Hariri Institute Graduate Student Fellow Spotlight: Xinyi Hu, CDS PhD Student

Hu’s research focus using multimodal data integration to predict neurodegenerative diseases

by John Eaton

Xinyi (Selena) Hu is an incoming Hariri Institute Graduate Student Fellow and a second year PhD student in the Center for Computing & Data Science studying algorithmic and statistical modeling, as well as machine learning. Since starting the doctoral program, Hu has been working with Professors Swathi Kiran (Sargent), Margrit Betke (Computer Science), and Engineering Professors Archana Venkataraman, Prakash Ishwar and Janusz Konrad on a project evaluating machine learning approaches to evaluate the predictive ability of structural and functional brain integrity on language recovery ability in post-stroke aphasia.

With a bachelor degree in Applied Mathematics and in Psychology from Tufts University, Hu seeks to further explore her interest in machine learning algorithms and the applications of these algorithms in language function neural deficiency. 

The Hariri Institute asked Hu about her research focus using multimodal data integration to predict neurodegenerative diseases:

Hariri Institute: Can you describe your research focus and its applications?

Hu: My research focuses on multimodal data integration, with a current emphasis on using this approach to predict neurodegenerative diseases. By combining various types of data—such as imaging and clinical information—I aim to deepen our understanding of neural substrates, improve the accuracy of early diagnosis, and develop more effective and individualized treatment plans for patients. 

Hariri Institute: How did you become interested in this? Was there something that inspired this area of interest?

Hu:  I have always been fascinated by the complexity and intricacy of the human brain. The potential of multimodal analysis to provide a more comprehensive understanding of brain function and disorders inspired me to pursue this research area. I believe that integrating diverse data sources can unlock new insights into how the brain works and develop customized treatment plans for patients with neural deficits.

Hariri Institute: What are the main goals or objectives of your research?

Hu:  While my research is still evolving, one of my main objectives is to explore the potential of using reinforcement learning models to assist aphasia patients. Additionally, I aim to investigate how large language models (LLMs) can help these patients communicate more effectively in their daily lives. By developing tools that cater to both aspects, I hope to customize treatment plans and ultimately improve their quality of life.

Hariri Institute: Has there been a recent development or finding that you find particularly exciting?

Hu: Yes, my lab has recently developed multimodal machine learning models that predict the severity of aphasia with greater accuracy and reliability than single-modality models. This advancement holds significant promise for enhancing diagnosis and treatment planning for patients with this condition.

Hariri Institute: What advice do you have for students entering the first year of a PhD program? 

Hu: It’s crucial to maintain a detailed record of your research process. PhD projects can span several years, and having comprehensive documentation will help you track progress, replicate results, and stay organized throughout your research journey.

Hariri Institute: How do you plan on using this fellowship opportunity?

Hu: I plan to use the fellowship funds to attend conferences. 

The Hariri Institute’s Graduate Student Fellowship recognizes outstanding PhD Student Researchers at BU and supports their development through collaboration with other Graduate Student Fellow cohorts to organize and support Institute-sponsored events and through internal networking between fellows and the wider Hariri Institute community. Learn more about current Graduate Student Fellows and the program here.