Semi-automated approaches to optimize deep brain stimulation parameters in Parkinson’s disease

Abstract

Background Deep brain stimulation (DBS) is a treatment option for Parkinson’s disease patients when medication does not sufficiently manage their symptoms. DBS can be a highly effect therapy, but only after a time-consuming trial-and-error stimulation parameter adjustment process that is susceptible to clinician bias. This trial-and-error process will be further prolonged with the introduction of segmented electrodes that are now commercially available. New approaches to optimizing a patient’s stimulation parameters, that can also handle the increasing complexity of new electrode and stimulator designs, is needed. Methods To improve DBS parameter programming, we explored two semi-automated optimization approaches: a Bayesian optimization (BayesOpt) algorithm to efficiently determine a patient’s optimal stimulation parameter for minimizing rigidity, and a probit Gaussian process (pGP) to assess patient’s preference. Quantified rigidity measurements were obtained using a robotic manipulandum in two participants over two visits. Rigidity was measured, in 5Hz increments, between 10–185Hz (total 30–36 frequencies) on the first visit and at eight BayesOpt algorithm-selected frequencies on the second visit. The participant was also asked their preference between the current and previous stimulation frequency. First, we compared the optimal frequency between visits with the participant’s preferred frequency. Next, we evaluated the efficiency of the BayesOpt algorithm, comparing it to random and equal interval selection of frequency. Results The BayesOpt algorithm estimated the optimal frequency to be the highest tolerable frequency, matching the optimal frequency found during the first visit. However, the participants’ pGP models indicate a preference at frequencies between 70–110 Hz. Here the stimulation frequency is lowest that achieves nearly maximal suppression of rigidity. BayesOpt was efficient, estimating the rigidity response curve to stimulation that was almost indistinguishable when compared to the longer brute force method. Conclusions These results provide preliminary evidence of the feasibility to use BayesOpt for determining the optimal frequency, while pGP patient’s preferences include more difficult to measure outcomes. Both novel approaches can shorten DBS programming and can be expanded to include multiple symptoms and parameters.

Multidimensional Bayesian Estimation for Deep Brain Stimulation Using the SafeOpt Algorithm

“…To test this, we require an efficient, safe optimization algorithm. Here, in order to develop such a tool, we extend the BayesOpt algorithm whose successful application to DBS settings we previously published 11 , using, as a test bed, a simulated cost function constructed for biological plausibility, with measurement noise based on our own experimental data with Parkinson’s gait.…”

“…it concentrates test points close to the current estimate of the optimum, where model uncertainty is low in preference to exploring high-uncertainty regions, which are more informative, and where a better optimum might lie. Our prior publication 11 used Matlab’s “expected improvement plus” acquisition function which avoids this by modifying its behavior when it is overexploiting an area. It defines, for a putative test point x a measure of overexploitation (σ x 2 - s 2 ) - ( r 2 • s 2 ) where σ x is the SD of the GP estimate at x , s is the SD of measurement noise, and r is the “exploration ratio” hyperparameter.…”

“…In our earlier publication 11 we relied on a Matlab library routine for the BayesOpt implementation. For this publication, we reimplemented a version of BayesOpt in Julia.…”

“…We have previously published a proof of concept study in which we used the BayesOpt Bayesian estimation algorithm successfully for one-dimensional optimization of DBS treatment of rigidity in people with Parkinson’s disease 11 . This proof of concept can be extended to more dimensions.…”

A New Application of Functional Zonal Image Reconstruction in Programming for Parkinson's Disease Treated Using Subthalamic Nucleus–Deep Brain Stimulation

“…Other sessions are often organized during follow-up visits to manage stimulation-induced side effects (e.g., speech problems and stimulation-induced dyskinesias) or worsening of the underlying parkinsonism. While the utility of these reprogramming sessions is well-established, no guidelines are available, and most of these changes rely on the results of a few open-label studies (10)(11)(12). In fact, although DBS has been used for almost three decades, systematic programming protocols remain lacking, leading to inconsistent and inefficient stimulation adjustments, as well as numerous or unnecessary patient visits.…”