Real-time EEG forecasting achieves faster and more accurate predictions
REVBRAIN: Real-time Probabilistic EEG Forecasting Using Residual Vector Quantization
Machine Learning
Summary
Forecasting brain signals in real time is useful for devices that respond to brain activity as it happens. The authors improved on an existing deep learning model called WaveNet, which was accurate but too slow for real-time use. Their new method, REVBRAIN, uses a clever way to represent the brain signals that lets it predict faster and more accurately. This speedup makes it possible to use these predictions live in brain-computer interfaces and other brain sensing applications.
What this means in practice
- •For bci developers: Implement real-time brain state detection systems that respond instantly using faster and accurate probabilistic EEG forecasts.
- •For wearable device engineers: Build brain-monitoring wearables that can predict neural activity quickly for improved brain-state-dependent stimulation treatments.
Authors
Hanna Pankka, Jaakko Lehtinen, Timo Roine
Abstract
Real-time forecasting of resting-state EEG signals is required by applications like brain-state-dependent stimulation, where computer generated actions are coupled with ongoing neural oscillations. Recently, the deep learning model WaveNet was proposed for the forecasting task. It shows superior accuracy compared to the prevailing method, the linear autoregressive (AR) model, and provides a built-in uncertainty measure due to its probabilistic modeling framework. However, inference using WaveNet is too slow for real-time use due to its sample-by-sample autoregressive operation: a single forecast requires evaluating the model tens or hundreds of times depending on the length of the prediction window. In this paper, we present REVBRAIN, a probabilistic deep learning method that produces forecasts in real time by generating the forecast over the whole prediction horizon at once. This is achieved by tokenizing the output space using residual vector quantization (RVQ) and generating forecasts over the items in the tokenized space. We tested REVBRAIN for forecasting 200 ms of EEG time series in the alpha-frequency band (8-13 Hz) and compared our results against WaveNet model. We found that REVBRAIN achieved real-time inference on a laptop GPU and an over $50\times$ speedup compared to WaveNet, while reaching superior prediction accuracy. By achieving real-time inference speed, REVBRAIN enables incorporating accurate and probabilistic EEG forecasting into BCI applications, opening up new ways for investigating the functioning of the human brain.