Inhale Insight: A Data-Driven Approach to Breathing Pattern Analysis
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Respiratory illnesses continue to be a serious global health concern, making early and reliable diagnosis increasingly important. This work investigates how machine learning techniques can assist in identifying and classifying respiratory conditions through the analysis of breathing sounds. The proposed framework operates in three main stages: collecting and preprocessing respiratory audio signals, extracting important sound characteristics such as wheezes and crackles, and classifying diseases using a Convolutional Neural Network (CNN) model. The system was developed and tested using 920 annotated respiratory recordings obtained from 126 patients diagnosed with conditions such as COPD, lower respiratory tract infections (LRTI), and upper respiratory tract infections (URTI). The results show strong classification performance, suggesting that automated respiratory sound analysis can support early detection and continuous monitoring. Such a system has the potential to be used in real-time clinical settings, enabling faster treatment decisions while reducing healthcare costs. Overall, this approach demonstrates how machine learning can contribute to more accurate diagnoses and improved patient care.
Fields of Study & Subjects
Phonocardiography and Auscultation Techniques
Artificial Intelligence in Healthcare
Non-Invasive Vital Sign Monitoring
Publication details
- Journal / source
- International Conference on Sentiment Analysis and Deep Learning (ICSADL)
- Pages
- 974–980
- Publication date
- 18 February 2026
- Language
- EN