Statistical Analysis of Sleep Quality from EEG Signal with Aromatic Stimuli

Abstract

Generally, adults need to sleep for about 8 hours every night to keep the body in a good condition. In addition to adequate sleep, good sleep quality is also an important factor to note. But when the sleep time and sleep quality was disrupted, of course many bad effects are lurking. The quality of one's sleep can be known by reading the bio potential signals in the brain commonly called electroencephalograph (EEG). To be able to detect this signal, an EEG sensor is required at some point around the head. The data of signal measurement by EEG sensor will be recorded and can be processed directly by computer, without having to go through the signal conditioning circuit first. This process is usually called the data acquisition process. In the first stage of digital processing on a computer, the signal will have pre-processing (cutting, filtering and segmentation). Then, the signal will be extracted by the method of Discrete Wavelet Transform and Fast Fourier Transform. The output of this system is the signal characteristics of the frequency and amplitude of the EEG signal of each sleep data segment. The amplitude results will be used as input in analyzing using the t test method paired samples. It is obtained that lavender aromatherapy with alpha and theta waves frequency of 0.023 and 0.021, respectively provide the smaller of significance level about 0.05.