DFP FILE
The Boston University College of Engineering. A project led by Archana Venkatarman, associate professor in the College of Engineering, and Dr. Swathi Kiran, director of the Center for Brain Research, recently received a $3.2 million grant from the National Institute of Health.
Boston University’s Center for Brain Research, which received a $3.2 million National Institutes of Health grant from the National Institute on Deafness and Other Communication Disorders, is spearheading a new project to develop tools that can reliably predict the outcomes of personalized treatment plans for patients with post-stroke aphasia.
Aphasia, a neurogenic language disorder that affects language comprehension and expression, is challenging to treat since the recovery paths vary amongst patients, according to the American Speech-Language-Hearing Association.
The initiative is being spearheaded by Archana Venkataraman, associate professor in the College of Engineering, and Swathi Kiran, director of the Aphasia Research Laboratory.
“It’s a very devastating impairment,” Venkataraman said. “Often patients can’t speak. They can’t understand and as you can imagine, it’s very detrimental to quality of life”
Very little is currently known to determine what treatment works best for a particular aphasia patient, or even how to identify whether they are recovering, Venkataraman said.
The project aims to address this gap by capturing images of the brain in different contexts through multimodal imaging, and then analyzing patterns within the brain’s neural wiring using artificial intelligence.
“We can leverage AI techniques and automatically get information to predict … [H]ow is this brain actually evolving in patients who improve or who respond to treatment or improve naturalistically over time?” she said.
What makes machine learning useful for the task, she said, is the human brain can process a limited amount of patterns, whereas AI sees these patterns and automatically learns them and their interactions.
“If you think about the sort of the level of patterns that a human brain can conceptualize, it’s fairly limited,” she said. “Through this other modality, the AI might be able to see it and automatically learn [it].”
Senior Lauren Dang, vice president of BU’s Mind and Brain Society, noted AI is becoming a more necessary and useful tool, especially in fields like neuroscience.
“Neuroscience is a very interdisciplinary subject, so as AI develops, a lot more students are having to learn how to work with it,” she said. “There’s definitely a rise in popularity with the computational path now, mostly because of the interest in computation and AI.”
AI is still a tool that requires caution, especially in computational research, Dang said.
“Things can slip through the cracks when you’re like, automating research or using AI for research,” she said. “But I think it is like a really useful tool, and it helps researchers process large chunks of data, which would usually take years and years to do.”
Sophomore Denali Schlesinger, vice president of the BU Artificial Intelligence society, said maintaining ethical responsibility in AI research is important.
“The data is never going to be not biased, but we have to understand how it’s biased,” he said. “And you have to set standards for that.”
Schlesinger added AI has the potential to make research projects more accessible.
“If you want to do a project in neurology, you used to have to be a professor, have 10 years of research to lead a project,” he said. “But now, since we have public data with a lot of open source data, [this] empowers anybody who has a problem.”
Venkataraman said while AI neural networks are a useful tool, they won’t solve all medical issues.
“There’s 500 years of knowledge, and it doesn’t make sense to throw all of that away,” she said. “I think that the best strategy is to figure out how to kind of bridge those two worlds.”