Towards A Comprehensive Software Suite for Stereotactic Neurosurgery

Stereotactic neurosurgery is a subspeciality within neurosurgery that relies on accurately targeting structures within the brain for diagnostic and therapeutic purposes. Surgical implantation of electrodes is a common procedure in stereotactic neurosurgery. For instance, deep brain stimulation (DBS) is an effective treatment option for individuals living with movement disorders while stereoelectroencephalography (SEEG) provides invaluable information from individuals living with drug-resistant epilepsy. In both procedures, electrophysiology data is acquired and used to identify brain characteristics such as surgical target location in DBS surgery and the seizure onset zone in SEEG surgery. Accurate surgical targeting of the electrodes is crucial, with millimeter deviations resulting in unwanted side effects and minimal clinical benefit. Among the medical centres that perform stereotactic neurosurgery, most employ preoperative magnetic resonance imaging (MRI) visualization of the surgical target for preoperative surgical planning. Neuronavigation is the term given to a set of tools or software that a neurosurgeon can use to navigate the brain. Currently, closed-source commercial neuronavigation software is used for preoperative trajectory planning. To ensure optimal positioning of DBS electrodes, intraoperative electrophysiology recording and test stimulation are performed to accurately place the electrode(s) within the surgical target nucleus. The preoperative MRI provides the anatomical border of the selected target nucleus while the intraoperative recordings provide the electrophysiological border. Once the final electrodes are implanted, the commercial neuronavigation software can be used to qualitatively assess the postoperative position of the implanted electrode(s) based on postoperative medical imaging data. Unfortunately, the commercial neuronavigation software is costly, is not easily customized, employs proprietary data formats, and often does not implement the most cutting-edge algorithms that may improve targeting accuracy. Few open-source tools have been developed that perform similar functions to the commercial neuronavigation software while also maintaining the flexibility for users to modify and improve the tool. The work in this thesis explores open-source solutions to stereotactic neurosurgery planning, data storage, and data visualization. In Chapter 3, a anatomical fiducial placement protocol is validated in a set of clinical imaging data for potential use in surgical planning. Chapter 4 explores an open-source pipeline for detecting and navigating stereotactic space using several common head frame systems. In Chapter 5, an open-source data storage structure for electrophysiology data is described and a preprocessing pipeline is introduced, called ephysPrep, which extracts electrophysiology signal features for machine learning applications. Finally, in Chapter 6, an open-source neuronavigation software called trajectoryGuide is introduced, which provides modules for preoperative surgical planning, data visualization, and postoperative electrode localization with electrode stimulation field modelling. Overall, the goal of this work is to provide a foundation of open and transparent tools that can be used and built upon by members of the clinical neuroscience community