Quantification and validity of modified Romberg tests using three-axis accelerometers

Abstract

The Romberg test is a simple and inexpensive test of the vestibular spinal response. It is difficult to quantify, has low sensitivity and specificity. We used four three-axis accelerometers to supplement the Romberg test by modified the test as a test of the regulation of upright position in order to determine its clinical validity. The Romberg measures taken at different locations, such as the head and COG (Center of gravity), were correlated with one another, we computed Pearson's coefficient of correlation for each pairwise combination of locations then performed independent two-sample t-tests on the mean values of the repeated Romberg tests at each location. Data from the head sensors were significantly different between the control and the vestibule-neuritis groups as the data from the COG sensors and data from both legs were also significantly different between the groups. © 2012 Springer-Verlag.