Validation and Comparison of Three Positioning Protocols of Inertial Measurement Units for Measuring Trunk Movement

摘要

Postures and movements of the trunk are of ergonomic concern when evaluating the risks at work. Technical measurement methods can be used for measurements of trunk movements for long duration with high accuracy, and are therefore increasingly used in practice and research. However, currently there is no standardized protocol for the sensor placement for trunk measurement. Three placement protocols of inertial measurement units (IMUs), including placement on C7, T4 and sternum (St), in combination with S1 spinous process, were compared with an optical motion capture (OMC) system. Four subjects performed a movement test including forward to backward bending, sideward bending and twisting of the trunk, and a symmetrical lifting task. Root-mean-square differences (RMSDs) and Pearson's correlation were calculated between the two systems. For the movement tests, the RMSDs of the forward inclination at the 10~(th), 50~(th) and 90~(th) percentiles from the three IMUs were all smaller than 7.3°. Larger differences were shown for C7 of the sideward inclination at 90~(th) percentile (10.8°). Also for the twisting, larger differences were shown, especially for C7-S1 and T4-S1 (RMSD= 16.5° and 19.8°). For the lifting tests of forward inclination, St had the smallest differences compared to OMC (RMSDs < 4.1°), while slightly larger errors were found for C7 and T4 at the 90~(th) percentile (RMSDs= 8.1° and 8.2°). Different positioning protocols seem to have a slightly different effect on the measurement accuracy of trunk movement. Considerations should be taken when comparing results across studies applying different protocols.