Somatosensory evoked potentials are used along with motor evoked potentials for ensuring patient safety during scoliosis surgery.
The operating room is a difficult recording environment with electromagnetic noise overlapping the few μV signal that is used
to asses the status of the somatosensory tract of the patient during the operation. In this paper we study three methods to
improve the signal-to-noise ratio of the recorded samples. The widely used arithmetic mean is compared with median and trimmed
mean methods.
Arithmetic mean smoothed the curves and median and trimmed mean preserved higher frequencies. The calculated signal-to-noise
ratios using the three methods yielded similar values with some higher figures during suppression EEG epochs using median
and trimmed mean. As a conclusion all methods gave results of our data, but none was significantly better than the others.
Keywords Somatosensory evoked potentials - Signal-to-noise ratio - Ensemble averaging - Anaesthesia