Sleep evaluation using audio signal processing

Yaniv Zigel, Ariel Tarasiuk, Eliran Dafna

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

1 Scopus citations


Millions of people worldwide experience sleep disorders; poor sleep is associated with increased risk for accidents, morbidity, and mortality. Early diagnosis and treatment can improve quality of life and health. However, the number of sleep testing sites is limited, and many people are left undiagnosed and untreated. The biomedical engineering field of sleep evaluation is on a “fast track” toward simple home-based testing. Existing technology for home-based sleep evaluation uses body-contact sensors that may lead to poor data acquisition and may also disturb sleep. Recently, new non-contact approaches were suggested for a simple sleep evaluation.

In this chapter, several audio-based signal processing algorithms for sleep evaluation are reviewed. These algorithms evaluate snoring and breathing, sleep-wake stages, sleep quality parameters (such as total sleep time), and obstructive sleep apnea. Our findings show that sleep and respiratory activity can be estimated by analysis of audio signals using non-contact sensors, such as microphones. The basic idea is that respiratory activity, which contains information regarding sleep stages and sleep breathing disorders, generates breathing (and snoring) sounds that can be acquired by a sensitive non-contact microphone and a digital audio recording device. Such an audio and breathing sound analysis approach may facilitate a solution for better accessibility to sleep evaluation, leading to early diagnosis and treatment.
Original languageEnglish
Title of host publicationBreath Sounds From Basic Science to Clinical Practice
PublisherSpringer International Publishing
Number of pages18
ISBN (Electronic)9783319718248
ISBN (Print)9783319718231
StatePublished - 12 Apr 2018

ASJC Scopus subject areas

  • General Medicine
  • General Nursing
  • General Engineering


Dive into the research topics of 'Sleep evaluation using audio signal processing'. Together they form a unique fingerprint.

Cite this