Multivariate Spectral Analysis of Transabdominally Recorded Intrauterine Acoustic Signals (TRIAS) Across Gestation

Area of Science:

• Biomedical Engineering
• Maternal-Fetal Medicine
• Acoustics

Background:

• Understanding the intrauterine acoustic environment is crucial for monitoring fetal well-being.
• Existing methods for intrauterine sound recording are often invasive or limited in scope.

Purpose of the Study:

• To develop and validate a multivariate analytical framework for transabdominally recorded intrauterine acoustic signals (TRIAS).
• To evaluate spectral changes in TRIAS across different stages of gestation.
• To explore the relationship between acoustic signals and maternal/fetal physiological factors.

Main Methods:

• Acquisition of 60 TRIAS recordings from 44 pregnant women (≥14 weeks) using a digital stethoscope.
• Calculation of Power Spectral Density (PSD) in the 30-2000 Hz range.
• Analysis using Ordinary Least Squares (OLS) regression, Principal Component Analysis (PCA), and Partial Least Squares Regression (PLSR).

Main Results:

• A consistent spectral peak was identified between 300-400 Hz.
• OLS regression showed a significant PSD decrease within 400-600 Hz during the second trimester (p<0.05).
• PCA indicated maternal heart rate as a major factor, while PLSR demonstrated independent associations of cardiovascular dynamics and gestational progression with acoustic features.

Conclusions:

• TRIAS analysis provides a non-invasive window into the maternal-fetal acoustic environment.
• Observed second-trimester spectral changes may indicate alterations in amniotic fluid dynamics or uterine structure.
• TRIAS holds potential as a baseline for future intrauterine monitoring research.