Date of Award

Fall 2008

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)


Biomedical Engineering

First Advisor

Schmit, Brian D.

Second Advisor

Hornby, George T.

Third Advisor

Bansal, Naveen K.


We conducted 2 separate experimental studies to I. Evaluate the effect of short duration, high intensity, high velocity body weight supported treadmill training on the locomotor function in ambulatory subjects with motor incomplete SCI and 2. Identjfy the differences in the variability of muscle activation patterns between robotic-assisted and therapist-assisted treadmill walking in subjects with motor incomplete spinal cord injury. In the first study, we assessed the effect of a short bout of high intensity fast walking treadmill training in subjects with incomplete SCI. Treadmill training with body weight support was provided at 5Kmph for 10 minutes with a 5 minute break period midway. EMG and kinematic data were collected during training. Prior to and after the training, over ground walking measurements were made using a Gaitmat. The Spasticity Index and integrated area under the EMG were calculated as a basic measure of muscle activity. A model was constructed to capture the muscle activation patterns. Using the model we were able to precisely compare the activity of muscles in one or more phases of gait. We observed significant increases in over ground walking speed stride length and cadence post training. A corresponding increase in integrated EMG area was seen in the Vastus Lateral is muscle. Specifically we saw significant increase in VL EMG activity during pre-swing / early swing compared to the more conventional early stance phase of the gait cycle. The change in timing of VL activation during pre/swing - swing phase suggests that the quadriceps muscle may be assisting the calf muscles in propelling the subject forward. Previous studies suggest that the hip proprioceptive feedback and spinal reflexes triggered by hip extension may cause this exaggerated vasti activity. In the second study, we identified the difference in variability in muscle activation patterns between therapist-assisted and robot-assisted (Lokomat) treadmill training. Twelve incomplete SCT subjects participated in the study. The robotic and therapist assisted walking we:re randomized between subjects, each of which involved walking at 3Kmph for 10 minutes with body weight support. EMG and kinematic data were collected during training. After filtering, rectification and smoothening of the EMG data, we constructed a 95% confidence interval and over laid plots for visual comparison. To quantify the difference in variability in muscle activation we calculated the Muscle Activation Timing Index (MAPI). The 95% confidence bands and the MATI scores, revealed significantly higher variability in muscle activation in therapist-assisted walking conditions. Error-inducing training has previously been shown to be more potent in adaptive learning in neurological injured population over error-reducing methods. Variability in muscle activation is seen here as a mechanism to adapt to error in stepping. The Lokomat, which reduces kinematic error involved in gait by following a rigid walking pattern, also appears to reduce the variability in muscle activation.



Restricted Access Item

Having trouble?