A reductionist approach to the analysis of learning in brain–computer interfaces

• Published in: Biological cybernetics Vol. 108; no. 2; pp. 183 - 201
• Main Author: Danziger, Zachary
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.04.2014; Springer; Springer Nature B.V
• Subjects: Adaptive algorithms; Adult; Algorithmics. Computability. Computer arithmetics; Algorithms; Applied sciences; Artificial Intelligence; Bioinformatics; Biological; Biological and medical sciences; Biomechanical Phenomena; Biomedical and Life Sciences; Biomedicine; Brain-Computer Interfaces; Complex Systems; Computer Appl. in Life Sciences; Computer science; control theory; systems; Computer systems and distributed systems. User interface; Cybernetics; Dynamical systems; Dynamics; Electrodiagnosis. Electric activity recording; Exact sciences and technology; Female; Human; Human-computer interaction; Human-computer interface; Humans; Investigative techniques, diagnostic techniques (general aspects); Learning; Male; Medical sciences; Nervous system; Neurobiology; Neurosciences; Original Paper; Sex Characteristics; Software; Task Performance and Analysis; Theoretical computing; User-Computer Interface; Young Adult; Cyberglove; Motor learning; Brain–computer interface; Machine learning; Closed loop; Brain; Glove; Adaptive algorithm; Sex; Online algorithm; Brain-computer interface; User interface; Learning algorithm; Artificial intelligence
• ISSN: 0340-1200; 1432-0770; 1432-0770
• DOI: 10.1007/s00422-014-0589-3

The complexity and scale of brain–computer interface (BCI) studies limit our ability to investigate how humans learn to use BCI systems. It also limits our capacity to develop adaptive algorithms needed to assist users with their control. Adaptive algorithm development is forced offline and typically uses static data sets. But this is a poor substitute for the online, dynamic environment where algorithms are ultimately deployed and interact with an adapting user. This work evaluates a paradigm that simulates the control problem faced by human subjects when controlling a BCI, but which avoids the many complications associated with full-scale BCI studies. Biological learners can be studied in a reductionist way as they solve BCI-like control problems, and machine learning algorithms can be developed and tested in closed loop with the subjects before being translated to full BCIs. The method is to map 19 joint angles of the hand (representing neural signals) to the position of a 2D cursor which must be piloted to displayed targets (a typical BCI task). An investigation is presented on how closely the joint angle method emulates BCI systems; a novel learning algorithm is evaluated, and a performance difference between genders is discussed.