A Brain Computer Interface is defined as a system that measures and analyze brain signals and converts them in real-time into outputs that do not depend on the normal output pathways of peripheral nerves and muscles. Nowadays, a BCI system is used in two different ways: 1) as tool for enabling communication with external environment or controlling some external devices; 2) as rehabilitation tool for stimulating brain plasticity in all those patients who shown motor deficit after stroke. In the latter case, BCI interfaces are controlled through the amplitude of a well-known cerebral rhythm, the sensory motor rhythm, which is evident over the sensorimotor areas in 8-13 Hz frequency band. Since previous findings in the field described plasticity phenomena at the basis of the motor recovery as modifies in brain networks, it could be interesting to develop BCI systems using as controlling features, measures derived by such brain networks. The estimation of brain networks from scalp EEG signals represents a help in moving a step forward in this field. However, it is difficult to estimate brain connectivity in an on-line setting (BCI use) due to the scarce amount of data samples available for the estimate. What is still missing is: i) a methodological solution in estimating brain connectivity ideally in on-line environment; ii) a toolbox for brain connectivity estimation to be integrated in a BCI system. The main objective of the proposed project consists in the development and implementation of an accurate and reliable toolbox for brain connectivity estimation to be used in a BCI system. Such toolbox will be the result of an integration between advanced methodologies for brain connectivity analysis and Machine Learning. The developed toolbox will be integrated in a BCI system that will be tested in a group of 10 healthy subjects performing the motor imagery task. The project will provide a new BCI system able to train brain plasticity after a damage in motor functions.
This project will provide a toolbox for brain connectivity analysis to be used in all those contexts in which few data samples are available for brain connectivity estimation process. Furthermore, this toolbox will be tested on a sample of subjects demonstrating the ability of new system to detect a particular brain circuit at the basis of motor imagery task. In future, the toolbox could be used on a greater experimental sample and on real patient with a motor disability and could have a strong impact on several clinical application.
In conclusion, the innovation of the project cover several aspects:
-methodological: 1) the project will produce a new toolbox for brain connectivity evaluation able to track brain circuits even if few data samples are available. This could open the way for the use of multimodal BCIs in which there are many different signals considered at the same time. 2) Transition from features derived from automatic process (evoked potentials), to features directly extracted from estimated brain connectivity patterns; this last information could be integrated in a classical BCI system in order to provide an additional information to improve performances of the mentioned system, in term of accuracy.
- clinical: there are several clinical conditions in which this new BCI system could be used:1) in disabilities associated with brain stroke in which the brain structures face a reorganization and the damaged motor functions could be restored via neuroplasticity. In this context, the new system could be used as analysis tool in order to increase the comprehension of the neural plasticity. 2) In epilepsy detection. Most available commercial epilepsy-related systems primarily focus on monitoring or off-line analysis of brain activities. This new BCI system with an integrate brain connectivity information could help neurologists during the diagnosis and during the detection of epileptic outbreak. Furthermore, on-line brain connectivity estimation could provide a description of brain circuits at the basis of an epileptic seizure for improving the knowledge about this clinical condition.
- neurophysiological: because many of cognitive processes are not stationary, could be very useful having a tool that can take into account inter-trial variability. The proposed approach could be employed in all those contexts in which is important evaluate this variability such as social neuroscience or during cognitive task. In fact, it was proved that there is a relationship between the fluctuations of EEG brain signals and behavioral performance during a cognitive task. These fluctuations are often considered as noise and treated as a variance of no explanatory value. For this reason, this new toolbox could be used for evaluate inter-trial variability in term of brain connectivity instead of considering behavioral performance. Furthermore, this could lead to a greater comprehension about temporal connection during the execution of cognitive task.