified biometric features
Solution Approach 1:
The patent extracts the QRS wave complex from the ECG signal as a specific feature for authentication. By isolating and analyzing this particular waveform component, the system achieves both high measurement precision for authentication and enhanced security, as the QRS wave's morphological features are difficult to forge compared to other biometric traits
Solution Approach 2:
The patent transforms the ECG signal into a feature vector by extracting morphological parameters of the QRS wave complex, such as amplitude, duration, and waveform shape. This parameter transformation converts raw physiological data into discriminative authentication features that maintain high accuracy while providing security against forgery
2Adaptability or versatility
If ECG variations due to physiological states, age, and health conditions are considered, then authentication robustness is improved, but discrimination feature quality deteriorates
Solution Approach 1:
The patent applies local quality by focusing analysis on specific segments of the ECG signal - the QRS wave complex - rather than the entire signal. By concentrating on this localized feature that exhibits consistent morphological characteristics across different physiological states, the system maintains high discrimination quality while achieving robustness to variations
Solution Approach 2:
The patent segments the ECG signal into distinct components, specifically isolating the QRS wave complex for authentication purposes. This segmentation allows the system to focus on the most discriminative and stable portion of the signal, maintaining feature quality while being adaptable to physiological variations that affect other parts of the ECG
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are electrocardiogram (ECG)-based authentication and training. An authentication method includes generating a feature vector of an ECG obtained from an entity or a person based on a dictionary, classifying the ECG through a classifier based on the feature vector, and performing authentication based on a classification result.