Related Experiment Video
Updated: Dec 10, 2025
08:20
Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
Published on: October 27, 2023
2.2K
Automated recognition of white blood cells using deep learning
Amin Khouani1, Mostafa El Habib Daho1, Sidi Ahmed Mahmoudi2
1University of Tlemcen, Tlemcen, Algeria.
Biomedical Engineering Letters
|August 28, 2020
Summary
This study presents a deep learning method for automatic white blood cell detection and segmentation in medical images. The efficient approach aids hematologists in cancer diagnosis, achieving 95.73% accuracy in under one second.
Area of Science:
- Medical Imaging
- Computational Biology
- Artificial Intelligence in Medicine
Background:
- Accurate detection, counting, and segmentation of white blood cells (WBCs) are crucial for diagnosing various cancers.
- Manual analysis of cytological images is time-consuming for hematologists.
Purpose of the Study:
- To develop an efficient deep learning-based method for automatic recognition and segmentation of WBCs in peripheral blood and bone marrow images.
- To reduce the workload of hematologists in clinical practice.
Main Methods:
- A deep neural network model was adapted for cell localization and segmentation after image pre-processing.
- Model outputs were refined using combined predictions and corrections.
- A novel algorithm integrating model results with spatial information enhanced segmentation quality.
Main Results:
- The proposed method demonstrated high efficiency, power, and speed compared to existing state-of-the-art techniques.
- Achieved a high accuracy of 95.73% for WBC detection and segmentation.
- Provided rapid predictions in less than one second.
Conclusions:
- The developed deep learning approach offers a promising solution for automated WBC analysis.
- The method significantly improves efficiency and accuracy in hematological image analysis.
- This tool can alleviate the burden on hematologists, potentially improving diagnostic workflows.
