Interests:

Developing intuitive, user-centered tools for scientific discovery and data extraction from electronic health records, with a focus on natural language processing (NLP), large language models and autonomous agents, cohort discovery, and clinical trials.

Background:

Dr. Nic Dobbins received his BA in History and Asian Languages and Literatures from the University of Minnesota, his MLIS in Library and Information Science from the University of Washington (UW) iSchool, and his PhD in Biomedical Informatics from UW. Dr. Dobbins joined the faculty in Johns Hopkins Biomedical Informatics and Data Science in 2024.

Research:

Dr. Dobbins leads efforts in large-scale de-identification of clinical notes, evaluation and application of large language models, and development of cohort discovery tools. His doctoral dissertation pioneered NLP methods and fine-tuning of large language models to generate database queries for identifying patients eligible for clinical trials. This work was the first to empirically demonstrate that machine learning methods can surpass human programmers in identifying eligible patients using actual clinical trials and structured database queries on real-world data.
He is the creator and developer of Leaf, one of the most widely used cohort discovery tools globally. His extensive experience includes full-stack software engineering, user interface design, database schema design and management, and publishing high-impact peer-reviewed manuscripts on information extraction from clinical notes, large language models, automated biomedical reasoning, knowledge representation, social determinants of health, and de-identification.