Breathing Sound Diagnosis Device with Parameter Importance Quantification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disease diagnosis models using deep neural networks can detect abnormal breathing sounds but fail to provide detailed basis for disease classification, making it difficult for medical institutions to trust the classification results.
Innovation Solution
A diagnosis device that generates diagnosis information and basis information by obtaining breathing sound data, analyzing it to extract feature parameters, and applying these parameters to a pre-trained differential diagnosis model to classify diseases and quantify the importance of each parameter, providing output through a user interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a deep neural network model is used to detect abnormal breathing sounds and classify diseases, then diagnosis accuracy is improved, but the ability to provide detailed classification basis is worsened
Solution Approach 1:
The patent introduces an intermediary module that generates diagnosis basis information by analyzing the relationship between input breathing sound data and the model's classification output. This intermediary component extracts and presents relevant feature parameters and their contributions to the diagnosis, bridging the gap between the black-box neural network and interpretable medical decision-making without compromising diagnosis accuracy.
2Loss of information
If multiple feature parameters are analyzed to provide detailed diagnosis basis, then information completeness is improved, but device complexity is worsened
Solution Approach 1:
The patent extracts only the essential feature parameters and their contribution metrics from the complex neural network processing, presenting a simplified set of diagnostic basis information. This extraction approach selectively outputs the most relevant parameters (such as abnormal breathing sound characteristics and their temporal-frequency features) without requiring the full computational complexity of the underlying model, thereby reducing system complexity while maintaining information completeness.
Data Source
AI summary
Disclosed is a method of operating a diagnosis device for diagnosing a disease of a user, which includes obtaining breathing sound data by measuring a breath of the user, generating feature parameter information including a plurality of parameters corresponding to at least one abnormal breathing sound based on an analysis of the breathing sound data, generating diagnosis information indicating a target disease with a highest model output value among a plurality of diseases by applying the feature parameter information to a pre-trained differential diagnosis model, generating diagnosis basis information that quantifies an importance of the plurality of parameters used to determine the target disease, and outputting the diagnosis information and the diagnosis basis information through a user interface device of the diagnosis device.