Machine learning: a new direction for assisting the application of deep brain stimulation

Abstract

In recent years, deep brain stimulation has been widely used in the treatment of various difficult diseases of the nervous system. The rapid development of machine learning algorithms has greatly promoted the application and development of deep brain stimulation. Machine learning can help in preoperative screening of surgical candidates, outcome prediction and surgical planning. It can assist target localization during operation. More importantly, the local field potential signals recorded by macroscopic electrodes can be analyzed in real time after surgery, which provides a basis for the development of a closed-loop stimulation system. Of course, the application of machine learning also has its limitations and challenges, such as dimensionality reduction of high-dimensional data, development and validation of new models and etc, which need to be further explored and improved.