Related Experiment Video
Updated: Jun 11, 2025
08:59
Morphology-Based Distinction Between Healthy and Pathological Cells Utilizing Fourier Transforms and Self-Organizing Maps
Published on: October 28, 2018
7.1K
Image Enhancement Using Bidimensional Empirical Mode Decomposition and Morphological Operations for Brain Tumor
Giang Hong Nguyen1,2, Yen Thi Hoang Hua1,3, Linh Chi Nguyen1,3
1Department of Physics and Computer Science, Faculty of Physics & Engineering Physics, University of Science, Ho Chi Minh City, Vietnam.
Asian Pacific Journal of Cancer Prevention : APJCP
|September 30, 2024
Summary
This study introduces a novel three-step method for brain tumor image analysis, enhancing preprocessing for better noise reduction and contrast. The approach shows high accuracy in tumor detection and classification, paving the way for advanced computer-aided diagnosis (CADx) systems.
Area of Science:
- Medical Imaging
- Artificial Intelligence
Background:
- Brain image processing, including preprocessing, segmentation, and classification, is crucial for patient care.
- Existing methods require enhancement, particularly in noise removal and contrast improvement during preprocessing.
Purpose of the Study:
- To present a novel three-step method for brain tumor image analysis.
- To emphasize preprocessing techniques for noise reduction and contrast enhancement.
Main Methods:
- Utilized fast and adaptive bidimensional empirical mode decomposition, anisotropic diffusion, and modified top-hat/bottom-hat transforms for preprocessing.
- Employed Fast C-means clustering for tumor detection and ensemble learning for classification.
Main Results:
- Achieved 99% accuracy and specificity, with 90% sensitivity and precision in tumor detection.
- Developed an ensemble learning model with 96.7% training accuracy and 76.7% testing accuracy.
- Classified tumor images with 75% accuracy, misclassifying three pituitary tumor cases.
Conclusions:
- The proposed method demonstrates significant potential for developing computer-aided diagnosis (CADx) software.
- This technology can serve as a valuable reference tool for physicians in treating brain tumors.
