Pressure Support Mask Identification Using Acoustic Spectrum Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge is to accurately identify the mask type, size, and brand used by patients with obstructive sleep apnea syndrome (OSAS) during pressure assisted ventilation support, as patients often switch masks without clinician knowledge, affecting data understanding, reimbursement, and device management.
Innovation Solution
A system utilizing a trained machine learning model to analyze breathing sounds, specifically exhalation and inhalation sounds, to automatically identify the mask type, size, and brand by generating acoustic spectrum data and employing a convolutional neural network (CNN) for classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If patients are supplied with a fit pack containing all mask sizes, then ease of operation is improved, but loss of information worsens because clinicians and DMEs cannot know which mask the patient is actually using
Solution Approach 1:
The system captures acoustic signals from the patient's breathing through the mask and uses machine learning to identify the mask type, size, and brand. This feedback loop automatically provides mask identification information to clinicians and DMEs, resolving the information loss problem while maintaining the ease of use benefits of fit packs
Solution Approach 2:
The patent replaces manual mask identification methods (visual inspection, patient reporting) with acoustic sensing and machine learning-based automatic identification. The system uses acoustic spectrum analysis of breathing sounds to determine mask characteristics, substituting mechanical/visual identification with acoustic field-based detection
2Adaptability or versatility
If patients can switch mask sizes without restriction, then adaptability is improved, but measurement precision worsens because the actual mask type cannot be accurately determined
Solution Approach 1:
The system replaces unreliable manual tracking of mask changes with acoustic-based automatic identification. By analyzing the acoustic signature of breathing sounds transmitted through different mask types, the system precisely identifies the actual mask being used, maintaining measurement precision despite patient flexibility in mask selection
Solution Approach 2:
The machine learning model analyzes multiple acoustic parameters (frequency spectrum, amplitude, temporal patterns) of breathing sounds to identify mask characteristics. By monitoring changes in these acoustic parameters, the system accurately determines mask type, size, and brand even when patients switch masks freely
3Device complexity
If manual tracking of mask changes is used, then device complexity is reduced, but loss of information worsens because clinicians cannot remotely know mask type for data interpretation
Solution Approach 1:
The system performs self-identification of mask type without requiring external intervention. The PAP device automatically captures acoustic signals, processes them through the machine learning model, and identifies the mask being used, eliminating the need for manual tracking while providing complete mask information to clinicians
Solution Approach 2:
The system implements automatic feedback by continuously monitoring acoustic signals and providing real-time mask identification information to the clinical team. This feedback mechanism ensures clinicians have accurate mask usage data for interpreting patient data and making treatment decisions
Data Source
AI summary
A system and associated method for automatically identifying a mask used in a pressure support system for delivering a flow of breathing gas to the airway of a patient. The system includes a controller implementing a trained machine learning model. The controller is structured and configured to receive a sound signal, the sound signal being indicative of breathing sounds (e.g., exhalation and/or inhalation sounds) captured from the patient during use of the mask in the pressure support system, generate acoustic spectrum data indicative of an acoustic spectrum of the exhalation sounds based on the sound signal, provide the acoustic spectrum data to the trained machine learning model, and determine a brand, type and/or size of the mask in the trained machine learning model based on the provided acoustic spectrum data.