Intelligent Headband System for Evaluating Rehabilitation Effectiveness

Abstract

Stroke is an acute cerebrovascular condition causing damage to cranial nerves and requires subsequent rehabilitation treatment. In clinical practice, the effectiveness of rehabilitation is usually subjectively assessed by experienced physicians or using global prognostic scales. Several brain imaging techniques, such as positron emission tomography, functional magnetic resonance imaging, and computed tomography angiography, can be applied in rehabilitation effectiveness evaluation, but their complexity and long measurement times limit the activity of patients during measurement. This paper proposes an intelligent headband system based on near-infrared spectroscopy. An optical headband continuously and noninvasively monitors changes in hemoglobin parameters in the brain. The system’s wearable headband and wireless transmission provide convenience of use. According to the change of hemoglobin parameters during rehabilitation exercise, several indexes were also defined to evaluate the state of cardiopulmonary function and further build the neural network model of the cardiopulmonary function evaluation. Finally, the relationship between the defined indexes and the cardiopulmonary function state were investigated and the neural network model for the cardiopulmonary function evaluation was also applied in the rehabilitation effect evaluation. The experimental results show the cardiopulmonary function state could reflect on most of the defined indexes and the output of neural network model, and the rehabilitation therapy could also improve the cardiopulmonary function.

Keywords

• Cardiopulmonary function
• near-infrared spectroscopy
• neural network model
• rehabilitation effectiveness
• stroke