Related Experiment Video
Updated: Sep 30, 2025
09:11
Automation of the Micronucleus Assay Using Imaging Flow Cytometry and Artificial Intelligence
Published on: January 27, 2023
2.3K
A Computer-Aided Diagnosis System of Fetal Nucleated Red Blood Cells With Convolutional Neural Network
Chao Sun1, Ruijie Wang2, Lanbo Zhao1
1From the Department of Obstetrics and Gynecology, The First Affiliated Hospital of Xi'an Jiaotong University, Xi'an, Shannxi, China (Sun, Zhao, Han, Ma, Liang, L. Wang, Tuo, Zhang, Li).
Archives of Pathology & Laboratory Medicine
|March 16, 2022
Summary
This study developed a computer-aided diagnosis system using a convolutional neural network for rapid and accurate recognition of fetal nucleated red blood cells (fNRBCs). The system achieved high accuracy in identifying fNRBCs from umbilical cord blood samples.
Area of Science:
- Biomedical Engineering
- Computational Pathology
- Hematology
Background:
- Rapid recognition of fetal nucleated red blood cells (fNRBCs) is challenging.
- Accurate identification of fNRBCs is crucial for prenatal diagnostics.
Purpose of the Study:
- To establish a computer-aided diagnosis (CAD) system for rapid fNRBC recognition.
- To utilize a convolutional neural network (CNN) for automated fNRBC identification.
Main Methods:
- fNRBCs were isolated from umbilical cord blood using density gradient centrifugation and magnetic-activated cell sorting.
- The cell-block method and H&E staining were employed for sample preparation.
- A CNN-based CAD system was developed for feature discrimination and fNRBC recognition, including automated cell segmentation and feature vector encoding.
Main Results:
- The system was trained and tested on 4760 images from 260 cell slides.
- In the test set, the CNN achieved 96.5% sensitivity, 100% specificity, and 98.5% accuracy.
- The training set demonstrated 100% accuracy.
Conclusions:
- A computer-aided diagnosis system based on CNN was successfully established.
- The developed system provides effective and accurate recognition of fNRBCs.
- This technology holds potential for improving prenatal diagnostic efficiency.
