Abstract
Parkinson's disease (PD) is a severe, progressive, neurological disorder. PD is not a single disease, but rather resembles a syndrome. PD includes two types of pathogenesis (i.e., classical PD and new PD). Clinically, PD patients present with a range of motor symptoms including decreased spontaneous movement, bradykinesia, muscle rigidity, changes in speech, and resting tremors. PD patients also often exhibit non-motor symptoms such as fatigue, sleep disorders, and emotional and mental health disturbances. Deep brain stimulation (DBS) performed in clinical neurosurgery has demonstrated considerable efficacy in the treatment of dyskinesia that occurs in PD patients. However, the specific neural mechanism of DBS remains unknown and is limited by several shortcomings that have hampered the popularization and development of the procedure. To address this issue, this study established a theoretical model of DBS for PD to investigate and understand the mechanism of DBS using several artificial intelligence (AI) algorithms. This model was used to investigate both classical PD and unheard-of new PD. The research described in this paper was as follows: a single neuron was used to establish a theoretical model of the basal ganglia circuit and to simulate the characteristic indicators of the potential release of the basal ganglia circuit in both normal and PD states. The state of the deep brain electrical stimulation in PD was then analyzed to identify the critical electrical stimulation index and the optimal target. We showed that the use of AI algorithms such as particle swarm optimization and other AI algorithms was beneficial for more detailed exploration and understanding of the mechanisms of DBS compared to those used in previous studies. This discovery may lead to advances in DBS technology and provide better treatment options for neurological diseases such as PD.