NLP Seminar: Sehi L’yi, Democratizing Biomedical Data with Visualization and AI

Content

Image

Sehi L’yi

Speaker

Sehi L’yi (Harvard University)

Abstract

Biomedical data, such as genomics and electronic health records (EHRs), hold immense potential to advance our understanding of human health and precision medicine. However, their size and complexity make them challenging to analyze, even with the recent innovations in artificial intelligence (AI). I develop interactive visualization systems as strategic tools to make this data more accessible and transparent to biomedical experts. Unlike prior work, my visualization systems scale to large biomedical data and support a wide variety of real-world visualization designs while significantly lowering technical barriers to use.

In this talk, I will demonstrate how these systems allow biomedical experts to build and use expressive biomedical data visualization in real-world settings. I will discuss Gosling, a declarative grammar that provides the foundational infrastructure for AI-driven, interactive, and multiscale biomedical data visualization; Chromoscope, a scalable system built on Gosling that enables the analysis of large-scale cancer cohorts; and Blace, which integrates Large Language Models (LLMs) with Gosling to allow biomedical experts to create complex visualizations with ease. I will describe how human-centered design and scalable software architectures can empower biomedical experts to navigate the increasing complexity of modern data ecosystems.

Speaker Bio

Sehi L'Yi is an NIH K99/R00 Postdoctoral Fellow in Biomedical Informatics at Harvard Medical School. He develops and deploys interactive visualization systems that bridge the gap between human–computer interaction (HCI), AI, and biomedical informatics. Before joining Harvard, he received his Ph.D. in Computer Science and Engineering from Seoul National University. His research is published in leading interdisciplinary venues, including Nature Methods, Bioinformatics, PLoS Computational Biology, IEEE VIS, ACM CHI, and ACM UIST. Dr. L’Yi’s work is recognized by several prestigious honors, including NIH K99/R00 Pathway to Independence Award, Best Paper Honorable Mention at IEEE VIS, and Best Abstract Award at ISMB BioVis. His visualization systems have been widely adopted in academia and industry; notably, his visualizations are integrated into cBioPortal, one of the world’s most utilized platforms for cancer researchers and clinicians.