To get started is not an easy task, if you are working on fundamental BCI, then your first thought is to scrutinize what is your IV and DV that you wanna examine. If you are working on applied BCI, definitely you want to first target which signals you primarily want to work on (SSVEP, P300, MI, or hybrid) and what kind of end results you would like to have. If you are working on affective BCI, definitely you want to determine what emotion or cognition you want to work to classify/predict.
Most of these answers can be found in papers, and getting the right papers will speed up your research process
Best papers to read as a starter is always the review paper. Review paper will give you a general idea what the field is doing, and what are the current problems facing the field (which leads you to find your potential topic!).
There are many excellent review papers in BCI including,
- Review of BCI in general, Another one
- Review of deep learning in BCI, Another one, An example
- Review of Riemannian geometry in BCI
- Sparse Bayesian Method in BCI
- Review of SVM in BCI
- Summary of Transfer Learning in BCI
- Review of classification algorithms
- Review of EEG BCI Design Paradigms
- Review of BCI spellers
- Review of Hybrid spellers, Another one, An example
- Review of visual P300 spellers
- Review of RSVP BCI, RSVP target
- Review of Emotion recognition in BCI
- Summary of common pitfalls
- Review of Artifact removal in BCI
- Comparison study of signal decomposition in BCI
A good reference book will be Brain Computer Interfacing: An Introduction - Rajesh
After getting an overview, you can start drilling more specific publications. When you are reading, try to focus on learning research methodologies which include IV, DV, metrics, preprocessing techniques, topography, EEG validity, models, etc. as this will often help you determine your topic.