Recalibration of neuromodulation parameters in neural implants with adaptive Bayesian optimization

Abstract

Objective. Neuromodulation technology hold promise to treat several conditions where physiological mechanisms of neural activity have been affected. Personalization of neurostimulation protocols is a mandatory step to make the treatment efficient and the devices highly effective. The interfaces between the targeted nervous tissue and the neurotechnology (i.e., human-machine link or neural interface) usually requires a constant re-calibration of the neuromodulation parameters, due to many different biological and microtechnological phenomena happening over-time. This adaptation of the optimal stimulation parameters generally involves an expert-mediated re-calibration, with correspondent economic burden, compromised every-day usability and efficacy of the device, and consequent loss of time and increased discomfort of patients going back to clinics to get the device tuned. We aim to construct an adaptable AI-based system, able to compensate for these changes autonomously. Approach. We exploited Gaussian Process-based Bayesian Optimization (GPBO) methods to re-adjust the neurostimulation parameters in realistic neuroprosthetic data by integrating the temporal information into the process to tackle the issue of time variability. To this aim, we built a predictive model able to tune the neuromodulation parameters for two separate crucial scenarios where re-calibration is needed. In the first one, we built a model able to find the optimal active sites to use from a multichannel electrode, i.e. able to cover a certain function for a neuroprosthesis, which in this specific case was the evoked-sensation location variability. In the second one, we propose an algorithm able to adapt the injected charge required to obtain a functional neural activation (e.g., perceptual threshold variability). Main results. Our automatic algorithm can successfully adapt neurostimulation parameters to perceptual threshold changes and to evoked-sensation location changes over-time. These findings show an automatic and fast way of tackling the inevitable variability of neurostimulation parameters over time. It will potentially enable a widespread and a better usability of this technology upon validation in other frameworks, while decreasing the time and the costs of the treatment. This work suggests the exploitation of AI-based methods for developing the next generation of “smart” neuromodulation devices.

Online Bayesian optimization of vagus nerve stimulation

“…The neural stimulation optimization using a Bayesian method, BO as suggested in this study, is more flexible and efficient, particularly for high-dimensional parameter spaces. Several simulation studies on neuromodulation models, that enable a precise comparison with some ground truth, show its promising performance when optimizing parameters (Grado et al 2018, Mao et al 2023, Aiello et al 2023. Evaluations using simulated data have consistently showcased BO's superior performance compared to random and grid searches-a point we highlight in this study as well-or greedy strategies (Bonizzato et al 2023).…”

“…BO algorithms have recently emerged as promising optimization tools for stimulations of the nervous system in several studies: for optimizing epidural spinal cord stimulation in rats (Desautels et al 2015), for intracortical stimulation in monkeys (Laferriere et al 2020), for distal limb movement in monkeys (Losanno et al 2021, Bonizzato et al 2023, treatment of epilepsy (Park et al 2020, Stieve et al 2023), and improved memory (Ashmaig et al 2018) in deep brain stimulation. Simulation studies, utilizing models based on neuromodulation data, demonstrate the potential of BO in comparison to other optimization methods (Grado et al 2018, Mao et al 2023, Aiello et al 2023.…”

Personalized strategies of neurostimulation: from static biomarkers to dynamic closed-loop assessment of neural function

“…In the same vein, Reihart and Nguyen demonstrated fast working-memory performance improvement in elders by applying HD-tACS with frequencies tuned in an individualized fashion ( Reinhart and Nguyen, 2019 ). Nowadays, the importance of personalization is further emphasized in the clinical usage of personalized neurostimulation for epilepsy ( Beumer et al, 2021 ), sensorimotor disorders ( Gupta et al, 2023 ), cognitive function ( Albizu et al, 2023 ), and dysfunction ( Hunold et al, 2022 ; Reinhart, 2022 ; Aiello et al, 2023 ), and several other applications. In fact, researchers are actively exploring tailored approaches that leverage computational models, machine learning, and Bayesian optimization algorithms to optimize stimulation parameters and enhance treatment outcomes.…”

Gaussian-process-based Bayesian optimization for neurostimulation interventions in rats

“…As demonstrated in the primary publication 1 and elsewhere, 2 , 3 , 4 , 5 , 6 Gaussian-process (GP)-based Bayesian optimization (BO) algorithms are a powerful framework to automatically optimize the efficacy of neurostimulation. It has been shown to outperform other strategies to simultaneously find the optimal values of multiple stimulation parameters (i.e., the optimal combination of parameter values) to maximize a chosen feature of the evoked response (e.g., the movement amplitude or the electromyographic [EMG] burst amplitude).…”