Fig 1.
Experimental setup and training protocol.
(A) The 3 degrees-of-freedom robot (WristBot) and the placement of the vibro-tactile actuators in the ipsilateral and contralateral configuration. The 4 vibro-tactile actuators were placed on the forearm to indicate right, left, up and down directions in order to correct movement trajectory. (B-C) Overall training protocol. All groups underwent proprioceptive and motor assessment which consisted of joint position sense testing and the wrist tracking task. Training executed by PT and PTVTF groups consisted of a center-out reaching task with haptic and vibro-tactile feedback.
Table 1.
Settings for providing vibro-tactile feedback.
Fig 2.
Motor learning related changes in the center-out pointing task during training.
Each data point represents the group mean for a block of 24 trials (each training day consisted of 7 blocks; group mean = mean of individual subject means). Error bar represents the Standard Error. (A) Maximum Lateral Deviation (MLD). (B) Movement Time (MT). (C) Number of movement units (MU). PT = proprioceptive training; PTVTF = proprioceptive training + vibro-tactile feedback.
Fig 3.
Effect of proprioceptive training on proprioceptive acuity.
The boxplot shows the pre- and post-training distribution of the position matching error (ME) for each experimental group. Medians are indicated by the solid line inside the box. The lower and upper edges of the box represent the 25th and 75th percentiles, respectively. The whiskers indicate the 1st and 99th percentile. Asterisks (*) indicate statistically significant differences for between-group (POST: PT vs PTVTF; NOPT vs PTVTF) and within-group (PTVTF: PRE vs POST) comparisons.
Table 2.
Tracking error and magnitude of haptic force in the untrained motor task.
Fig 4.
Effect of proprioceptive training on the untrained wrist tracking task.
Pre to posttest change in the use of haptic force feedback (HF) as a function of tracking error (TE). Each data point represents the coordinates of the respective group means. The error bars represent Standard Error. The percentage of change was calculated as (initial value—final value) / initial value. Positive values correspond to a decrease of the variable with training. Note that the PTVTF group had significantly decreased the amount of haptic feedback after training.
Fig 5.
Effect of vibrotactile feedback location.
Matching error (ME), haptic force feedback (HF) and tracking error (TE) values for PTVTFright (light gray) and PTVTFleft (dark gray) groups. Each point represents group mean, pre and post training and the bars represent the SE. The comparison shows no difference between the two groups for each variable.