The use of electromyography (EMG) is limited, particularly in the investiga
tion of children, by the invasive nature of needle electrodes. Surface elec
trode techniques are an attractive alternative but the detected signals are
greatly influenced by volume conductor effects, thus making their interpre
tation problematic. Using finite element analysis we investigated the relat
ionship between surface potential distribution and motor unit depth, incorp
orating anisotropic conductivity to model muscle tissue and a range of subc
utaneous fat thicknesses. The modeling results were used to analyze data re
corded with a 16-channel surface electrode array, from 10 normals subjects
and 12 patients with motor neuron disease, Differences in the motor units b
etween the two groups were statistically significant (P < 0.01) and are con
sistent with reinnervation and increased motor unit territory in the patien
t group. This noninvasive technique shows promise as a more acceptable alte
rnative to the use of conventional needle electrodes for neurophysiological
investigations. (C) 2001 John Wiley & Sons, Inc.