Related Experiment Video
Updated: Nov 9, 2025
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
496
Cervical cytology screening facilitated by an artificial intelligence microscope: A preliminary study.
Hong-Ping Tang1, De Cai2, Yan-Qing Kong1
1Department of Pathology, Affiliated Shenzhen Maternity and Child Healthcare Hospital, Southern Medical University, Shenzhen, China.
Cancer Cytopathology
|April 7, 2021
Summary
An AI microscope improves cervical cancer screening by assisting pathologists in real time. This technology enhances diagnostic accuracy and efficiency for cervical cytology, aiding in the early detection of abnormalities.
Area of Science:
- Medical technology
- Artificial intelligence in healthcare
- Pathology
Background:
- Cervical cytology screening is often time-consuming and lacks consistent diagnostic accuracy.
- An artificial intelligence (AI) microscope has been developed to offer real-time, onsite diagnostic support for cervical cytology.
Purpose of the Study:
- To evaluate the performance of an AI microscope in assisting cervical cytology screening.
- To determine if AI assistance improves diagnostic accuracy and consistency compared to manual reading.
Main Methods:
- A reader study was conducted using 486 cervical cytology slides, comparing manual reading with AI-assisted reading.
- The AI microscope highlighted potentially abnormal cells in real time, providing binary and multiclassification results according to the Bethesda system.
Main Results:
- AI assistance significantly improved sensitivity for detecting low-grade squamous intraepithelial lesions (0.837 to 0.923) and high-grade squamous intraepithelial lesions (0.830 to 0.917).
- Inter-observer agreement, measured by κ scores, improved for atypical squamous cells of undetermined significance (ASCUS) and for both binary and multiclassification tasks.
Conclusions:
- The AI microscope effectively provides real-time assistance for cervical cytology screening.
- The technology enhances both the efficiency and accuracy of cervical cytology diagnosis.
