Related Experiment Video
Updated: Oct 4, 2025
Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone
Published on: April 8, 2015
Automated knowledge-assisted mitosis cells detection framework in breast histopathology images
Xiao Jian Tan1, Nazahah Mustafa2, Mohd Yusoff Mashor2
1Centre for Multimodal Signal Processing, Department of Electrical and Electronic Engineering, Faculty of Engineering and Technology, Tunku Abdul Rahman University College (TARUC), Jalan Genting Kelang, Setapak 53300, Kuala Lumpur, Malaysia.
Automated mitosis cell detection in breast cancer histopathology images is crucial for grading. This study introduces a knowledge-assisted framework that significantly reduces false positives, improving detection accuracy and outperforming existing methods.
Area of Science:
- Digital pathology
- Computational biology
- Medical image analysis
Background:
- Accurate mitosis cell detection is vital for breast carcinoma grading using the Nottingham Histopathology Grading (NHG) system.
- The heterogeneity of histopathology images presents challenges for automated mitosis detection.
Purpose of the Study:
- To develop an automated image processing framework for mitosis cell detection in breast histopathology images.
- To incorporate domain knowledge and histopathologist strategies to enhance detection accuracy and reduce false positives.
Main Methods:
- The framework employs color normalization and hyperchromatic nucleus segmentation.
- A novel knowledge-assisted false positive reduction method is introduced.
- Support Vector Machine (SVM) classifier is used for feature extraction and classification of mitosis candidates.
Main Results:
- The knowledge-assisted false positive reduction method successfully eliminated at least 87.1% of false positives across two datasets.
- The framework achieved high F1-scores: 89.1% on a custom dataset and 88.9% on the MITOS dataset.
- The proposed method demonstrated superior performance compared to recent works in mitosis detection.
Conclusions:
- The developed knowledge-assisted framework offers a promising approach for accurate automated mitosis cell detection in breast histopathology.
- This method effectively addresses the challenges posed by image heterogeneity and improves grading accuracy.
- The framework's performance suggests its potential for clinical application in breast cancer diagnosis.
