An AI-based Lung Sound Classification System
Session Number
1
Advisor(s)
Lichuan Liu, Northern Illinois University
Location
A121
Discipline
Medical and Health Sciences
Start Date
15-4-2026 10:15 AM
End Date
15-4-2026 11:00 AM
Abstract
When screening for respiratory diseases, the primary form of diagnosis is auscultation, which, without proper training and experience, can yield unreliable or incorrect results. This presents an issue, as many parents lack the experience or knowledge to diagnose children at an early age. Within this SIR, I’ve explored various systems that are utilized or closely related to the development of an AI-based classification system and digital signal processing. Included in these techniques, I’ve learned to use MATLAB to graph, analyze auditory signals, and design filters for these signals. I’ve researched and read about past systems used by both my mentor and other researchers on respiratory signal classification. As a part of this project, I’ve been given access to multiple open-source databases by my mentor, which include various auditory recordings of infants and young children breathing. I’ve learned to trim these files into only the necessary audio for a trained AI system.
An AI-based Lung Sound Classification System
A121
When screening for respiratory diseases, the primary form of diagnosis is auscultation, which, without proper training and experience, can yield unreliable or incorrect results. This presents an issue, as many parents lack the experience or knowledge to diagnose children at an early age. Within this SIR, I’ve explored various systems that are utilized or closely related to the development of an AI-based classification system and digital signal processing. Included in these techniques, I’ve learned to use MATLAB to graph, analyze auditory signals, and design filters for these signals. I’ve researched and read about past systems used by both my mentor and other researchers on respiratory signal classification. As a part of this project, I’ve been given access to multiple open-source databases by my mentor, which include various auditory recordings of infants and young children breathing. I’ve learned to trim these files into only the necessary audio for a trained AI system.