Design and evaluation of pediatric gait rehabilitation robots

Thesis Abstract

Gait therapy methodologies were studied and analyzed for their potential for pediatric patients. Using data from heel, metatarsal, and toe trajectories, a nominal gait trajectory was determined using Fourier transforms for each foot point. These average trajectories were used as a basis of evaluating each gait therapy mechanism.

An existing gait therapy device (called ICARE) previously designed by researchers, including engineers at the University of Nebraska-Lincoln, was redesigned to accommodate pediatric patients. Unlike many existing designs, the pediatric ICARE did not over- or under-constrain the patient’s leg, allowing for repeated, comfortable, easily-adjusted gait motions. This design was assessed under clinical testing and deemed to be acceptable.

A gait rehabilitation device was designed to interface with both pediatric and adult patients and more closely replicate the gait-like metatarsal trajectory compared to an elliptical machine. To accomplish this task, the nominal gait path was adjusted to accommodate for rotation about the toe, which generated a new trajectory that was tangent to itself at the midpoint of the stride. Using knowledge of the bio-mechanics of the foot, the gait path was analyzed for its applicability to the general population.

Several trajectory-replication methods were evaluated, and the crank-slider mechanism was chosen for its superior performance and ability to mimic the gait path adequately. Adjustments were made to the gait path to further optimize its realization through the crank-slider mechanism.

Two prototypes were constructed according to the slider-crank mechanism to replicate the gait path identified. The first prototype, while more accurately tracing the gait path, showed difficulty in power transmission and excessive cam forces. This prototype was ultimately rejected. The second prototype was significantly more robust. However, it lacked several key aspects of the original design that were important to matching the design goals. Ultimately, the second prototype was recommended for further work in gait-replication research.