Related Experiment Video
Updated: Jun 7, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
Published on: July 5, 2024
Performance of a Generative Pre-Trained Transformer in Generating Scientific Abstracts in Dentistry: A Comparative
Caio Alencar-Palha1, Thais Ocampo1, Thaisa Pinheiro Silva1
1Division of Oral Radiology, Department of Oral Diagnosis, Piracicaba Dental School, University of Campinas, Piracicaba, São Paulo, Brazil.
Generative Pre-trained Transformer (GPT) models show promise in generating dental radiology abstracts that are indistinguishable from human writing. However, AI detection tools can identify GPT-generated content, revealing its artificial origin.
Area of Science:
- Dental Radiology
- Artificial Intelligence in Science
Background:
- Generative Pre-trained Transformer (GPT) models are increasingly used in scientific writing.
- Evaluating the efficacy of GPT in generating scientific abstracts is crucial for academic integrity.
Purpose of the Study:
- To assess the performance of GPT in creating scientific abstracts within the field of dentistry.
- To compare human and AI-generated abstracts for detectability and quality.
Main Methods:
- Ten original abstracts from dental radiology articles were collected.
- Ten new abstracts were generated using ChatGPT based on article methodology and results.
- Five evaluators assessed abstracts on accuracy, formatting, and writing quality.
- AI detection tools assessed 'Human Score' and plagiarism levels.
Main Results:
- Human evaluators found GPT-generated abstracts comparable to human-written ones.
- AI detection tools identified GPT-generated abstracts with an average "Human Score" of 16.9%.
- Orthography and punctuation were key indicators for AI-generated abstracts.
Conclusions:
- GPT demonstrates a strong capability in generating scientific abstracts that are difficult for humans to distinguish.
- AI detection tools are effective in identifying GPT-generated content, highlighting the need for transparency.
Related Concept Videos
Improving Translational Accuracy
Source Transformation for AC Circuits
Source Transformation
It is essential to note that when...
Extraction: Advanced Methods

