To measure the posture of a paraplegic while standing up, an economical met
hod has been developed. Markers, glued over the joints, hook onto strings w
hich run to rotary potentiometers mounted on a fixed frame. Springs maintai
n a near-constant small tension in the strings, and the potentiometers rota
te as the strings wind onto pulleys. The positions are calculated by triang
ulation, using two potentiometers per joint, assuming that body segment len
gths are constant, or three potentiometers without this assumption. Using a
second-order polynomial fit, the random error in length measurement for ea
ch potentiometer is less than +/-2 mm for the string length from 0 mm to 11
00 mm, or less than +/-1 mm in actual range from 600 mm to 1000 mm. With tw
o potentiometers per joint, using a second-order polynomial fit and assumin
g the ankle position is known exactly, an estimate of the resulting errors
in the knee and hip marker positions are 4 and 8.5 mm, respectively. (C) 19
99 TPEM, Published by Elsevier Science Ltd. All rights reserved.