AR spectral analysis technique for human PPG, ECG and EEG signals.
Journal of medical systems
confidence
Key findings
Comparative spectral analysis (FFT and AR) of PPG, ECG, EEG signals under PEMF; no clinical/biological endpoints reported.
View source on PubMed (PMID 18444356) ↗
- Sample size
- Not reported
- Population
- Human subjects (PPG, ECG, EEG signal recordings)
- Dosing
- PEMF at extremely low frequency (ELF)
- Duration
- Not reported
- Route
- PEMF exposure
- Blinding
- not_reported
- Controls
- not_reported
- Drug class
- physical modality
Full abstract
In this study, Fast Fourier transform (FFT) and autoregressive (AR) methods were selected for processing the photoplethysmogram (PPG), electrocardiogram (ECG), electroencephalogram (EEG) signals recorded in order to examine the effects of pulsed electromagnetic field (PEMF) at extremely low frequency (ELF) upon the human electrophysiological signal behavior. The parameters in the autoregressive (AR) method were found by using the least squares method. The power spectra of the PPG, ECG, and EEG signals were obtained by using these spectral analysis techniques. These power spectra were then used to compare the applied methods in terms of their frequency resolution and the effects in extraction of the features representing the PPG, ECG, and EEG signals. Some conclusions were drawn concerning the efficiency of the FFT and least squares AR methods as feature extraction methods used for representing the signals under study.