Abstract

Each year, as a result of strokes, over 1,000,000 people are left with hemiplegia, or impaired motor ability on one side of their body. A major goal in the rehabilitation of these people is to restore their ability to walk. Their quality of life would be dramatically improved by the development of rehabilitation strategies that were more effective in this regard. One of the next great challenges in biomedical engineering will be to apply sound engineering principles to this problem.

The purpose of this proposal is to investigate lower extremity dysfunction in post-stroke hemiplegia using a combination of experimental and computer modeling techniques. The biomechanical objective in most lower limb tasks is to coordinate the muscles of the legs in order to produce a desired force at the foot. We hypothesize that this ability is impaired in hemiplegic persons because of limitations in the number of muscles that their nervous system can independently control. We call these limitations abnormal coordination constraints. Examples include inappropriate synergies (muscle co-excitation), inappropriate interlimb coupling (excitation of muscles in one limb when muscles in the other limb are active), and hyperactive stretch reflexes. Although these constraints are well known in the clinic, the way in which their biomechanical implications affect a person's ability to accomplish the wide variety of tasks required for daily living is not yet fully known.

We will test the hypothesis using an ergometer pedaling paradigm. Ten age-matched control and thirty-four hemiplegic subjects will perform a variety of tasks on an ergometer. Three-dimensional kinematic data and EMG data from sixteen muscles will be recorded. The tasks will represent a wide range of biomechanical objectives with varied levels of motor program complexity, and will include isometric force generation (pushing or pulling against a stationary crank), discrete movements (moving the crank through a half cycle), and cyclical movements (pedaling). The discrete and cyclical movements will be done both concentrically (the leg propelling the crank) and eccentrically (the leg resisting a motor-driven crank).

The key element of this proposal is the use of computer models to analyze the experimental results. Because the biomechanics of human multijoint tasks are so complex, computer models are necessary to accurately interpret the contribution that each muscle makes to an overall task. We will develop a three-dimensional computer model of the musculoskeletal system and the pedaling task, and perform simulations to identify the experimentally observed abnormal coordination constraints and assess their biomechanical implications.

The expected outcome of this work will be a better understanding of the way in which abnormal coordination constraints limit the hemiplegic person's ability to accomplish the biomechanical requirements of different tasks. This will be the first step toward our long-term objective of improving lower extremity function in persons with hemiplegia.