VARIABILITY IN LATERAL ASYMMETRY MEASUREMENTS WITH ISIS2

Abstract

ISIS2 is a surface topography system measuring the three-dimensional shape of the back in scoliosis patients using digital photography with structured light. Lateral asymmetry is the ISIS clinical parameter estimating the curve of the spine in the coronal plane [1]. The shape of the back changes with patient stance, breathing and muscle tension. Although ISIS2 uses bony landmark markers to minimise the effect of stance, there will still be variations from measurement to measurement. The aim of this work is to quantify the variability in lateral asymmetry measurements. The patients were asked to stand in the patient stand in a relaxed normal pose; the feet were placed just outside the blocks on the footplate, the abdomen rested lightly against the crossbar of the stand, and the arms were supported away from the sides of the body by the arm rests. Two photographs were taken with the patient walking around the room between them. The mean difference between pairs of measurements on 62 patients was 0.12°, the standard deviation was 1.64° and the 95% limits of agreement were −3.10° to 3.34°. A plot of difference against mean showed no significant evidence of a relationship between them (r = −0.10). The standard deviation for intraob-server measurement of Cobb angle has been reported as ranging from 1.4° to 3.3° [2,3,4] and clinically significant change is generally regarded as greater than 5°. This experiment shows that intraobserver variability in lateral asymmetry is thus sufficiently low to detect clinically significant changes in the curve of the spine.