Related Experiment Video
Updated: Jul 3, 2025
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
1.5K
ChatGPT's innovative application in blood morphology recognition.
Wan-Hua Yang1,2,3, Yi-Ju Yang1, Tzeng-Ji Chen4,5,6
1Department of Pathology and Laboratory, Taipei Veterans General Hospital Hsinchu Branch, Hsinchu, Taiwan, ROC.
Journal of the Chinese Medical Association : JCMA
|February 13, 2024
Summary
Generative artificial intelligence (AI) shows promise in identifying blood cell morphology, with ChatGPT-4 achieving 88% accuracy for normal cells. While it aids diagnosis, AI currently cannot replace expert hematopathologists for complex cases.
Area of Science:
- Hematology
- Artificial Intelligence
- Medical Diagnostics
Background:
- Advancements in generative AI, like ChatGPT-4, are impacting medical image recognition.
- Accurate hematological diagnosis, especially blood morphology identification, is crucial.
- Manual identification by hematopathologists is subjective and prone to errors, necessitating AI assistance.
Purpose of the Study:
- To investigate the potential of ChatGPT-4 in assisting blood morphology identification.
- To evaluate AI's accuracy in classifying normal and abnormal blood cell morphologies.
Main Methods:
- A retrospective study utilized blood images from the American Society of Hematology (ASH).
- Images included both normal and abnormal blood cell morphologies.
- ChatGPT-4 classification was compared against expert technician analysis.
Main Results:
- ChatGPT-4 achieved 88% accuracy in identifying normal blood cells.
- Accuracy for identifying abnormal blood cells was 54%, slightly surpassing the manual method's 49.5% accuracy.
- ChatGPT-4 demonstrated particular strength in recognizing red blood cell morphology and inclusion bodies.
Conclusions:
- Generative AI, including ChatGPT-4, has potential as an auxiliary tool for clinical diagnosis in hematology.
- Current AI capabilities do not replace the need for professional medical judgment.
- Further improvements in AI accuracy are necessary to enhance overall diagnostic standards.
