Proposition of the Artificial Intelligence GACPC© Algorithm for DBS Nuclei Targeting Optimization and Postoperative Programming

Objective: The objective of this study was to develop a mathematical solution to standardize nucleus targeting and DBS programming.

Background: Despite many sophisticated computational tools and neuroimaging information available for neurosurgeons, accurate targeting remains mostly an intuitive human task with inter-surgeon and intra-surgeon variabilities. Deep brain stimulation (DBS) programming also, remains an essay-error paradigm. Due to those uncertainty, there is still the need to perform clinical testing and neurophysiological monitoring during DBS surgery.

Methods: The volumes of a computationally reconstructed subthalamic nucleus (STN) and of a simulated volume of tissue activated (VTA) were measured according to the LPBA40 average brain template and Anne Morel’s Atlas of basal ganglia, with MNPS software. The intersection volume was calculated, and several different parameters of stimulation were simulated. Three indexes of comparison between volumes were calculated at each VTA: Paddick conformity index, Jaccard and Sorensen similarity indexes. All indexes exhibit a slope of variation as VTA increases; the maximum value corresponds mathematically to the best match possible between STN and VTA.

Results: The best possible DBS stimulation parameter can eventually be determined. At LPBA40, at 50% the distance of STN center of mass and its posterior limit: (x,y,z) −11.161, −1.465, −3.639; for the whole STN; with 1.000 Ω impedance, 90 μs, 130 Hz, was at 2.0 V, 2.0 mA. For targeting, given a certain VTA, the best possible coordinate to stimulate can be also computationally determined, since it is at the site of maximum value of the indexes.

Conclusions: This study may be one way of pushing DBS targeting and programming to a new paradigm. The GACPC© algorithm developed here will allow neurosurgeons to determine the best target coordinate for each patient and facilitate programming sessions for the best clinical results.