Breath Sound Analysis Using Frequency-Shifted Reference Spectra

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing technologies face difficulties in accurately distinguishing and analyzing multiple abnormal breath sounds that overlap on the same frequency axis, making it challenging to diagnose health conditions effectively.

Innovation Solution

A biological sound analyzing apparatus and method that involves obtaining frequency information from biological sounds, shifting reference spectra based on this information, and calculating the ratio of each sound type using frequency-shifted reference spectra to accurately analyze and output the composition of mixed breath sounds.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional spectrum analysis methods are used to distinguish abnormal breath sounds, then normal breath sounds can be identified, but multiple overlapping abnormal breath sounds cannot be accurately divided and detected

Engineering Contradiction:

Improvedetection accuracy of abnormal breath soundsVSAvoidability to handle multiple overlapping sound types

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent segments the complex spectrum analysis problem into multiple steps: obtaining frequency information, shifting reference spectra based on frequency characteristics, and calculating ratios for each sound type separately. This segmentation allows the system to handle multiple overlapping abnormal breath sounds by processing them through distinct analytical stages rather than attempting to identify all sounds simultaneously.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter of reference spectra by applying frequency shifts based on obtained frequency information. This parameter transformation allows the reference spectra to adapt to the actual frequency characteristics of the biological sounds being analyzed, enabling accurate detection of multiple sound types that would otherwise overlap and be indistinguishable using fixed reference spectra.

Inventive Principle:

Principle #35Parameter changes

2Device complexity

If fixed reference spectra are used for sound classification, then the classification process is simple, but accurate analysis of mixed breath sounds with different frequency characteristics is difficult

Engineering Contradiction:

Improvecomplexity of reference spectrum processingVSAvoidaccuracy of sound type ratio calculation

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent transforms the static reference spectra into dynamic references by shifting their frequency positions based on the actual frequency characteristics obtained from the biological sounds. This dynamic adjustment allows the reference spectra to adapt to varying frequency conditions while maintaining a relatively simple processing framework, thus improving accuracy without significantly increasing system complexity.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent modifies the frequency parameter of the reference spectra by applying shifts calculated from the obtained frequency information. This parameter change enables the reference spectra to match the actual frequency characteristics of the biological sounds being analyzed, significantly improving the accuracy of sound type ratio calculations while maintaining computational efficiency.

Inventive Principle:

Principle #35Parameter changes

3Quantity of substance

If multiple abnormal breath sounds overlap on the same frequency axis, then the sounds mix together, but conventional methods cannot divide and detect each sound type individually

Engineering Contradiction:

Improvenumber of detectable sound typesVSAvoiddifficulty of separating overlapping sounds

Core Design Contradiction:

Quantity of substanceVSDifficulty of detecting and measuring

Solution Approach 1:

The patent applies local quality by calculating the ratio of each reference spectrum independently at different frequency positions. By obtaining frequency information and shifting reference spectra locally according to these frequency characteristics, the system can distinguish and quantify each sound type even when they overlap on the same frequency axis, effectively increasing the number of detectable sound types without proportionally increasing detection difficulty.

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS10441243B2Biological sound analyzing apparatus, biological sound analyzing method, computer program, and recording medium

Publication Date: 2019.10.15 AIR WATER INC

AI summary

A body sound analysis device includes: acquisition means (210, 220) that acquires information relating to a frequency corresponding to a prescribed feature of a spectrum of a body sound; shifting means (230) that shifts, according to the information relating to the frequency, a plurality of reference spectrums that are references for classifying body sounds, and acquires a frequency shift reference spectrum; and output means (240, 250) that outputs a ratio of a plurality of reference spectrums included in the body sounds on the basis of the body sounds and the frequency shift reference spectrum. According to this body sound analysis device, a plurality of types of sounds included in the body sounds can be favorably analyzed.