Skip to main content Skip to navigation

Advancing Problem-Based Learning in Biomedical Engineering in the Era of Generative AI

Project Overview

The document explores the integration of generative AI in biomedical engineering education, specifically through a Problem-Based Learning (PBL) framework. It underscores the necessity for effective AI education tailored to biomedical contexts, identifying challenges within traditional curricula. A three-year case study conducted at Georgia Tech and Emory University illustrates the positive outcomes achieved, including enhanced learning experiences, improved interdisciplinary collaboration, and the potential of generative AI to overcome limitations in scalability and resources. Overall, the findings suggest that generative AI can significantly enrich educational practices in biomedical engineering, making learning more effective and engaging while preparing students for future challenges in the field.

Key Applications

Integration of Generative AI as Educational Support Tools

Context: Graduate and undergraduate students in biomedical engineering courses at institutions such as Georgia Tech and Emory University, focusing on personalized instruction, virtual patient simulations, and automated feedback systems.

Implementation: Incorporating generative AI tools, including large language models, to enhance personalized learning experiences through virtual simulations and provide automated feedback, while employing a Problem-Based Learning (PBL) framework to address real-world biomedical AI challenges via interdisciplinary collaboration.

Outcomes: Enhanced personalized learning experiences, improved student engagement, high research productivity with 16 student-authored publications, positive peer evaluations, and strengthened capabilities in addressing biomedical challenges.

Challenges: Scalability of PBL, diverse student backgrounds, limited computational resources, ensuring accuracy of AI outputs, maintaining academic integrity, managing AI biases, and addressing ethical concerns.

Implementation Barriers

Scalability

Traditional Problem-Based Learning (PBL) requires considerable faculty resources and ongoing curricular updates, making it difficult to implement widely.

Proposed Solutions: Integrating advanced AI technologies to facilitate guided problem exploration and knowledge acquisition, thereby reducing faculty workload.

Accuracy

Generative AI may produce plausible yet incorrect information (hallucinations), posing risks in educational contexts.

Proposed Solutions: Implementing rigorous validation protocols and training students to critically evaluate AI outputs to mitigate risks associated with inaccuracies.

Ethical

Educational use of AI presents challenges related to privacy, data security, and academic integrity.

Proposed Solutions: Establishing clear institutional policies for AI use and ensuring transparency in AI's role in education to address ethical concerns.

Equity and Inclusion

Generative AI may reinforce existing educational inequalities, particularly for non-native speakers and diverse backgrounds.

Proposed Solutions: Developing inclusive strategies and monitoring AI effectiveness across demographic groups to ensure equitable access to education and address disparities.

Project Team

Micky C. Nnamdi

Researcher

J. Ben Tamo

Researcher

Wenqi Shi

Researcher

May D. Wang

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Micky C. Nnamdi, J. Ben Tamo, Wenqi Shi, May D. Wang

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18