Abstract A new lung sound analysis software program has been developed. It can automatically select a typical lung sound spectrogram and calculate lung sound parameters using machine learning programs. This study aimed to clarify lung sound characteristics in early infants using this program. Using the program, the characteristics of lung sounds in healthy 1- and 4-month-old infants were examined. The lung sounds were assessed in the supine position for 1-month-old infants and in both the sitting and supine positions for 4-month-old infants. We compared the characteristics of the infant lung sounds with those of healthy 3-year-old children. The lung sound parameters of the 1-month-old infants (n = 58) were affected by gender, height, and birth weight. However, those of the 4-month-old infants (n = 50) obtained in the sitting or supine position were not affected by these factors in the study. The lung sound parameters obtained in the sitting and supine positions were not significantly different, and they were not related to a history of wheezing or allergy. PAP0 was higher for the 1-month-old infants than for the 4-month-old infants, and RPF50p and RPF75p were also higher for the 1-month-old infants than for the 4-month-old infants. The PAP0, FAP0, RPF50p, RPF75p, A3a/AT, and B4a/AT of the 4-month-old infants were significantly higher than those of the 3-year-old children (n = 80). Conclusion The new program confirmed the specificity of lung sound parameters in early infants. Further studies are needed to clarify why the specificity differs from that of 3-year-old children. <Table Float="No" ID="Taba"> <tgroup cols="1"> <colspec align="left" colname="c1" colnum="1" /> <tbody> <row> <entry align="left" colname="c1"> What is Known: • Lung sound analysis has been evaluated to some extent as safe and straightforward for estimating lung function in toddlers and older humans. However, it has not been applied to early infants. </entry> </row> <row> <entry align="left" colname="c1"> What is New: • A lung sound analysis method using new analysis software has clarified the characteristics of lung spectra in early infants. This technique may facilitate the early diagnosis and treatment of respiratory diseases in infants. </entry> </row> </tbody> </tgroup> </Table>

Characteristics of lung sounds in early infants using automated analysis

期刊论文
出版模式：

开放获取
发表日期：

2025年5月13日

查看全文

Yoshifumi Murayama,
Hiroyuki Mochizuki,
Hidetoshi Yano,
Shigeki Ochiai,
Mayumi Enseki,
Takashi Koike,
Hiroyuki Furuya,
Yoshiyuki Yamada,
Atsushi Uchiyama

摘要

Abstract

A new lung sound analysis software program has been developed. It can automatically select a typical lung sound spectrogram and calculate lung sound parameters using machine learning programs. This study aimed to clarify lung sound characteristics in early infants using this program. Using the program, the characteristics of lung sounds in healthy 1- and 4-month-old infants were examined. The lung sounds were assessed in the supine position for 1-month-old infants and in both the sitting and supine positions for 4-month-old infants. We compared the characteristics of the infant lung sounds with those of healthy 3-year-old children. The lung sound parameters of the 1-month-old infants (n = 58) were affected by gender, height, and birth weight. However, those of the 4-month-old infants (n = 50) obtained in the sitting or supine position were not affected by these factors in the study. The lung sound parameters obtained in the sitting and supine positions were not significantly different, and they were not related to a history of wheezing or allergy. PAP₀ was higher for the 1-month-old infants than for the 4-month-old infants, and RPF_50p and RPF_75p were also higher for the 1-month-old infants than for the 4-month-old infants. The PAP₀, FAP₀, RPF_50p, RPF_75p, A_3a/AT, and B_4a/AT of the 4-month-old infants were significantly higher than those of the 3-year-old children (n = 80).

Conclusion

The new program confirmed the specificity of lung sound parameters in early infants. Further studies are needed to clarify why the specificity differs from that of 3-year-old children.

What is Known:

• Lung sound analysis has been evaluated to some extent as safe and straightforward for estimating lung function in toddlers and older humans. However, it has not been applied to early infants.

What is New:

• A lung sound analysis method using new analysis software has clarified the characteristics of lung spectra in early infants. This technique may facilitate the early diagnosis and treatment of respiratory diseases in infants.