Pilot Study of Sleep Apnea Detection with Wavelet Transform

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60461373%3A22340%2F17%3A43914777" target="_blank" >RIV/60461373:22340/17:43914777 - isvavai.cz</a>

Výsledek na webu

<a href="http://www2.humusoft.cz/www/papers/tcp2017/032_schatz.pdf" target="_blank" >http://www2.humusoft.cz/www/papers/tcp2017/032_schatz.pdf</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Pilot Study of Sleep Apnea Detection with Wavelet Transform

Popis výsledku v původním jazyce

The sleep apnea syndrom is defined as repeated pauses in breathing during sleep period which leads to interrupts in sleep and decreases in oxyhemoglobin saturation. It is well understood that quantity and quality of sleep could significantly affect work productivity. In this study multimodal analysis of breathing is done with two different sensors. The first sensor measures nasal air flow and the second sensor measure abdomen effort during breathing. As it is needed to manually go through records of whole night sleep to confirm some of an automatic classification of events that can disturb sleep, it is very important to have accurate classifier. This papers aim is to present results of pilot study of competitive neural network (CNN) classifier based on Wavelet transform, with which is possible to evaluate sleep apnea from multimodal breathing data with accuracy of 94.2 % with comparison to classification of Sleep apnea by doctor. Evaluation of the whole output of CNN is complicated as the neural network was trained without target data. It can detect all apnea events from length of 5 seconds, including those that are missing in the classification by a doctor.

Název v anglickém jazyce

Pilot Study of Sleep Apnea Detection with Wavelet Transform

Popis výsledku anglicky

The sleep apnea syndrom is defined as repeated pauses in breathing during sleep period which leads to interrupts in sleep and decreases in oxyhemoglobin saturation. It is well understood that quantity and quality of sleep could significantly affect work productivity. In this study multimodal analysis of breathing is done with two different sensors. The first sensor measures nasal air flow and the second sensor measure abdomen effort during breathing. As it is needed to manually go through records of whole night sleep to confirm some of an automatic classification of events that can disturb sleep, it is very important to have accurate classifier. This papers aim is to present results of pilot study of competitive neural network (CNN) classifier based on Wavelet transform, with which is possible to evaluate sleep apnea from multimodal breathing data with accuracy of 94.2 % with comparison to classification of Sleep apnea by doctor. Evaluation of the whole output of CNN is complicated as the neural network was trained without target data. It can detect all apnea events from length of 5 seconds, including those that are missing in the classification by a doctor.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2017

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

24th Annual Conference Proceedings Technical Computing Prague 2017

ISBN

978-80-7592-002-7

ISSN

2336-1662

e-ISSN

neuvedeno

Počet stran výsledku

11

Strana od-do

"32-1"-"32-11"

Název nakladatele

Vysoká škola chemicko-technologická v Praze

Místo vydání

Praha

Místo konání akce

Praha

Datum konání akce

8. 11. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—