Respiratory sound classification : sensor design

• Main Author: Goh, Xin Ni
• Other Authors: Ser Wee
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Electrical and electronic engineering::Control and instrumentation::Medical electronics
• Online Access: https://hdl.handle.net/10356/149535
• Institution: Nanyang Technological University

Pulmonary Edema is caused by the extravascular movement of fluid into the pulmonary interstititium and alveoli. This condition is a precursor to diseases that are principal causes of death such as Pneumonia and Chronic Obstructive Lung Diseases/Chronic Obstructive Pulmonary Disease. With the introduction of chest auscultation and later, the development of the stethoscope, practitioners had a more objective means in detecting Pulmonary Edema through the detection of adventitious sounds when patients inspire. To maintain the benefits that a stethoscope can offer and overcome the limitations of other more advanced and expensive detection methods, research on the development of an electronic stethoscope with acoustic-based techniques are conducted and gaining traction within the medical research community. This project aims to develop the electronic stethoscope chest piece using 3D printing technique and explore the various sensor housing designs based on the replication of an actual stethoscope chest piece. The project utilises the Autodesk Fusion 360 software to develop a total of 16 sensor housing design renderings, across 4 Series and 4 Types. The 16 sensor housing designs are 3D printed over five 3D printing sessions at both the Garage@EEE and SPMS Making and Tinkering Space. To facilitate the 3D printing process, the MakerBot Application software was used to convert the Fusion 360 renderings to printable files for 3D printing. Finally, the 3D printed models undergo a test where audio clips are recorded through the sensor housing and compared against the baseline audio clips to determine the sensor housing with the highest accuracy of sound detection and replication. To record and analyse the audio clips, the Audacity software was used. The project concluded that all sensor housing designs were able to replicate the baseline audio clips of varying quality. It was deemed that the Series 4 Type C sensor housing design has the best sensor design as it was able to replicate the baseline audio clips accurately and of high degree. The sensor design could potentially be explored further such as with the use of different materials etc.