It’s a scene echoed in thousands of dorm rooms and basements around the world: one person playing a video game, onlookers craning eagerly over his shoulder. But this player, Nawwaf Aleisa, doesn’t have a controller in his hands. The penguin racing down a snowy slope on his screen is controlled by Aleisa’s brain signals, captured and translated by the headband he’s wearing.
Aleisa is an undergraduate researcher working with Xiaodong Qu, an assistant professor of computer science in the George Washington University School of Engineering and Applied Science. A junior computer science major and the lab’s most proficient racer—“I guess they do talk about me like I’m the champion, but it took a lot of practice”—Aleisa and his teammates are working on one of the most compelling technological frontiers of the 21st century: the brain-computer interface, or BCI.
A BCI is a technology that establishes direct communication between the human brain and an external device, like a computer or prosthetic. Brain signals are captured and processed to identify patterns that correspond with user intent. An algorithm then analyzes that data in real time, translating neural activity into triggers for specific action in the external device, like moving a prosthetic limb or controlling an onscreen character. In Qu’s lab, for instance, players first calibrate the algorithm by thinking “turn right,” “turn left,” “move forward” and “move back” when prompted. Then the program refines the patterns it has detected, eliminating irrelevant data from facial movements, blinking or other “noise” that can interfere with a model’s accuracy, and applies those patterns to the movement of the penguin down his ice slide.
Applications of the BCI could be endless, and perhaps their most valuable quality is the speed at which they could theoretically operate. The brain’s neural network is faster than any human-created system, Qu said, with an ability to communicate complex concepts almost instantaneously. By investigating the BCI, researchers hope to link our powerful internal processor with an external network to make person-to-computer and person-to-person communication swifter and more complete. If realized, such technology could potentially both save time and enable deeper understanding—whether between professors and students, researchers and their collaborators, medical professionals and their patients or any other relationship that involves transmission of knowledge.
“When I’m talking to students, I can sentence by sentence deliver the message,” Qu said. “[But] if my brain can directly connect to the internet, the speed will be much faster than I can speak, or I can type.”
Brain health and the BCI
The penguin-racing game is an attention-grabbing demonstration, but commercial applications of the BCI, like gaming, are just half the focus of Qu’s lab. The other half is clinical, applying machine learning methods to facilitate faster, more accurate and more affordable electroencephalogram (EEG) data collection and predictive analysis.
An EEG is a noninvasive procedure that monitors neurological activity through small metal disks called electrodes, which, when placed on a patient’s scalp, detect the electrical signals passing between brain cells. These signals are amplified and translated onto a display, where they appear as waveforms. The penguin gaming interface uses relatively simple headsets with five electrodes resting against the forehead and behind the ears; clinical EEGs use quadruple that number, while researchers may apply 64 electrodes or more to a subject.
Due to the number of sensors and the complexity of the signals they capture, this data can require an enormous amount of time and processing power to analyze. Much of Qu’s lab’s recent work has focused on machine-learning methods that streamline that process, more efficiently translating brain signals—often from patients with debilitating neurological conditions, including Alzheimer’s, Parkinson’s and depression—into legible information. From there, clinicians can make data-driven decisions on care, while researchers can spot patterns and investigate possible solutions.
One of Qu’s research projects used machine learning to analyze EEG signals from patients experiencing first-time psychosis. The algorithm used pattern recognition to identify clinically and functionally distinct subgroups among these patients—machine-learning-driven classifications that, the research found, corresponded to meaningful differences in cognitive and functional outcomes.
“This approach has the potential to help clinicians stratify patients more effectively, leading to earlier interventions tailored to specific neurophysiological profiles,” Qu said. “The ability to extract actionable insights from complex brain data not only shortens research timelines but also enhances clinical decision-making for mental health disorders.”
Fertile ground for student research
For Qu, getting students excited is a crucial part of his job—instilling a love for learning in the next generation of computer scientists. “Research is hard,” he said. “So especially for young researchers in their sophomore or junior year, their passion, their interests, their motivation is so important.”
And he’s proud to bring students along as part of his research vanguard. Qu mentors both his undergraduate and graduate researchers, who’ve published papers in peer-reviewed journals and presented papers at prominent conferences. In the summer of 2024, Aleisa traveled to Barcelona to showcase his research at the International Conference on Knowledge Discovery and Data Mining.
“Traveling to Barcelona to present our research on generative AI was an incredibly meaningful experience for me,” Aleisa said. “It was both an opportunity and a responsibility to represent our work, Professor Qu’s lab and GW on an international stage. Being able to share our research with a global audience, engage with leading experts, and receive feedback from top minds in the field was both humbling and exhilarating.
“Additionally, it was inspiring to see the breadth of AI applications being explored worldwide. The experience deepened my passion for research and motivated me to continue pushing the boundaries of generative AI and machine learning.”
Aleisa and another undergraduate student in Qu’s lab, Zeina Nweasha, also attended the 2025 American Association for the Advancement of Science (AAAS) annual meeting in Boston, where they demonstrated the racing game for peers and potential collaborators.
“What I enjoy most with working with Professor Qu and specifically in this area is that he makes a great environment to work in and he encourages us to try new things,” Aleisa said. “I just like how it balances creativity and practicality together.”
Emily Flanagan, a sophomore double majoring in computer science and business, is another of Qu’s undergraduate researchers. For her, Qu’s dual focus on the clinical and commercial applications of the BCI comes naturally—and being part of the lab has given her a taste of a collaborative professional future.
“Getting involved as an undergraduate was huge for me, because it allowed me to learn how to work with a team,” Flanagan said. “A lot of our assignments in our computer science courses are specifically individual, that we learn how to code on our own. But it's super important that we also learn how to interact with a team, because in the job market you will be working with other people. It’s given me the ability to communicate complex ideas with other people and really learn how to break down a problem into simpler steps.”
For Qu, the BCI is the site of the next major scientific paradigm shift. “We envision that if we know a little bit about how the brain works and connect the brain to the computer a little bit more, we may have the next big technology innovation breakthrough similar to the AI breakthrough we are currently experiencing.”
That next breakthrough could entail revolutionary new ways to map, assist, augment or heal neurological and motor functions. And Flanagan, Aleisa and Qu’s other students will be on the wave’s breaking edge.
“It almost feels like something out of science fiction,” Aleisa said. “But actually bringing that to life.”
