Loading...
QUALITATIVE ANALYSIS OF AI IN SUMMARIZATION OF ORTHODONTIC RESEARCH ARTICLES
Wittstein, Matthew
Wittstein, Matthew
Citations
Altmetric:
Genre
Thesis/Dissertation
Date
2025-08
Advisor
Committee member
Group
Department
Oral Biology
Subject
Permanent link to this record
Collections
Research Projects
Organizational Units
Journal Issue
DOI
https://doi.org/10.34944/vna3-4f89
Abstract
Objectives: Applications of artificial intelligence (AI) in health sciences have taken over the previous systems to process, analyze, and utilize information. Currently, a lesser application of AI is information synthesis in literature. Specific aims of this study include: (1) Evaluate capabilities of Unriddle.ai in summarizing orthodontic research articles. (2) Compare quality and comprehensiveness of summaries to orthodontic research articles by AI-driven programs versus humans in the orthodontic field. (3) Analyze potential strengths and weaknesses of utilizing, or not utilizing, an AI-driven program for information synthesis. (4) Discuss implications for the use of artificial intelligence to advance dental knowledge, and practical implications for orthodontists and other dental professionals.
Methods: 32 articles published in the American Journal of Orthodontics and Dentofacial Orthopedics from November 2023 to July 2024 were selected for this study. Summaries and critical analyses were collected for each article and similarly, one was generated by each of the AI-models available on Unriddle.ai: GPT-4.0 and Claude-3.5-Sonnet. These summaries and analyses were qualitatively compared to evaluate differences in quality and comprehensiveness.
Results: The results of the study revealed Unriddle.ai, utilizing the language model Claude-3.5-Sonnet, demonstrated a commendable ability to summarize orthodontic research articles in a concise manner. However, a comparative analysis with human-generated summaries indicated while the AI program was generally effective, it occasionally omitted critical details that a human summarizer included. Critical analyses produced by AI were notably different, with human-generated insights often reflecting a deeper contextual interpretation and clinical understanding.
Conclusions: AI models had strengths in efficiency; contrasted by its limitations in comprehensiveness and deeper clinical understanding and interpretation. Overall, the findings suggest that while AI-driven programs can serve as valuable tools for information synthesis in orthodontics, reliance solely on these systems may lead to gaps in understanding crucial nuances present in the literature.
Description
Citation
Citation to related work
Has part
ADA compliance
For Americans with Disabilities Act (ADA) accommodation, including help with reading this content, please contact scholarshare@temple.edu