Identification of Responders to the FES Interventions in Stroke Population

Hemiplegia with associated foot drop occurs in 50% of the stroke survivors and frequently impairs an individual's ability to walk. Functional Electrical Stimulation (FES) based neuroprosthetic devices have been developed to correct foot drop. The efficacy of these devices were initially examined by Liberson el al. who demonstrated that electrical stimulations could assist in restoring functional movements in paralyzed limbs. In addition to assistance with foot drop, these devices have showed significant improvements in biomechanical variables such as walking speed, distance, stride length and physiological cost for individuals with stroke. In order to comprehensively understand the effect of electrical stimulations on gait recovery, it is critical to analyze the dynamic aspects of gait and measure gait variability during the functional electrical stimulation intervention. In the proposed investigation, we will determine the 'gait symmetry' of FES assisted walking using bilateral cyclograms of the ankle and knee over a period of 6 months. This novel approach will account for the dynamics and complexity of balance by measuring the deviations of joints from a line of symmetry at every instance of gait cycle and will provide better measure of gait symmetry. Utilization of this outcome measure will allow us to understand the role of electrical stimulation at ankle and how this effect gets translated to the knee and hip joints during walking. The changes in the surface electromyograms (EMGs) of selective muscle groups will demonstrate how FES can contribute to muscle re-training after stroke. We will use advanced signal processing algorithms to remove FES artifact from the EMG signal in order to comprehensively analyze the carry-over effect of the FES intervention. Finally, we will employ Principal Component Analysis (PCA) - an advanced data mining technique to track and quantify the overall gait recovery process of individuals with stroke using pattern classification algorithms. The gait symmetry measure and the EMGs will be statistically classified to see their clear separation at baseline and 6 month intervals. This classification will allow us to identify the individuals who were most responsive to the intervention. This information is critical and will allow researchers and clinicians to re-strategize the rehabilitation process. Such scientific evaluation will provide the base for further development and implementation of FES devices or technologies, thus supporting the NINDS' fundamental goal of translating basic and clinical discoveries into better ways to treat neurological disorders.