Concurrent validation of a magnetometer-based step counter in various walking surfaces.

Background And Aim: Clinicians need a simple method for quantifying gait activity. The aim of this study was to develop and validate the reliability of a quantitative gait assessment based exclusively on one magnetometer located on the shank.

Methods: Twenty-five healthy volunteers were simultaneously equipped with a magnetometer (MAG system) on the right shank, and two validated step-counter systems: the StepWatch Activity Monitor (SAM) and three Force-Sensing Resistors (FSRs). Volunteers performed a standard circuit including level walking, up and down stairs and up and down a slope. The three step counting approaches were compared using the Pearson correlation coefficient and the Bland-Altman method for each of the surface-types.

Results: The step counts measured by the MAG and FSR were highly correlated for all the surfaces (r>.83). Congruently, the Bland-Altman analysis revealed an overall ± 5% limit of agreement. The step counts measured by the MAG and SAM were also well correlated for the level-surface condition (r=.85), with a Bland-Altman ± 5% limit of agreement but comparisons were less satisfying for the other surfaces.

Conclusions: These results demonstrate that the use of a single magnetometer is an accurate tool for step counting over varied surfaces. These small sensors are easy to set up and to use and the signal processing is robust, making the MAG method highly applicable for clinical purposes, especially for the analysis of long walking periods in daily life conditions.