Seeking Optimal Montage for Single-Pair Transcranial Direct Current Stimulation Using Bayesian Optimization and...

Area of Science:

• Neuroscience
• Computational Neuroscience
• Biomedical Engineering

Background:

• Transcranial direct current stimulation (tDCS) is an emerging neuromodulation technique.
• Optimizing tDCS montages is critical due to the significant impact of electrode position and current intensity on stimulation effects.
• Conventional optimization methods are limited by high computational load, restricting the search space of electrode positions.

Purpose of the Study:

• To introduce and evaluate the Bayesian optimization and hyperband (BOHB) method for optimizing tDCS montages across the entire human scalp.
• To overcome the constraints of traditional optimization approaches that use limited electrode position candidates.
• To seek optimal tDCS montages without conventional limitations.

Main Methods:

• The study employed the Bayesian optimization and hyperband (BOHB) method, a probabilistic approach for iterative montage selection.
• BOHB was compared against random search and greedy search for tDCS simulation suitability.
• The objective function was defined as maximizing the average electric field norm (|E|) in the motor area (M1) and left dorsal lateral prefrontal cortex.

Main Results:

• The BOHB method demonstrated superior performance compared to random search in both regions of interest (ROIs) for the same number of iterations.
• Statistically significant improvements were observed, with smaller iteration indices for BOHB in M1 (p < 0.0001).
• BOHB yielded a higher normalized electric field (|E|) in the M1 region compared to random search.

Conclusions:

• The BOHB method is a feasible and effective approach for optimizing transcranial direct current stimulation (tDCS) montages.
• It outperforms traditional random search methods in finding optimal stimulation parameters.
• BOHB can serve as an efficient tool for seeking optimal tDCS montages, adjustable via control parameters.