210P Advancing upper limb motor function evaluation in Duchenne muscular dystrophy and spinal muscular atrophy via kinematic parameterization with the wearable device “ArmTracker”

Abstract

Assessing upper limb motor function in Duchenne Muscular Dystrophy (DMD) and Spinal Muscular Atrophy (SMA) traditionally relies on the Performance of Upper Limb (PUL) and Revised Upper Limb Module (RULM) scales, respectively. While considered gold standards, using these scales in isolation presents some challenges, notably in capturing subtle changes in motor function over time or in response to treatments. Inertial Measurement Units (IMUs) are inertial sensors that provide objective and quantifiable movement data. We hypothesize that integrating IMU measurements into these scales could provide complementary data for a more comprehensive assessment of motor function in individuals with neuromuscular diseases. Ten children with DMD (aged 12-17, Brooke score 2-5), 10 children with SMA (aged 6-13, Brooke score 2-5), and 6 healthy control children (aged 5-17, Brooke score 1) performed the PUL and RULM scales while wearing the ArmTracker device, equipped with 7 IMUs (Xsens Dot, Xsens Technologies) placed on the back of the hands, forearms, arms, and torso. Each IMU provided quaternion data. A sensor-to-segment calibration process was conducted with subjects seated in a chair with forearms resting on a table. Photographs were taken in frontal and lateral planes during calibration, and this visual information was integrated with IMU data to enhance calibration accuracy. Euler angles YZ'Y’’ for the shoulder, and ZX'Y’’ for the elbow and wrist were utilized. Maximum reachable area of hands, workspace area, and range of motion of shoulder, elbow, and wrist were evaluated and correlated with motor function scale scores using the Spearman correlation coefficient. Workspace area was normalized by the maximum achievable area by individuals (in %). We found significant correlations between the workspace area of upper limbs, notably on the frontal plane, and the scores of both the PUL and RULM. Additionally, significant correlations were observed between the scores of both motor function scales and the range of motion (ROM) of the shoulder. Furthermore, a correlation matrix analyzing the angles of the three primary upper limb joints revealed compensation patterns, which proved particularly valuable in identifying compensatory movements during the shoulder abduction task within the scales. Employing inertial sensors during the administration of functional motor scales in individuals with neuromuscular diseases yields valuable variables for assessing motor function, with particular interest in workspace area of the upper limbs and ROM of the shoulder. These and other variables are currently under further investigation within the same cohort of individuals who have undergone evaluations at home and school spanning four days utilizing the ArmTracker. This ongoing research aims to ascertain the system's potential for conducting assessments at home and assessing real-life movements in everyday scenarios.