SUMMARY
Santiago Romero-Brufau, M.D., Ph.D., is the director of AI and Systems Engineering in the Department of Otolaryngology — Head and Neck Surgery. His work focuses on the development of machine learning, deep learning and other data science tools. This includes the implementation of those tools into clinical practice.
Dr. Romero-Brufau uses a pragmatic combination of systems engineering, data science and machine learning operationalization (MLOps) to bridge the gap from model development to clinical impact. He is particularly interested in developing best practices that lead to better translation of models from the lab into clinical practice.
Dr. Romero-Brufau's professional background includes these accomplishments:
- Consultation services provider for hospitals, health ministries, and universities across the U.S., Latin America, Africa and Europe.
- Inventor of intellectual property licensed to multiple third-party companies.
- Adviser for several health tech AI startups.
Dr. Romero-Brufau also has a keen interest in data science education. He is an adjunct assistant professor in the Department of Biostatistics at the Harvard T.H. Chan School of Public Health. There he teaches two core courses for the Health Data Science Master's Program: one on deep learning and another covering the implementation of data science solutions into clinical practice.
At Mayo Clinic, Dr. Romero-Brufau is involved with the Clinical Informatics Fellowship and has developed courses to teach clinicians about data science and machine learning. He also has been invited to teach health care leaders in America, Europe, Africa and Asia.
Focus areas
- Automation of clinical registries with the use of large language models. Clinical registries often need to manually extract information from free-text clinical notes and operative reports. Dr. Romero-Brufau's group is building natural language models and pipelines to automatically process the free-text data into structured fields that can feed the registries.
- Improving the efficiency of patient triage. As a quaternary care center, Mayo Clinic receives many referral requests. It's important to prioritize patients with complex care needs who would benefit the most from specialized care. Dr. Romero-Brufau's team is designing a system that uses AI and machine-learning models to efficiently triage patient requests and referrals.
- Using large language models for semi-automation of the response to patient messages. People frequently send queries, which make up a significant volume of the messages that are received.
- Identifying candidates for cochlear implants. Some people with serious hearing loss can improve their hearing with the help of cochlear implants. These implants are often covered by Medicare as well as most insurance. However, it is estimated that only 3% to 15% of people who would benefit from a cochlear implant receive one. Dr. Romero-Brufau and his colleagues are building machine-learning models to improve the identification of people who would benefit from cochlear implants.
Significance to patient care
Dr. Romero-Brufau's work has a direct impact on the efficiency and quality of patient care. Improving the efficiency of health care delivery is the next frontier to improve health care access, reduce cost and improve quality. His work focuses on three pillars:
- Developing tools that help health care professionals make better patient-focused decisions that are informed by data.
- Automating routine tasks, so that health care professionals are free to focus on human-centered tasks that require their unique expertise.
- Increasing process efficiency that can result, for example, in an increased number of surgical cases allowing more patients to have quicker access to ENT and head and neck surgery.
Professional highlights
- Department of Otolaryngology — Head and Neck Surgery, Mayo Clinic.
- Member, ENT Innovation Group, 2023-present.
- Member, ENT Research Committee, 2021-present.
- Member, Enterprise AI Translation Advisory Board, Mayo Clinic, 2022-present.
- Member, executive committee, Master's in Health Data Science, Harvard T.H. Chan School of Public Health, 2021-present.
- Certificate of Distinction in Teaching, Harvard University, 2021-2022.
- Health Informatics Certification, American Medical Informatics Association, 2013.