Classification of motor activities through derivative dynamic time warping applied on accelerometer data.

In the context of tele-monitoring, great interest is presently devoted to physical activity, mainly of elderly or people with disabilities. In this context, many researchers studied the recognition of activities of daily living by using accelerometers. The present work proposes a novel algorithm for activity recognition that considers the variability in movement speed, by using dynamic programming. This objective is realized by means of a matching and recognition technique that determines the distance between the signal input and a set of previously defined templates. Two different approaches are here presented, one based on Dynamic Time Warping (DTW) and the other based on the Derivative Dynamic Time Warping (DDTW). The algorithm was applied to the recognition of gait, climbing and descending stairs, using a biaxial accelerometer placed on the shin. The results on DDTW, obtained by using only one sensor channel on the shin showed an average recognition score of 95%, higher than the values obtained with DTW (around 85%). Both DTW and DDTW consistently show higher classification rate than classical Linear Time Warping (LTW).

Difference in sensorimotor adaptation to horizontal and vertical mirror distortions during ballistic arm movements.

When learning a novel motor task, the sensorimotor system must develop new strategies to efficiently control the limb(s) involved, and this adaptation appears to be developed through the construction of a behavioral map known as an 'internal model'. A common method to uncover the mechanisms of adaptation and reorganization processes is to expose the system to new environmental conditions, typically by introducing visual or mechanical distortions. The present study investigated the adaptation mechanisms of the human sensorimotor system to horizontal and vertical mirror distortions (HMD and VMD) during the execution of fast goal-directed arm movements. Mirror distortions (MDs) were created by means of virtual visual feedback on a computer screen while the movement was executed on a graphics tablet. Twenty healthy adult participants were recruited and assigned to one of two groups of 10 people each. Tests were divided in two subsequent blocks of five trials. The first block consisted of trials with no mirror distortion (NMD), while the second block was recorded when exposing one group to HMD and the other to VMD. Both MDs resulted in kinematic changes: during the tests with the MDs the participants did not reach the performance level found at the NMD test. Motor performance during HMD appeared to be globally better than during VMD and the adaptation process to VMD appeared to be slower than to HMD, but data interpretation was hampered by large within-participant and between-participant variability. In-depth analyses of the data revealed that most of the motor performance information was contained in the direction of movement. The data supported the idea that the internal model for HMD was already partially built.