Related Experiment Video
Updated: Jul 30, 2025
04:25
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
Published on: December 15, 2023
2.5K
Segmentation and classification of brain tumors using fuzzy 3D highlighting and machine learning
Khalil Mowlani1, Mehdi Jafari Shahbazzadeh2, Maliheh Hashemipour1
1Department of Computer Engineering, Kerman Branch, Islamic Azad University, Kerman, Iran.
Journal of Cancer Research and Clinical Oncology
|May 11, 2023
Summary
Accurate brain tumor classification is vital for early cancer diagnosis. Machine learning methods, including genetic algorithm-deep neural networks (GA-DNN) and grasshopper optimization algorithm-support vector machines (GOA-SVM), achieved over 97% accuracy in classifying tumors.
Area of Science:
- Medical imaging analysis
- Artificial intelligence in oncology
- Neurological diagnostics
Background:
- Brain tumors are a leading cause of cancer death, necessitating early detection.
- Machine learning (ML) aids radiologists in non-invasive tumor diagnosis.
- Challenges include effective deep learning framework development and time-consuming manual segmentation of tumors.
Purpose of the Study:
- To develop and evaluate an automated method for brain tumor segmentation and classification using ML.
- To compare the performance of two distinct ML approaches for improved diagnostic accuracy.
Main Methods:
- A fuzzy 3D highlighting technique was used for brain tumor segmentation from MRI scans.
- Feature extraction was performed on segmented tumor regions.
- Classification employed a genetic algorithm-deep neural network (GA-DNN) and a grasshopper optimization algorithm-support vector machine (GOA-SVM).
Main Results:
- Both GA-DNN and GOA-SVM demonstrated high classification accuracy.
- The average classification accuracy reached 97.53% for GA-DNN and 97.65% for GOA-SVM.
- The proposed methods significantly outperformed existing alternatives.
Conclusions:
- The developed ML strategies offer a rapid and effective approach for brain tumor detection.
- These findings represent a significant advancement in neurological diagnosis, aiding in the identification of cancerous tumors and lesions.
