Related Experiment Video
Updated: Mar 10, 2026
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
1.3K
Assessing the capabilities of AI-based large language models (AI-LLMs) in interpreting histopathological slides and
Khanisyah E Gumilar1,2, Grace Ariani3, Priangga A Wiratama3
1Graduate Institute of Biomedical Science, China Medical University, Taichung, Taiwan, ROC.
Biomedicine
|March 9, 2026
Summary
ChatGPT-4 demonstrated superior performance in interpreting histopathology and scientific images compared to other AI-LLMs. This advancement holds potential for improving medical diagnostics and enhancing scientific education.
Area of Science:
- Biomolecular Sciences
- Medical Imaging Analysis
- Artificial Intelligence in Healthcare
Background:
- Artificial intelligence-large language models (AI-LLMs) are emerging tools for complex scientific tasks.
- AI-LLMs can enhance comprehension and accessibility of scientific information, aiding professionals and students.
- Applications include interpreting histopathology slides and scientific figures for diagnostics and education.
Purpose of the Study:
- To evaluate the capability of AI-LLMs in interpreting histopathological slides and scientific images.
- To assess AI-LLM performance in supporting diagnostics and improving comprehension in biomolecular sciences.
Main Methods:
- Two-part study: interpretation of histopathology slides and scientific figures.
- Three leading chatbots (ChatGPT-4, Gemini Advanced, Copilot) tested on 12 images each.
- Expert raters evaluated responses on relevance, clarity, depth, focus, and coherence using a 5-point Likert scale; statistical analysis included ANOVA and regression.
Main Results:
- ChatGPT-4 significantly outperformed Gemini Advanced and Copilot in interpreting both histopathology and scientific images (P < 0.001).
- ChatGPT-4 achieved higher scores across all evaluated parameters: relevance, clarity, depth, focus, and coherence.
- Superior performance attributed to advanced algorithms, extensive training data, specialized modules, and user feedback.
Conclusions:
- ChatGPT-4 excels in scientific image interpretation, potentially improving diagnostic accuracy and reducing pathologist workload.
- Enhanced understanding of complex images benefits students and promotes interactive learning.
- ChatGPT-4 shows significant potential to improve patient care and enrich educational experiences.
