I. Introduction
In recent years, the number of patients with hand dysfunction has increased, and most of them lack the motor function and somatic sensation of the hand. The hand is one of the most important parts of how we interact with the world. Reaching, opening, and grasping are the basic hand behaviors that we used in our daily life. Those behaviors are achieved by our specific muscles that produce target torque to corresponding joints. The somatic sensation is another important function of our hands, and it makes it possible for us to interact with the world. The neural pathways of patients with hand dysfunction are impaired, and there is an urgent need for a technique to restore the neural pathway. Braincomputer interface (BCI) as a novel human-machine interaction interface makes it possible to communicate with the brain directly by decoding the neural signals [1]. BCI fundamentally relies on the fact that information within the nervous system is conveyed by the patterns of electrical activity in neurons. Because BCI can not only recording the neural signals but also stimulate the nervous, a BCI system can be a bidirectional neural pathway and bypassed the damaged neural system. We describe BCI as a neural prosthesis and the framework of the bidirectional BCI is illustrated in Fig. 1. In this review, we will focus on BCI used for restoring the hand function.