Article

Using robotic technology, we examined the ability of a visually guided
reaching task to assess the sensorimotor function of patients with stroke.
METHODS: Ninety-one healthy participants and 52 with subacute stroke of mild to
moderate severity (26 with left- and 26 with right-affected body sides) performed
an unassisted reaching task using the KINARM robot. Each participant was assessed
using 12 movement parameters that were grouped into 5 attributes of sensorimotor
control. RESULTS: A number of movement parameters individually identified a large
number of stroke participants as being different from 95% of the controls-most
notably initial direction error, which identified 81% of left-affected patients.
We also found interlimb differences in performance between the arms of those with
stroke compared with controls. For example, whereas only 31% of left-affected
participants showed differences in reaction time with their affected arm, 54%
showed abnormal interlimb differences in reaction time. Good interrater
reliability (r > 0.7) was observed for 9 of the 12 movement parameters. Finally,
many stroke patients deemed impaired on the reaching task had been scored 6 or
less on the arm portion of the Chedoke-McMaster Stroke Assessment Scale, but some
who scored a normal 7 were also deemed impaired in reaching. CONCLUSIONS: Robotic
technology using a visually guided reaching task can provide reliable information
with greater sensitivity about a patient's sensorimotor impairments following
stroke than a standard clinical assessment scale.

Langue : ANGLAIS