Related Experiment Video
Updated: Mar 11, 2026
08:56
Evaluation of Lung Metastasis in Mouse Mammary Tumor Models by Quantitative Real-time PCR
Published on: January 29, 2016
10.4K
Accuracy of Automatic Quantitative Analysis of Pathological Tissue Images Using ChatGPT Data Analyst: Comparison With
Yuta Sannomiya1, Takuya Sakamoto2,3, Yuka Hiramatsu2
1Department of General and Digestive Surgery, Kanazawa Medical University, Kahoku, Ishikawa, Japan.
Microscopy Research and Technique
|March 10, 2026
Summary
ChatGPT Data Analyst offers a novel, accurate method for quantifying tumor regions in pathological images. This AI tool shows high correlation and agreement with traditional analysis, improving efficiency in cancer research.
Area of Science:
- Pathology
- Computational Biology
- Artificial Intelligence
Background:
- Colorectal cancer metastasis, particularly to the liver, significantly impacts patient prognosis.
- Accurate image analysis of pathological tissues is crucial for tumor research and diagnosis.
- Current manual or semi-automated methods for image analysis are often inefficient and require specialized expertise.
Purpose of the Study:
- To evaluate the efficacy of ChatGPT Data Analyst for automated quantification of tumor regions in H&E-stained liver metastasis mouse models.
- To compare the performance of ChatGPT Data Analyst with conventional ImageJ analysis for area and count measurements.
- To identify optimal staining protocols for enhanced accuracy in pathological image analysis.
Main Methods:
- Utilized ChatGPT Data Analyst to analyze H&E-stained liver metastasis sections from a mouse model (MC38 cells).
- Quantified tumor area, total liver area, tumor proportion, and tumor count.
- Compared results with ImageJ analysis and investigated optimal hematoxylin and eosin staining times for accuracy.
Main Results:
- ChatGPT Data Analyst demonstrated high correlation (r > 0.99) and agreement (ICC > 0.9) with ImageJ for area measurements, with excellent reproducibility.
- Identified specific staining conditions (hematoxylin 2 min, eosin 30 s) crucial for improving extraction accuracy.
- Benchmarking showed advanced LLMs achieved accuracy comparable to specialized AI frameworks and commercial software.
Conclusions:
- ChatGPT Data Analyst provides a novel, prompt-based approach for pathological image analysis with accuracy comparable to established methods.
- Optimized staining protocols enhance the reliability of AI-driven image analysis.
- Future updates and standardized conditions will further advance the application of LLMs in tumor research and diagnosis.
