Evaluation of New Parameters for Assessment of Stroke Impairment
Authors: Kathrin Tyryshkin, Janice Glasgow and Stephen Scott.
Abstract:
This paper presents new parameters for the evaluation of stroke impairment using data collected with KINARM robot (Kinesiological Instrument for Normal and Altered Reaching Movements). The new parameters evaluated in this study were cross-correlation, low frequency, and high frequency. The data were collected from control (people with no neurological disorders) and stroke subjects performing a center outreach task. In this task the subjects were instructed to move the examined arm quickly and accurately from the central target position to a randomly illuminated target, and to maintain the hand at this target for the remainder of the trial. The collected data for each of eight individual reaching movements to eight different targets can be viewed as a time series. For each subject a cross-correlation between the reaching movement to each of eight targets and a straight line fitted between these targets was computed. In addition, high and low frequencies were calculated from the time series data using a Fourier transform. The results showed that the new parameters identified the same or a higher percentage of stroke participants as abnormal, compared to previously reported parameters [1], especially in the experiments performed with the non-affected arm. Therefore, the new parameters can facilitate the detection of abnormalities in the movements of stroke patients and may be used as features for the classification of stroke patients.