Xuezhu Cai’s research combines two of her passions: medical imaging technology and machine learning. By marrying these two fields, Cai hopes to teach computers to read and translate medical images for human doctors and researchers, resulting in faster diagnoses and better treatment outcomes.
Since being accepted to Northeastern’s PhD program Bioengineering in 2015, Cai has worked and studied as a research assistant in the Center for Translational Neuroimaging (CTNI) under her advisor, Professor Craig Ferris. “When I started, the lab was mostly focused on neurodegenerative diseases like Alzheimer’s and Parkinson’s, which I was very interested in,” she recalls, “I feel that those are two very common diseases that cause people to suffer a lot. So, I hoped I could use medical imaging to better understand the diseases, how they progress, whether there were better ways to assess their effects.”
Since then, she has worked on a wide variety of research projects, from neuropathic pain to drug addiction to stress conditions, all using animal subjects. “In all of these areas, the goal is to find biomarkers for each condition and use that information to help with diagnosis and to assess treatment,” she explains, “we also do behavior studies which we use with the imaging data to better understand the problems we’re tackling.”
Her interest in machine learning manifested as a response to the repetitive and labor-intensive process of analyzing medical imaging scans, a time-consuming and often tedious process. “But if we can use machine learning to automate this process, that would be a big improvement in terms of speed,” says Cai, “Secondly, AI can be used to generate predictive modeling to help with diagnoses, or to recommend treatment paths that might not be considered otherwise.”
Cai used a 2018 summer internship with Sanofi Genzyme, a Cambridge biotech company, to test this theory. On her own time and out of pure interest, she tried to automate their MRI pipeline, applying machine learning algorithms in order to speed up the process of analyzing the images. She was successful, and felt the results were positive enough to pursue further. Today she is optimistic about both her prospects, and the future of machine learning in the medical imaging field. After completing a LEADERs Fellowship at MERCK, Cai is now Senior Scientist in the Image Analytics IT group there.
