Sonification of Polysomnographic Sleep Recordings
Ribeiro, Laura (2017)
Ribeiro, Laura
2017
Tieto- ja sähkötekniikan tiedekunta - Faculty of Computing and Electrical Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2017-08-16
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201708241722
https://urn.fi/URN:NBN:fi:tty-201708241722
Tiivistelmä
The capacity of sleep clinics to evaluate all those that suffer from sleep disorders is very limited.
The development of new home available methods of sleep analysis started to allow an affordable, extensive recording of sleep data. However, this data is hard to understand by users without extensive training and experience.
In this work we aimed to determine if sonification (which has long been used as a way of representing hard to understand data) can be used to solve this challenge.
We developed a new sonification methodology of sleep recordings by extracting relevant features from the EEG, EOG, EMG and oxygen saturation signals and their combination into functions that modulate characteristics of preexistent sounds.
Results show that separability among classes for the sonification of sleep stages is quite high, and the respiratory and movement functions have generally higher values for unhealthy patients. Listening tests show that 8 out of 10 listeners were able to correctly identify all recordings as healthy or unhealthy, and all the participants would be willing to listen to these sounds on a regular basis.
We concluded that sonification can be a very valuable tool in solving the data interpretation step of sleep recordings.
The development of new home available methods of sleep analysis started to allow an affordable, extensive recording of sleep data. However, this data is hard to understand by users without extensive training and experience.
In this work we aimed to determine if sonification (which has long been used as a way of representing hard to understand data) can be used to solve this challenge.
We developed a new sonification methodology of sleep recordings by extracting relevant features from the EEG, EOG, EMG and oxygen saturation signals and their combination into functions that modulate characteristics of preexistent sounds.
Results show that separability among classes for the sonification of sleep stages is quite high, and the respiratory and movement functions have generally higher values for unhealthy patients. Listening tests show that 8 out of 10 listeners were able to correctly identify all recordings as healthy or unhealthy, and all the participants would be willing to listen to these sounds on a regular basis.
We concluded that sonification can be a very valuable tool in solving the data interpretation step of sleep recordings.