Towards a Measure of Muscle Function in Spinal Muscular Atrophy (SMA) and Duchenne Muscular Dystrophy (DMD): Using Wearable Sensors

Topic:

Other

Poster Number: T336

Author(s):

Arafat Rahman, University of Virginia, Shashwat Kumar, University of Virginia, Allison McCrady, University of Virginia, Chelsea Masterson, BA, University of Virginia, Ann Sumner Thorp, University of Virginia, Laura Barnes, PhD, University of Virginia, Silvia S. Blemker, PhD, University of Virginia, Rebecca J. Scharf, MD, University of Virginia

Neuromuscular disorders may lead to weakness of muscles due to muscle degradation in Duchenne Muscular Dystrophy (DMD), or motor neuron loss in Spinal Muscular Atrophy (SMA). Traditional techniques for assessing these diseases like the Brooke score or the Children’s Hospital of Philadelphia Infant Test of Neuromuscular Disorders (CHOP-INTEND) are subjective, semiquantitative, and motivation-dependent. They are also limited in quantifying subtle changes in patient movements and may not be sensitive enough to assess incremental changes. Movement quality assessment in children may be challenging using wearable sensors because behavior and mood may affect results, and patients may move their arms slower and faster due to their inability to follow precise instructions. To overcome these challenges, we employed shape analysis to align arm movement trajectories collected from 25 patients with DMD, 11 with SMA, and 16 healthy controls over time while they were performing different arm movements wearing gyroscope sensors on their wrists. We then created a standard shape of reference trajectory for assessing the quality of arm movement and used functional principal component analysis (FPCA) to summarize different trials of arm movements (bicep curl and knocking). From FPCA, we identified two main variations in arm movements: the speed of the movement and the symmetry of the motion, especially when lifting against gravity. Patients with SMA, in particular, showed more pronounced asymmetry in their arm movements while those with DMD’s movement speed decreased with age, indicating temporal degeneration of muscles. By combining these movement variations using Canonical Correlation Analysis (CCA), we derived a formula that correlates the speed and asymmetry of arm motions with echogenicity (muscle fat infiltration), Brooke score, and age-related changes (r = 0.61). This formula proposes a new, promising index for assessing motor function, paving the way for more accessible and home-based evaluations for people with neuromuscular disorders.