AUTHOR=Rodriguez Fernando , He Shenghong , Tan Huiling 

TITLE=The potential of convolutional neural networks for identifying neural states based on electrophysiological signals: experiments on synthetic and real patient data

JOURNAL=Frontiers in Human Neuroscience

VOLUME=Volume 17 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2023.1134599

DOI=10.3389/fnhum.2023.1134599

ISSN=1662-5161

ABSTRACT=Processing incoming neural oscillatory signals in real-time and decoding from them relevant behavioural or pathological states is often required for adaptive Deep Brain Stimulation (aDBS) and other Brain-Computer Interface (BCI) applications. Most current approaches rely on first extracting a set of pre-defined features, such as the power in canonical frequency bands or various time-domain features, and then training machine learning systems that use those predefined features as inputs and infer what the underlying brain state is at each given time point. However, whether this algorithmic approach is best suited to extract all available information contained within the neural waveforms remains an open question. Here, we aim to explore different algorithmic approaches in terms of their potential to yield improvements in decoding performance based on neural activity such as measured through local field potentials (LFPs) recordings or electroencephalography (EEG). In particular, we aim to explore the potential of end-to-end convolutional neural networks, and compare this approach with other machine learning methods that are based on extracting pre-defined feature sets. To this end, we implement and train a number of machine learning models, based either on manually constructed features or, in the case of deep learning-based models, on features directly learnt from the data. We benchmark these models on the task of identifying neural states first based on simulated data with various features that have been previous identified as playing a role in physiological and pathological functions. We then compare the performance of these models in decoding movements based on local field potentials recorded from the motor thalamus of patients with essential tremor. The results, based on both the simulated and real patient data, show that end-to-end deep learning-based methods may outperform feature-based methods, especially when the patterns of interest within the data are either unknown or not readily quantifiable, or when there are potential unknown features which might contribute to decoding performance. The methods proposed here could be leveraged in the context of aDBS, and can also be used in other brain computer interface applications.