Audio Signal Processing for Heart Rate Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current electronic systems for non-contact extraction of heart rate information from human voice audio signals are limited in accuracy and require training configurations, making them inefficient for real-time and adaptive applications.
Innovation Solution
An electronic system that processes audio signals in the time domain to calculate cardiovascular heartbeat information, including heart rate and variability, by detecting the fundamental frequency of vowel sounds, filtering noise, and generating time domain intermediate signals to capture frequency, amplitude, and phase, allowing for real-time heart rate monitoring without the need for training configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If non-contact sensing techniques are used to extract heart rate information from human voice audio signals, then contactless monitoring is achieved, but measurement precision and reliability are limited
Solution Approach 1:
The audio signal is segmented into multiple analysis windows, and the power spectral profile is computed for each window to identify fundamental frequency components. This segmentation allows the system to process the signal in manageable segments, improving measurement precision while maintaining non-contact monitoring capability.
Solution Approach 2:
The patent introduces an intermediary processing stage that computes the power spectral profile and identifies fundamental frequency components as intermediaries between the raw audio signal and the final heartbeat information. This intermediary approach filters out noise and enhances the reliability of heart rate extraction.
2Adaptability or versatility
If current electronic systems process audio signals to extract heart rate information, then non-contact monitoring is enabled, but the systems require training configurations and are not adaptive to different subjects
Solution Approach 1:
The system performs self-calibration by automatically detecting the fundamental frequency of the subject's voice through power spectral analysis. The denoising module automatically adapts to different voices without requiring manual training configurations, making the system self-sufficient and adaptable to individual subjects.
Solution Approach 2:
The system dynamically adjusts processing parameters based on the detected fundamental frequency of each subject's voice. By changing parameters such as filtering bandwidth and analysis window settings according to the individual's vocal characteristics, the system achieves adaptability without requiring complex training configurations.
3Productivity
If traditional methods process audio signals for heartbeat information, then analysis can be performed, but real-time processing capability is limited
Solution Approach 1:
The system performs preliminary actions by pre-computing the power spectral profile and identifying fundamental frequency components before the main heartbeat analysis. This preliminary processing enables the system to quickly analyze subsequent audio segments in real-time, reducing processing delay and improving productivity.
Solution Approach 2:
The system maintains continuous processing by continuously analyzing the audio signal in overlapping windows and continuously updating the heartbeat information. This continuous action eliminates interruptions and delays, enabling real-time monitoring while maintaining high productivity.
4Measurement precision
If the system processes audio signals to maintain phase information and provide beat-by-beat metrics, then measurement precision is improved, but computational complexity increases
Solution Approach 1:
The system applies local quality analysis by focusing the power spectral profile computation and fundamental frequency detection on specific local segments of the audio signal. This localized approach maintains phase information accuracy for each segment while reducing the overall computational complexity compared to analyzing the entire signal at once.
Data Source
AI summary
A device for calculating cardiovascular heartbeat information is configured to receive an electronic audio signal with information representative of a human voice signal in the time-domain, the human voice signal comprising a vowel audio sound of a certain duration and a fundamental frequency; generate a power spectral profile of a section of the electronic audio signal, and detect the fundamental frequency (F0) in the generated power spectral profile; filter the received audio signal within a band around at least the detected fundamental frequency (F0) and thereby generating a denoised audio signal; generate a time-domain intermediate signal that captures frequency, amplitude and/or phase of the denoised audio signal; detect and calculate heartbeat information within a human cardiac band in the intermediate signal.