Closing the loop between brain and electrical stimulation: towards precision neuromodulation treatments

Abstract

One of the most critical challenges in using noninvasive brain stimulation (NIBS) techniques for the treatment of psychiatric and neurologic disorders is inter- and intra-individual variability in response to NIBS. Response variations in previous findings suggest that the one-size-fits-all approach does not seem the most appropriate option for enhancing stimulation outcomes. While there is a growing body of evidence for the feasibility and effectiveness of individualized NIBS approaches, the optimal way to achieve this is yet to be determined. Transcranial electrical stimulation (tES) is one of the NIBS techniques showing promising results in modulating treatment outcomes in several psychiatric and neurologic disorders, but it faces the same challenge for individual optimization. With new computational and methodological advances, tES can be integrated with real-time functional magnetic resonance imaging (rtfMRI) to establish closed-loop tES-fMRI for individually optimized neuromodulation. Closed-loop tES-fMRI systems aim to optimize stimulation parameters based on minimizing differences between the model of the current brain state and the desired value to maximize the expected clinical outcome. The methodological space to optimize closed-loop tES fMRI for clinical applications includes (1) stimulation vs. data acquisition timing, (2) fMRI context (task-based or resting-state), (3) inherent brain oscillations, (4) dose-response function, (5) brain target trait and state and (6) optimization algorithm. Closed-loop tES-fMRI technology has several advantages over non-individualized or open-loop systems to reshape the future of neuromodulation with objective optimization in a clinically relevant context such as drug cue reactivity for substance use disorder considering both inter and intra-individual variations. Using multi-level brain and behavior measures as input and desired outcomes to individualize stimulation parameters provides a framework for designing personalized tES protocols in precision psychiatry.

Fig. 1. Block diagram of closed-loop stimulation: an engineering perspective.

Different parts of this system are constituting the output: brain state or behavior, Desired brain state: a predefined reference. Measured brain state (measurement): measured/quantified brain state known as a biomarker (as an indirect indicator of brain state; e.g., frontoparietal synchronization). Comparator: an algorithm that compares a measured brain state with a predefined desired brain state and sends the comparison result (named Error signal) to the controller. When there is no difference between desired and measured brain states, the comparator output is zero. Therefore, the controller input is also zero, which means that there is no need to change the stimulation parameters. Controller: an optimization algorithm that receives the difference between desired and measured brain states and tries to find optimum values of stimulation parameters based on minimizing differences between the desired and measured brain states. Stimulator: a neurostimulation device such as transcranial electrical stimulator (tES) that adjusts its parameters (e.g., phase and frequency) based on information received from the controller. Brain: the plant under-stimulation. Sensor: a hardware or device that records/quantifies the current brain state (e.g., fMRI system). Control signal/command: signal/command to titrate stimulation dose automatically. Actuating signal: electrical current stimulation signals applied to the brain. Behavior: the loop can be extended to behavior, and instead of target engagement biomarkers, a treatment response biomarker is recorded (e.g., drug craving self-report). Brain model (generated by sensors): A model that links the behavioral and clinical outcome and biomarker to a disease mechanism and defines the dynamic targets for engagement and change. Stimulation model: A model that includes technical properties of the stimulator and its mechanism of action to predict the optimized protocol based on the inputs from the stimulation model.

Fig. 2. tES-fMRI and real-time fMRI-neurofeedback integration into a closed-loop system to optimize dose titration in tES studies.

a Real-time fMRI-neurofeedback system (rtfMRI-NF). In rtfMRI-NF interventions, subjects are asked to regulate their brain functions based on an instruction to engage specific neural targets. BOLD signals are analyzed with a rapid algorithm in each loop, and the level of target engagement is visualized. Subjects regulate their brain activities based on the instruction to maximally activate predefined neural targets. b Concurrent tES-fMRI system. The tES stimulator is placed outside of the MR scanner to avoid associated noise, and fMRI data are collected simultaneously with tES to measure the modulation of the neural targets. c Closed-loop tES-fMRI. In a closed-loop tES-fMRI setting, the fMRI data are acquired and analyzed in a real-time approach in response to the stimulation to measure and reach the ideal target with optimized stimulation parameters. rtfMRI real-time functional magnetic resonance imaging, NF neurofeedback, tES transcranial electrical stimulation.

Fig. 3. The process of integrating tES with fMRI in a real-time closed-loop approach (Closed-loop tES-fMRI).

(1) Concurrent tES-fMRI starts with prior expectations about optimal tES parameters. (2) Targets are selected based on the clinical/behavioral outcome of interest and its corresponding neurocognitive function. The averaged BOLD signals are extracted from predefined targets. (3) To measure ongoing brain state (e.g., frontoparietal connectivity), the BOLD signal is segmented using a tapered sliding window, and dynamic similarities between extracted BOLD signals in frontoparietal regions of interests (ROIs) are calculated for each segment using the Pearson correlation coefficient (r). (4) Fisher’s Z transformation of those correlation coefficients is used to measure dynamic functional connectivity (FC), i.e., the dynamic correlation between the time series of frontal and parietal ROIs is defined as a model of current brain state over time. (5) The extracted measures are compared with the desired value and the results of the respective comparison are fed into an optimization algorithm. (6) Optimal stimulation parameters are determined to minimize the difference between ongoing FC and the desired value by maximizing the objective function in a defined parameter search space (e.g., 1D search space to optimize the frequency of the injected current or 2D search space to optimize phase difference and frequency simultaneously). (7) The stimulation device is updated with the optimal stimulation parameters for the next round, and this loop continues until predefined stopping criteria are reached. tES transcranial electrical stimulation, fMRI functional magnetic resonance imaging, BOLD blood oxygenation level-dependent.

Fig. 4. From closed-loop individualization to real-world clinical application.

The clinical utility of optimized parameters obtained from closed-loop tES-fMRI can be tested in two main types of subsequent trials. (1) Multi-session trials to test whether the online effect will be transferred as accumulating and long-lasting offline effect meaningful in the clinical setting and (2) Trials with home-based on-demand use (e.g., when a person with substance use disorders feels a high level of craving at midnight) of the online effects (e.g., momentary reduction in craving during stimulation) will have a meaningful effect on clinical outcomes (e.g., relapse or overdose prevention).

Fig. 5. Exemplary functional neuroimaging-informed intervention development pathway.

The pathway from the intervention (top) to clinical outcomes (bottom) is illustrated for the example of aiming for frontoparietal synchronization in drug addiction with a tES intervention in a closed-loop tES-fMRI system.