Page 146 - AIH-2-4

Artificial Intelligence in Health AI-driven personalized learning in residency

1. Introduction instruction in undergraduate medical education, their
application in graduate medical education, particularly
Artificial intelligence (AI) has recently transformed within the time-constrained and high-stakes context of
medical education by offering personalized, adaptive residency training, remains underexplored. Given the
9
learning experiences and content delivery. Traditional increasing demands for competency-based education
medical residency training programs (graduate medical and personalized learning environments in graduate
education) face challenges in meeting the diverse learning medical education, integrating AI into residency curricula
needs of trainees, as standardized approaches often fail may represent a transformative step in optimizing
to accommodate individual differences in knowledge knowledge acquisition, supporting just-in-time learning,
acquisition and clinical preparedness. Flipped classroom and ultimately improving patient care outcomes across
1
models, which shift passive learning outside the classroom 10-12
and emphasize active engagement during instructional healthcare systems. In this vein, through real-time
adaptation, the AI beings provide personalized support,
sessions, have demonstrated improved learner engagement, ensuring residents receive targeted reinforcement in
comprehension, and new language acquisition. The areas of difficulty. Prior research has demonstrated that
2,3
evolution of AI in education has followed both incremental AI-assisted learning enhances engagement, reduces study
and disruptive paths, supporting personalized learning
through increasingly sophisticated cognitive models time, and improves performance in formative as well as
13,14
that could enhance the flipped classroom methodology. summative assessments. Given the increasing demand
4
When enhanced by AI, flipped classroom models further for competency-based medical education, AI-driven
personalize instruction, offer real-time feedback, and platforms have the potential to bridge gaps in traditional
significantly accelerate the acquisition of new skills and learning methods by offering scalable, data-driven
solutions tailored to individual learning needs.
languages by adapting content to individual learning needs
and preferences. 3,5 Accordingly, the present study sought to evaluate
Traditional flipped classroom models, while effective the feasibility of integrating adaptive AI beings into a
in promoting active learning, often fail to account for the flipped classroom model for internal medicine residents.
heterogeneity of learner progression, intrinsic cognitive Addressing the aforementioned deficiencies of earlier
load, and the unpredictable demands of clinical training models, the AI-driven platform used in this study
environments. In fact, traditional flipped classroom leverages natural language processing and machine
6
models are inherently limited by their reliance on pre- learning algorithms to assess learner’s progress and
designed, static, repetitive content and a lack of responsive optimize instructional delivery. This novel integration
1,7
feedback mechanisms. These models often fail to account aims to advance the pedagogical utility of flipped learning
for the heterogeneity of learner progression, cognitive load, by offering a scalable, data-informed solution that delivers
and the dynamic nature of clinical training environments. real-time personalization within the demanding context of
As a result, residents may struggle to bridge understanding graduate medical education.
gaps or receive timely clarification on complex topics, 2. Data methods
particularly when self-directed study occurs outside of
scheduled instructional time. In this context, AI-driven 2.1. Study design and setting
educational platforms offer a compelling advancement This study employed a feasibility design to evaluate
by introducing real-time adaptability into the flipped integrating an AI-driven learning platform into an internal
classroom structure. A systematic review of AI applications medicine residency program. The AI-enhanced flipped
in higher education reveals widespread use in intelligent classroom model was implemented at HCA Florida Oak
tutoring, predictive analytics, and learning support tools. Hill Hospital’s Internal Medicine residency program, with
8
These systems can analyze user input, track performance participation of residents from post-graduate years 1 to 3.
patterns, and deliver content dynamically tailored to The 6-month educational intervention examined resident
individual learning trajectories. engagement, learning efficiency, and performance changes
Through automated feedback loops and continuous in preparation for in-service examinations. Residents
assessment, AI platforms can identify areas of difficulty, were introduced to the AI platform through detailed
reinforce core concepts, and adjust content delivery orientation sessions and continuous access to the platform
accordingly – features that are otherwise challenging to throughout the study period. The platform provided
implement in traditional didactic or blended formats. adaptive learning pathways tailored to each resident’s
1
While early studies have demonstrated the feasibility performance. Engagement was monitored through various
and learner satisfaction associated with AI-enhanced metrics, including total study time, frequency and duration

Volume 2 Issue 4 (2025) 140 doi: 10.36922/AIH025130023