Applying Artificial Intelligence in Otolaryngology

Innovations in Ear, Nose & Throat | Fall 2024

As artificial intelligence evolves at a rapid pace, the health care industry is increasingly applying its capabilities to transform care delivery. Technologies, including machine learning, natural language processing and predictive analysis, are being leveraged to advance patient outcomes, enhance diagnostics, improve efficiencies and reduce costs.

In this article, two experts at the University Hospitals Ear, Nose & Throat Institute share how they are integrating AI into current practices and predict its potential impact on otolaryngology and medical education.

Sanjeet Rangarajan, MD

“AI and similar computational techniques really entered the public zeitgeist in 2022 when the public large language model, ChatGPT, was released,” says Sanjeet Rangarajan, MD, a rhinologist and skull base surgeon and Vice Chair of Innovation and Community Partnerships within the UH Department of Otolaryngology – Head and Neck Surgery. “Since then, people began to understand more tangibly what these advanced computational techniques can do in real life, and there has been incredible interest in how we can utilize AI throughout health care.”

Of course, the integration of AI raises significant questions about regulation, ethics and bias. “As we experience the evolution of these tools, how can we ensure that we are applying them appropriately?” Dr. Rangarajan says. “It is important that people working in health care and other public-facing industries have a basic vocabulary about what AI is, its limitations and how the algorithms are trained.”

AI and the Impact on Otolaryngology

Dr. Rangarajan has co-authored articles on the early adoption of AI within his specialty, including a preliminary review of AI utilization to detect eosinophilic chronic rhinosinusitis, published in the International Forum of Allergy & Rhinology. “Radiographic imaging and interpretation of pathology specimens using AI computer vision algorithms has long been of interest,” he says. “It is allowing us to detect anomalies, disease or tumors that may be missed by the human eye.” He adds that AI significantly increases capacity and throughput to analyze large numbers of scans for research purposes.

Dr. Rangarajan is particularly interested in the power of AI to build predictive algorithms to minimize post-surgical complications. “My practice includes endoscopic skull-based surgery, where we remove brain tumors through the nose,” he says. “The possibility for complications after these procedures—while rare—can be life-threatening. The ability to predict who is likely to have a complication based on intraoperative findings or patients’ comorbidities has tremendous potential to improve our outcomes.”

AI and the Impact on Medical Education

Nina Zhao, MD

As Associate Program Director of the UH Otolaryngology Residency Program, Nina Zhao, MD, MAEd, has been monitoring the national implications of AI in medical education. “Conversation about the role of AI in clinical assessments of medical students are surging,” says Dr. Zhao, a Laryngologist within the UH Department of Otolaryngology – Head and Neck Surgery and Assistant Professor at the School of Medicine. “We are seeing students use AI to develop study materials or summarize documents.” She adds that AI tools can reinforce learning by generating resources such as practice questions or item banks.

Of particular interest to Dr. Zhao is the potential bias in human-AI interaction. “There is so much potential, but machine learning is only as good as how we input data and train the algorithm,” she says. “How can we look at outputs with a critical eye to avoid the perpetuation of unspoken biases or thought patterns? A large body of scholarly research is under way to address how we can maintain control as we bring AI into the fold.”

AI is also showing potential as a tool to help residency programs screen applicants. “For some time, people have been using natural language processing to scan personal statements for patterns that may identify students more likely to be successful residents,” Dr. Zhao says.

She has reviewed several papers looking at the use of AI to assess student characteristics in their written applications.

Currently, there are proprietary algorithms that enable residency program directors input values and preferences and output a suggested interview list. “Our otolaryngology program receives 300 to 400 applications annually for four residency slots,” Dr. Zhao says. “There is certainly an opportunity for additional inquiry about whether a tool like this is appropriate and how to integrate it into the time-consuming selection process so that we are not overlooking talented students.”

Moving Forward with Intention

At University Hospitals, AI oversight committees are bringing together stakeholders, including clinical pillar groups, researchers, venture teams and legal experts, who are tasked with ensuring AI’s safe and responsible implementation. “Leaders have been very thoughtful with how to evaluate and deploy these advanced technologies across our health care enterprise,” Dr. Rangarajan says.

For more information, contact Dr. Rangarajan at Sanjeet.Rangarajan@UHhospitals.org or Dr. Zhao at Nina.Zhao@UHhospitals.org.

Contributing Experts:
Sanjeet Rangarajan, MD
Rhinologist – Skull Base Surgeon
Vice Chair of Innovation and Community Partnerships
Department of Otolaryngology – Head and Neck Surgery
University Hospitals Ear, Nose & Throat Institute
University Hospitals Cleveland Medical Center
Associate Professor
Case Western Reserve University School of Medicine

Nina Zhao, MD, MAEd
Laryngologist – Voice, Airway, and Swallowing Center
Associate Residency Program Director
Department of Otolaryngology – Head and Neck Surgery
University Hospitals Ear, Nose & Throat Institute
Assistant Professor
Case Western Reserve University School of Medicine