Abstract:Predicting and understanding the changes in cognitive performance, especially after a longitudinal intervention, is a fundamental goal in neuroscience. Longitudinal brain stimulation-based interventions like transcranial direct current stimulation (tDCS) induce short-term changes in the resting membrane potential and influence cognitive processes. However, very little research has been conducted on predicting these changes in cognitive performance post-intervention. In this research, we intend to address this gap in the literature by employing different EEG-based functional connectivity analyses and machine learning algorithms to predict changes in cognitive performance in a complex multitasking task. Forty subjects were divided into experimental and active-control conditions. On Day 1, all subjects executed a multitasking task with simultaneous 32-channel EEG being acquired. From Day 2 to Day 7, subjects in the experimental condition undertook 15 minutes of 2mA anodal tDCS stimulation during task training. Subjects in the active-control condition undertook 15 minutes of sham stimulation during task training. On Day 10, all subjects again executed the multitasking task with EEG acquisition. Source-level functional connectivity metrics, namely phase lag index and directed transfer function, were extracted from the EEG data on Day 1 and Day 10. Various machine learning models were employed to predict changes in cognitive performance. Results revealed that the multi-layer perceptron and directed transfer function recorded a cross-validation training RMSE of 5.11% and a test RMSE of 4.97%. We discuss the implications of our results in developing real-time cognitive state assessors for accurately predicting cognitive performance in dynamic and complex tasks post-tDCS intervention

What problem does this paper attempt to address?

The aim of this paper is to address the problem of predicting changes in multitasking performance after long-term transcranial direct current stimulation (tDCS) through EEG (electroencephalogram) functional connectivity analysis and machine learning methods. Specifically, the goal of the study is to predict changes in cognitive performance after a 6-day tDCS intervention in a complex multitasking environment by combining functional connectivity metrics extracted from EEG data (such as Phase Lag Index and Directed Transfer Function) with various machine learning algorithms. This fills a gap in the existing literature on this type of predictive research and contributes to the development of real-time cognitive state assessment tools to accurately predict cognitive performance after tDCS intervention in dynamic complex tasks. The study found that the Multilayer Perceptron (MLP) and Directed Transfer Function (DTF) had RMSEs of 5.11% in cross-validation training and 4.97% on the test set, indicating that these models can effectively predict changes in cognitive performance.

Prediction of multitasking performance post-longitudinal tDCS via EEG-based functional connectivity and machine learning methods