Loading...

Automatic Detection of Sleep Arousal from EEG Signal Using Respiratory Information

Aghdaei, Elnaz | 2021

394 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54767 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Shamsollahi, Mohammad Bagher
  7. Abstract:
  8. Sleep is vital for physical and mental health, affecting neurocognitive, physiological, and psychopathology functions and performance. Arousals are linked with sleep and interrupt the sleep states, forming a sleep/arousal loop. Spontaneous arousals are part of a normal sleep/wake cycle. There are also different clinical conditions causing sleep fragmentation and arousals, including sleep apnea (obstructive, central, and mixed apnea), hypopnea, and non-apnea such as respiratory effort-related arousals (RERA), snoring, teeth grinding, and periodic leg movement.This research introduced a novel approach for automatic arousal detection inspired by extracting respiratory information from EEG signals. The dataset was provided by the PhysioNet/ Computing in Cardiology Challenge 2018. It comprised overnight polysomnographic (PSG) recordings of 1985 subjects at the Massachusetts General Hospital (MGH) labs.We introduced a novel approach based on Time-frequency mapping and a deep learning method to extract respiratory information from EEG signals, such that the information of the time domain with the airflow signal is well comparable. For this comparison, we used the phase synchronization criteria to examine the similarity between the dummy airflow signal (in this dissertation, we have called it the characteristic signal) and the reference airflow signal. After that, we classified arousals and non-arousals using the characteristic signal.Finally, to compare with the results of other articles, characteristic signal accompanied by other physiological signals have been used to classify arousals. According to the results, the characteristic signal obtained from the time-frequency mapping method can be a suitable alternative to the reference airflow signal
  9. Keywords:
  10. Electroencphalogram Signal ; Apnea Detection ; Phase Synchronization ; Airflow Signals ; Polysomnography (PSG) ; Sleep Arousal

 Digital Object List

 Bookmark

  • تشخیص arousal‌های خواب از روی سیگنال EEG با استفاده از اطلاعات تنفسی استخراج شده از آن
  • تشخیص arousal‌های خواب از روی سیگنال EEG با استفاده از اطلاعات تنفسی استخراج شده از آن
  • تشخیص arousal‌های خواب از روی سیگنال EEG با استفاده از اطلاعات تنفسی استخراج شده از آن