Konstantin Slavin

The idea of direct visualization of the human subthalamic nucleus (STN) is not new. Due to the high variability of its location in the depth of brain tissue, clinicians discovered early on that our standard atlas-based approach to targeting of thalamic and pallidal structures is not sufficient to reliably target the STN during functional stereotactic interventions. When the STN became the preferred substrate (target) for deep brain stimulation (DBS), mainly in the treatment of advanced Parkinson disease (PD), the ability to pinpoint the STN became particularly important because precision and accuracy of surgical intervention translates into improved outcome and reduced rate of complications.

Introduction of high-Tesla MR imaging allowed radiologists to combine high resolution with high contrast ratios, clearly defining the borders of deep midbrain nuclei—the red nucleus, the nucleus of Luys (STN), and the substantia nigra—and opened new opportunities for visualization of surgical targets. It just so happened that in my institution acquisition of a 3T MR scanner coincided with initiation of the DBS program (our first STN DBS procedure for PD took place in April 2001, a few months after the first 3T MR scanner was installed in the University of Illinois at Chicago). This coincidence prompted me and my radiology colleagues, led by Prof. Thulborn, to use this approach for my frame-based DBS cases; the only concern we had was whether distortion of 3T images would negatively affect surgical accuracy (this concern was alleviated by testing with a homemade phantom that consisted of a stereotactic frame mounted on a plastic head model and glued grapes, as well as intraoperative confirmation of target accuracy using microelectrode recording).

We used this approach routinely, assuming that every center that had a 3T MR scanner came up with the same conclusions on the usefulness and simplicity of such frame-based imaging that significantly shortened the imaging time and eliminated the need for image fusion. To my surprise, our presentations that included a description of this approach were almost uniformly met with resistance and disbelief from my colleagues, most of whom expressed concerns about the feasibility of our technique and the correctness of images that were based on placing a large metal frame into a high magnetic field. Along with my co-authors, I took this as a challenge, and collected my cases and representative images into a short paper in hopes that a publication in a highly respected peer-reviewed journal would add credibility to our method of STN visualization.

The following 8 years (the paper was published online in 2005) witnessed rapid proliferation of high-Tesla imaging devices and continuous refinement of STN visualization techniques. But it appears that our paper was on the forefront of this innovation. The practice that we pioneered has become widely accepted; it became rather a standard approach in many clinical institutions worldwide. The paper was quoted in many publications and presentations and eventually became the basis of more in-depth investigations, both clinical and scientific, including our own analysis of STN volume change in relation to progression in PD that was published few years later.

Read this article at AJNR.org . . .