Acoustic Speech Segmentation for Physiological State Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately assess the physiological state of a subject based on speech analysis, particularly in identifying conditions such as congestive heart failure, coronary heart disease, atrial fibrillation, chronic obstructive pulmonary disease, asthma, interstitial lung disease, pulmonary edema, Parkinson's disease, bipolar disorder, and other psychological disorders, due to limitations in time-normalization and feature extraction methods.
Innovation Solution
A system and method that maps a test speech sample to a reference speech sample or a speech model, dividing both into segments based on acoustic properties, computes distances between these segments, and communicates an output indicating the physiological state of the subject, using dynamic time warping and Hidden Markov Models or deep neural networks to align and quantify deviations in speech patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional time-normalization algorithms are used for speech analysis, then the processing speed is maintained, but the accuracy in detecting physiological states deteriorates
Solution Approach 1:
The speech sample is divided into multiple segments based on acoustic features, allowing localized analysis of different portions of the speech signal. This segmentation enables more precise detection of physiological state indicators while maintaining computational efficiency by focusing on relevant segments rather than processing the entire signal uniformly.
Solution Approach 2:
The system employs dynamic time warping to align speech samples of varying durations and characteristics. This dynamic approach adapts the analysis to the specific temporal structure of each speech sample, improving accuracy in detecting physiological states without requiring rigid fixed-duration processing windows.
2Reliability
If speech samples are divided into segments based on acoustic properties, then the detection accuracy of physiological conditions is improved, but the computational complexity increases
Solution Approach 1:
Different acoustic features are extracted and analyzed for different segments of the speech sample based on their local characteristics. This allows the system to focus computational resources on segments and features most relevant to specific physiological conditions, improving reliability while optimizing resource usage by avoiding uniform high-cost processing across the entire signal.
3Measurement precision
If dynamic time warping is used to align speech samples, then the accuracy in comparing speech patterns is improved, but the processing time increases
Solution Approach 1:
Speech samples are pre-processed to extract acoustic features and identify key segments before performing dynamic time warping alignment. This preliminary action reduces the dimensionality and complexity of the alignment task, improving speech pattern comparison accuracy while reducing the computational time required for the DTW process.
Data Source
AI summary
A method includes mapping, by a processor, a test speech sample, which was produced by a subject while a physiological state of the subject was unknown, to a reference speech sample, which was produced in a known physiological state. The method further includes, based on the mapping, computing a distance between the test speech sample and the reference speech sample, and in response to the distance, communicating an output indicating the physiological state of the subject while the test speech sample was produced. Other embodiments are also described.