At age 10, senior Bryant Liu developed a habit of lying on a mat daily for 15 to 30 minutes of mindfulness. Amid busy days filled with schoolwork, he has learned to value this time as an imperative part of his routine.
Experiencing the stress-reducing result of meditation and reading studies proving that it changes brain structure inspired Liu to pursue a project linking meditation and cognitive science at the 2022 online Scholar Launch summer research program.
Liu wrote his research paper in August before presenting his work online to two conferences in China: the 2022 International Communication Engineering and Cloud Community Conference in October and the 2022 International Conference on Computing and Big Data in November.
The Institute of Electrical and Electronics Engineers (IEEE) also accepted his paper for publication on the Xplore website in December, though it has not been published yet and will be released in future journals.
His original goal was to create a program to track people’s progress in meditation through brain changes, but he found finding and analyzing actual brain data to be too ambitious; instead, he switched to measuring changes in the Brain-Computer Interface (BCI), an external device that picks up on the brain’s electrical activity and allows the user to control devices such as a wheelchair or computer purely with brain activity.
Since BCI can be affected by meditation, Liu looked to use deep learning models to capture its precise effects. Previous research had found that people are better at controlling BCI when doing meditation, and Liu dived deeper into comparing the accuracy of meditation BCI versus normal BCI.
After taking a month to finalize his idea last June, he studied deep learning models and brainstormed methods before jumping into programming with Python and MATLAB in July.
“For models, it took longer actually learning about it than actually writing it,” he said. “Writing it didn’t take that long.”
Liu’s research is relatively unusual as it is not a continuation of previous work, but rather original work stemming from his personal interests. The only elements he borrowed from others’ research were the datasets his Scholar Launch deep learning neuroscience mentor Ganesh Mani found from a publicly available research paper published by the Oxford University Press. Liu used this data to train his model.
“[My project] lays the foundation for future deep learning models that people want to apply to BCI and annotation,” Liu said.
Receiving permission to use those datasets, however, turned out to be the most difficult part of the research process. He contacted around seven professors from a wide range of universities that had done research with BCI data but did not receive any replies.
With Mani’s guidance, Liu further solidified his interest in both neuroscience and meditation, leading him to choose neuroscience as his second-choice major behind computer science in his college applications and to start the meditation club in the Wellness Center this year with juniors Nancy Lei and Danni Deng as well as freshman Vivian Lei. A typical club meeting involves 10 to 20 minutes of guided meditation.
For people just starting with research, Liu recommends getting a mentor or connecting with an experienced researcher for guidance in developing their idea and planning how to carry it out.
“Researching by yourself is really hard because you probably don’t know what you need to do,” Liu said.
In the future, Liu hopes to see more applications of machine learning models to improve BCI accuracy and obtain more specific results on meditation’s impact on the brain.
“I’m looking forward to researching in computer science and mind-related areas in the future or something else related to applying tech to cognitive science,” Liu said. “This experience taught me to take risks and avoid losing focus when things get challenging.”