Indian Institute of Science Fetal Heart Sound Database (IIScFHSDB)

Abstract

The Indian Institute of Science Fetal Heart Sound Database (IIScFHSDB) has 60 fetal phonocardiography (fPCG) recordings obtained from expecting mothers (age 18–27 years) with gestation weeks 30–40. The recordings were made at St. John’s Hospital, Bangalore, using an electronic stethoscope. Data were collected from the antenatal and labor wards by placing the stethoscope on the lower abdomen of each subject. Each fPCG recording was acquired with a gain of 500X, has an average duration of 8 minutes, and was sampled at 2 kHz.

Background

Fetal monitoring during pregnancy and labor helps in identifying high-risk fetuses. The most commonly monitored parameter is the Fetal Heart Rate (FHR), from which clinically relevant features—such as baseline FHR, accelerations, and decelerations—can be derived. Various techniques exist for FHR determination, including fetal Doppler ultrasound, direct and indirect fetal electrocardiography (fECG), fetal magnetocardiography (fMCG), and fetal phonocardiography (fPCG). Devices based on fPCG are non-invasive, passive, simple, and user-friendly. However, extracting reliable FHR measurements from fPCG is challenging due to noise. This database provides fPCG recordings from a noisy clinical setting, enabling the development and evaluation of signal processing algorithms for fPCG denoising and FHR determination.

Methods

The fPCG recordings were obtained using an electronic stethoscope (SS30LA) connected to a data acquisition system (MP36, Biopac Systems Inc.) with a gain of 500X and a sampling frequency of 2 kHz. The gain of 500X indicates that the input signal is amplified 500 times before digitization. Recordings were made using Biopac Student Lab software and exported as .wav audio files for MATLAB analysis. Data were collected from pregnant women with gestation greater than 30 weeks and singleton pregnancies, without known maternal complications, at a location advised by a trained doctor.

Data Description

A total of 60 fPCG recordings were collected from 60 subjects, with an average recording time of around 8 minutes. The data files are stored in .wav format and can be analyzed using MATLAB. No filtering was applied to the data except for the Biopac’s built-in notch filter to remove line frequency interference.

The demographic and clinical information of the subjects is shown in the table below:

Each recording is labeled as subject_x. A .csv file is provided that contains information on each subject’s height, weight, gravida, age, clinical conditions (if present), and FHR values when available.

Usage Notes

The fPCG dataset was developed to encourage the development and evaluation of algorithms for monitoring fetal cardiac activity, especially under noisy conditions. In prior work [2], this dataset was used to develop an FHR extraction algorithm that employs autocorrelation and validation steps to accurately measure FHR for varying signal qualities. An example MATLAB code (example_code.m) is provided to demonstrate reading the audio file and applying a comb filter.

Release Notes

Version 1.0: Initial Release

Ethics

The study was approved by the Institutional Ethics Committee, St. John’s Hospital, Bangalore (IEC code No. 07/2018). Data were collected after obtaining informed consent.

Acknowledgements

We would like to thank Dr. Shirley George and the staff at the Department of Obstetrics and Gynaecology, St. John’s Hospital, Bangalore, for their support in data acquisition.

Conflicts of Interest

The authors declare no conflict of interest.

References

1. Ibrahim, Emad A., et al. "A comparative study on fetal heart rates estimated from fetal phonography and cardiotocography." Frontiers in Physiology 8 (2017): 764.
2. Amrutha B. et al., "Heart Rate Estimation and Validation Algorithm for Fetal Phonocardiography", Physiological Measurements.