Consistency of Interpretation of Lung Sounds between Experienced Physicians and Automatic Analysis Using a Newly Developed Algorithm Based on the Acoustic Characteristics.
Listening and interpreting lung sounds by a stethoscope has been an important component of screening and diagnosing lung diseases. However, this practice is vulnerable to inter-observer variations and difficulty in comparison with previous findings. In this study, we aimed to examine the consistency of interpretation of lung sounds between experienced physicians and automatic analysis using a newly developed algorithm based on the acoustic characteristics.
The auscultation of the respiratory system has been an important component of screening and diagnosing various lung diseases. Although it is inexpensive and non-invasive, the practice is vulnerable to poor quantitatively and reproducibility and inter-observer variations in interpretation. Also, it is sometimes difficult for physicians to evaluate temporal changes in lung sounds, such as before and after treatment. However, it remains unclear how consistent the interpretation is when lung sounds are repeatedly heard by a physician or when evaluated by multiple physicians. In addition, since interpretation of lung sounds can be different between observers, a tool for sharing lung sounds among healthcare workers would be helpful.
In this study, we aimed to examine the consistency of interpretation of lung sounds between live and recorded auscultation and between experienced pulmonologists. We also compared physicians’ evaluation of lung sounds and automatic analysis using a newly developed algorithm based on the acoustic characteristics.
Pulm Res Respir Med Open J. 2016; SE(1): S8-S11. doi:10.17140/PRRMOJ-SE-1-103