Low-Cost Classroom and Laboratory Exercises for Investigating Both Wave and Event-Related Electroencephalogram Potentials

Kylie Smith, Abbey Pilger, Marcio L. M. Amorim, Stanislav Marcic, Zach Reining, Nick Ristow, Dylan Miller, Aljoscha Leonhardt, Joseph C. Donovan, Matthias Meier, Timothy C. Marzullo, Etienne Serbe-Kamp, Adam. P. Steiner, Gregory J. Gage

https://doi.org/10.59390/YNPH4485

Electroencephalography (EEG) has given rise to a myriad of new discoveries over the last 90 years. EEG is a non-invasive technique that has revealed insights into the spatial and temporal processing of brain activity over many neuroscience disciplines, including sensory, motor, sleep, and memory formation. However, most undergraduate students lack laboratory access to EEG recording equipment or the skills to perform an experiment independently. Here, we provide easy-to-follow instructions to measure both wave and event-related EEG potentials using a portable, low-cost amplifier (Backyard Brains, Ann Arbor, MI) that connects to smartphones and PCs, independent of their operating system. Using open-source software (SpikeRecorder) and analysis tools (Python, Google Colaboratory), we demonstrate tractable and robust laboratory exercises for students to gain insights into the scientific method and discover multidisciplinary neuroscience research. We developed 2 laboratory exercises and ran them on participants within our research lab (N = 17, development group). In our first protocol, we analyzed power differences in the alpha band (8-13 Hz) when participants alternated between eyes open and eyes closed states (n = 137 transitions). We could robustly see an increase of over 50% in 59 (43%) of our sessions, suggesting this would make a reliable introductory experiment. Next, we describe an exercise that uses a SpikerBox to evoke an event-related potential (ERP) during an auditory oddball task. This experiment measures the average EEG potential elicited during an auditory presentation of either a highly predictable (“standard”) or low-probability (“oddball”) tone. Across all sessions in the development group (n=81), we found that 64% (n=52) showed a significant peak in the standard response window for P300 with an average peak latency of 442ms. Finally, we tested the auditory oddball task in a university classroom setting. In 66% of the sessions (n=30), a clear P300 was shown, and these signals were significantly above chance when compared to a Monte Carlo simulation. These laboratory exercises cover the two methods of analysis (frequency power and ERP), which are routinely used in neurology diagnostics, brain-machine interfaces, and neurofeedback therapy. Arming students with these methods and analysis techniques will enable them to investigate this laboratory exercise’s variants or test their own hypotheses.