Noninvasive measurement of spinal cord conduction: review of presently available methods.

The ability to measure spinal cord conduction velocity noninvasively is limited by available methodology. Surface recording of small spinal potentials, although feasible in infants and young children, is problematical in adults, especially when recording over the cervical spine. On the other hand, indirect methods designed to improve the signal-to-noise ratio, which include recording of somatosensory cortical evoked potentials, F-waves, or other muscle responses, are limited by the unproven assumptions necessary in calculating spinal conduction. Additionally, each method has its own particular limitations. The majority of presently available noninvasive methods take a restricted, or no, account of conduction through the motor pathways. Despite these often serious limitations, each of the reviewed methods does play a useful clinical role in the electrophysiologic investigation of cord disease that is not visible radiologically. Knowledge of them allows for sufficient diversity to tackle most relevant problems until an ideal physiologic method is developed.