Empowering Equality: AI and Gender Equity in Otolaryngology

PROGRAM DESCRIPTION: This session will address the critical issue of gender bias in artificial intelligence (AI) systems and its implications for the field of otolaryngology (ENT) as it relates to patient care and women practicing within the field. As AI technologies are increasingly integrated into healthcare, they hold significant potential to enhance diagnostic accuracy and treatment decision-making. Because AI is based off of pattern recognition, errors and biases in the training dataset are perpetuated into AI recommendations. This bias that results is referred to as algorithmic bias, which can result in systematic and unfair discrimination embedded in AI systems. These biases can lead to disparities in care, misdiagnoses, and unequal treatment recommendations.

An example of algorithmic bias is illustrated in a study by Seyyed-Kalantari et al. (2021) where they examined underdiagnosis in chest x-ray pathology classification. This study found systematic underdiagnosis biases in under-served subpopulations, including female patients, Black patients, Hispanic patients, and those of lower socioeconomic status. Garnering better understanding of AI and the implicit bias in the system is a crucial component to utilizing AI.

Beyond patient outcomes, AI has the potential to perpetuate bias amongst the professional field of ENT. In a study looking at the trends of women ENT, Ladlie et. al (2024) found remarkably low numbers of women in leadership positions, and as award and grant recipients. In a study by Menz et al. (2024) that investigated AI language learning models (LLM), they found she/her pronouns were assigned to 98% of nurses in prompts. Furthermore, they found that for certain LLM’s, stories focusing on senior medical doctors and surgeons were less likely to be she/her than stories focusing on junior medical doctors and surgeons.

The goal is to foster a greater understanding of the ethical and practical challenges of AI in otolaryngology and equip practitioners with the knowledge and tools to mitigate these biases in their clinical environments.

OUTCOME OBJECTIVE 1: Understand the concept of algorithmic bias in AI systems and its potential impacts on healthcare.

OUTCOME OBJECTIVE 2: Examine the manifestations of gender bias in the context of Otolaryngology, and explore the specific challenges faced by women and underrepresented groups within the medical field as influenced by AI models and data.

OUTCOME OBJECTIVE 3: Explore strategies for mitigating and foster discussion on the role of healthcare professionals in ensuring AI systems are used responsibly, without reinforcing societal inequities.

BACKGROUND STATEMENT: AI continues to play an increasingly prominent role in healthcare ad multiple facets of AI have already become integrated into the clinical setting. Physicians need to have a better understanding of the implicit biases reflected in AI and ways to mitigate them.

Comprehensive Otolaryngology Track Statement :