Preliminary study for the personal handheld device based snoring detection in ordinary sleep situation

Area of Science:

• Sleep Science
• Acoustics
• Biomedical Engineering

Background:

• Snoring is a key indicator of sleep disorders, impacting daily life quality.
• Accurate snoring detection is crucial for effective sleep disorder management.
• Previous research often lacks real-world sleep environment data.

Purpose of the Study:

• To characterize common sleep environment noises.
• To establish a dataset for preliminary snoring detection research.
• To explore the feasibility of using smartphones for sleep sound recording.

Main Methods:

• Utilized smartphone built-in sound recording in a typical bedroom setting.
• Designed an experimental protocol including common sleep-disrupting noises (e.g., cough, alarm, fan).
• Recorded sound data from 10 normal subjects during actual sleep.

Main Results:

• Acquired a dataset comprising 44 snoring instances.
• Collected a dataset of 75 distinct non-snoring sleep environment noises.
• Established a preliminary dataset for analyzing snoring and background sounds.

Conclusions:

• Smartphone recording is a practical method for capturing sleep sounds.
• Characterizing ambient sleep noises is essential for accurate snoring detection.
• The collected dataset serves as a valuable resource for future sleep disorder research.