Benefits of functional calibration for estimating elbow joint angles using magneto-inertial sensors: preliminary results

Magneto-inertial measurement unit (M-IMU) sensors are more and more used in motion analysis as they allow for ‘out-of-the-lab’ measurements. This technology combines acceleration, angular velocity and magnetic field information for the estimation of the sensor three-dimensional orientation. Since it does not deliver position information, bony landmarks palpation as proposed by the International Society of Biomechanics (ISB) (Wu et al. 2005) cannot be associated with this technology. A calibration is thus needed to establish the relation between each sensor orientation and its corresponding human segment (De Vries et al. 2010). Although the literature proposes several ways of calibration, it is not straightforward for the researcher and for the practitioner to know the pros and cons of each method for each joint studied. Moreover, most of the studies propose a common calibration for several joints (i.e. a unique standing pose) and do not take into account joint specificity. In addition, an orthonormal representation of segments is often associated to these calibrations, involving possible divergence between perpendicular axes and anatomical axes. Finally, validation of M-IMU calibrations is generally done using a gold-standard optoelectronic technique whereas other validity criteria such as limited cross-talk effect represent meaningful complementary information (Cutti et al. 2008). This work proposes to study different ways of generating segment axes, joint by joint. These segment axes are not enforced to be orthonormal (Dumas and Cheze 2007) and are built using either static postures or functional movements. This abstract presents preliminary results for the elbow joint angles.