In e.m.g.s recorded from patients with pathological tremours, the tremour cannot often be recognised by visual inspection. Demodulation of tremour e.m.g.s improves the detectability of the tremour, but the signal-to-noise ratio remains poor. In this study, demodulated artificially generated tremour e.m.g.s and e.m.g.s from a patient with a pathological tremour were subjected to digital spectral analysis. Power spectra, phase and coherence functions of the modulating and demodulated signals were calculated. The result is that the detectability is improved from a modulation depth of 60% in modulated noise to 2-4% after demodulation and spectral analysis. The results suggest a useful application of the demodulator and analysing software to tremour e.m.g.s in clinical practice. © 1983 IFMBE.
|Number of pages||4|
|Journal||Medical & Biological Engineering & Computing|
|Publication status||Published - Sept-1983|
- Spectral analysis