Non-Invasive Spatial Registration Using Customized Dental Bracket and Improved Genetic Algorithms

In order to implant the flexible electrodes into the brain safely, the titanium screws are driven into the skull to register the patient’s physical space with imaging spaces obtained from CT. The operation may cause bleeding risk, and the distribution of screws depends on the doctor’s subjective experience. This paper proposes a non-invasive registration approach based on a customized dental bracket, a series of registration balls are arranged on the bracket. Additionally, an improved genetic algorithm (IGA) is introduced to select the optimal registration balls automatically according to the implantation target position, which can promote the registration accuracy effectively. Experimental results show that the proposed method can achieve a registration error within 0.36mm without the screws implantation, which can meet the requirements for flexible electrode implantation. Additionally, compared to manual selection screws, the target registration errors are reduced by 61.3%, and the computing time of IGA is within 0.2 seconds.