STUDENT CORNER
Koji Koizumi
This was the first time I attended the International BCI meeting. This event took place on May 21 – 25, 2018 at the Asilomar Conference Center in Pacific Grove, California, USA. To tell the truth, this was the first conference I have attended. Just about the time I finished writing my graduation thesis at the University of Tokyo, I got to know about this meeting through the call for papers via its website. Because I was just beginning in BCI, I wanted to deepen my understanding of the field and benefit from the experience and advice of other researchers. Therefore, I immediately decided to submit an abstract. Fortunately, I received the Student Award at this meeting and received a travel grant from the IEEE Brain Initiative. I would like to take this opportunity to express my appreciation to IEEE Brain Initiative. In conclusion, this conference was truly amazing and I was really pleased having participated in this conference.
A characteristic of this meeting is that BCI researchers from all over the world gather at the Asilomar conference center, distant from the urban area, and soaked themselves with BCI every day. The program of the BCI meeting is quite dense in activities: Plenary talks and Keynote sessions, Poster sessions, a session on Neuroethics, Workshops, as well as Fundamental Didactic Sessions designed specifically for students to learn about the state-of-the-art and to answer and discuss about BCI practice and research. Every session was very attractive and there were many new topics for me, so I was able to learn a lot.
In a Keynote Session, a research director at Facebook came to talk about their vision of BCI devices. His talk was very interesting and made me excited. However, I see a gap between the ideals of the company and the actual research, and I felt that there are many ethical and technical issues to overcome. It was a good opportunity to make me think about these issues. His talk generated an active discussion among BCI researchers. The most interesting thing for me was the Master Class that was introduced for the first time in this conference. Each masterclass had a ‘master’, i.e. a senior researcher, and two students. Anyone else was free to attend and contributed but only the two students got to present their work. Each student presented his/her research (10-15 minutes), then the master provided wise and constructive comments and discussion was encouraged with the audience. The masterclass was informal and was intended to give students access to senior expertise and get good tips for their research. In my Master class, Professor Jonathan R Wolpaw (National Center for Adaptive Neurotechnologies, Wadsworth Center, USA) was the chairperson. I am deeply touched that such an expert advised me by listening to my presentation given that I had just started my research. After my presentation was over, the chairperson and the audience gave important insights. In particular, they taught me the importance of the definition of the words and the reliability of the results. Even the same words can be interpreted in different ways depending on the background and field of the researcher. Therefore, when introducing new words, you must start by defining them. Otherwise, communication is inconsistent, and it affects understanding. On the reliability of results, I used EEG to obtain brain signals in my research. In my study, when the power spectral density (PSD) of the high frequency band was extracted and used as the feature, the brain signal classification accuracy when visually imagining how an object moves in three axes was significantly higher than the PSD of the lower frequency band. Of course, we used the existing method to remove the artifacts, but it is known that the myoelectric potentials can contaminate the high frequency bands. Even if the myoelectric potential could be eliminated by 80-90%, it cannot be said that it was completely removed, and the possibility that the electromyographic artifacts affect the result cannot be ruled out. I thought that I had to study more data analysis and noise removal method.
In addition to the scientific program, researchers took breakfast, lunch and dinner at the same venue, so the discussion on research was blooming at every table during meals. For me, as someone just starting in BCI research, all the time I spent awake during the meeting I was learning new things. Also, as I was staying at in a quadruple room and was able to develop a friendship with my roommates. After the program of the day, dozens of researchers went to the beach and gathered around a bonfire, creating links among us while we grilled marshmallows and shared drinks; it was a wonderful experience. For us researchers who are engaged in research in their respective countries on a daily basis, it is extremely important for us to create connections, strengthen relationships and exchange valuable information beyond the boundaries of these borders. All that was possible at this conference.
BCI research is progressing with the cooperation of researchers in truly diverse fields such as brain science, psychology, and engineering both in academia and industry. I got acquainted with many researchers from various field, learned much, through the opportunity of participating in this conference. It became very meaningful for me. I strongly wanted to participate in this conference the next time. I am thankful to the organizers of BCI meeting from the bottom of my heart.
Biography
Koji Koizumi, is in the 1st year of a Master’s student at the University of Tokyo – Japan, majoring in mechanical engineering. He is studies Brain-Computer Interface, Cognitive Psychology and Machine Learning. In his BSc graduation thesis, he studied whether it is possible to identify the cognitive state by measuring the EEG when visually imaging the multiple movements of a drone. He gave an oral presentation on research at the International Conference of the IEEE Engineering in Medicine & Biology Society (EMBS) held in Hawaii in July 2018. His strong research interest is the human-machine cooperative control of spacecraft and aircraft using BCI. Contact Information: koizumi@hnl.t.u-tokyo.ac.jp