Sound based respiratory rate estimation

• Main Author: Lee, Ashley Hui Qing
• Other Authors: Ser Wee
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/77995
• Institution: Nanyang Technological University

Respiratory rate serves as an important vital sign to indicate the state of patient’s health. It is often measured to detect early diseases and monitor deteriorations in patients. However, due to the inadequacies of existing methods of respiratory rate measurement, there is a need to develop better estimation methods for simple and cost-effective monitoring. Recently, there have been more studies on non-invasive measures in the near environment of the patients. Respiratory rate estimation based on sound signals is one area explored. While existing sound based algorithms reflect positive performance, few examined the effect of varied conditions on its performance. The objective of this final year project is to develop a sound based respiratory rate estimation algorithm and study its performance under various conditions to improve its accuracy. The algorithm was evaluated using open source audio samples. The findings of this study revealed strong performance for the algorithm. The correlation coefficient between the estimated and reference respiratory rate is 0.98, with an average percentage accuracy of 95.15% across the test samples. The findings also concluded that the variation of parameters and injection of noise reflected an effect on the algorithm performance.