An automated navigation system for deep brain stimulator placement using hidden Markov models.

Abstract

Placement of deep brain stimulators (DBSs) currently involves the use of both image-based stereotaxy and intraoperative microelectrode recording (MER). Interpretations of MER data and integration with anatomical data are currently manual processes. Hidden Markov models (HMMs) are commonly used in signal processing, speech recognition, and a wide array of biologic applications. A 6-state HMM was designed and trained for evaluation in simulated surgery for subthalamic nucleus (STN) DBS. The accuracy of identifying the correct brain location was 98.5%. Sensitivity of detecting passes intersecting the STN was 100%, and specificity was 84.9%. Anatomical location of the MER passes was calculated with a mean error of 0.06 mm (95% confidence interval, -0.54 to 0.42 mm) in the medial-lateral axis. Automated DBS intraoperative navigation using HMMs may be feasible based on promising results of this prototype system.