Related Experiment Video
Updated: Aug 5, 2025
04:23
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
1.9K
Deep convolutional neural networks using an active learning strategy for cervical cancer screening and diagnosis
Xueguang Li1, Mingyue Du1, Shanru Zuo1
1The Key Laboratory of Model Animals and Stem Cell Biology in Hunan Province, School of Medicine, Hunan Normal University, Changsha, Hunan, China.
Frontiers in Bioinformatics
|March 27, 2023
Summary
A novel deep convolutional neural network (DCNN) effectively screens cervical cancer (CC) using whole-slide images. Active learning improved labeling efficiency and accuracy for better CC diagnosis.
Area of Science:
- Oncology
- Medical Imaging
- Artificial Intelligence
Background:
- Cervical cancer (CC) is a significant global health concern for women.
- Accurate screening and diagnosis are crucial for CC prevention.
- Deep convolutional neural networks (DCNNs) show promise in medical image analysis.
Purpose of the Study:
- To develop a robust DCNN for cervical cancer screening using whole-slide images (WSI).
- To evaluate the efficacy of an active learning strategy in improving DCNN performance and labeling efficiency.
Main Methods:
- Utilized whole-slide images (WSI) from ThinPrep cytologic test (TCT) slides of 211 CC patients and 189 normal cases.
- Implemented a deep convolutional neural network (DCNN) architecture for image analysis.
- Employed an active learning strategy for enhanced image labeling compared to traditional supervised methods.
Main Results:
- The best DCNN model achieved high performance metrics: 96.21% sensitivity, 98.95% specificity, and 97.5% accuracy for CC identification.
- Active learning demonstrated superiority over supervised learning in reducing costs and enhancing image labeling quality.
- The study provides freely available data and source code for reproducibility.
Conclusions:
- The proposed DCNN model offers a robust and accurate approach for cervical cancer screening.
- Active learning is an effective strategy for optimizing DCNN development in medical diagnostics.
- This research contributes to advancing AI-driven solutions for cervical cancer prevention and diagnosis.
