Related Experiment Video
Updated: May 7, 2026
06:48
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
8.5K
Computer-aided segmentation system for breast MRI tumour using modified automatic seeded region growing (BMRI-MASRG)
Ali Qusay Al-Faris1, Umi Kalthum Ngah, Nor Ashidi Mat Isa
1Imaging and Computational Intelligence Research Group (ICI), School of Electrical & Electronic Engineering, Universiti Sains Malaysia, Penang, Malaysia, alialfaris2009@gmail.com.
Journal of Digital Imaging
|October 9, 2013
Summary
This study introduces an automated system for breast magnetic resonance imaging (MRI) tumor segmentation. The novel method significantly improves segmentation accuracy compared to existing approaches, aiding in diagnosis.
Area of Science:
- Medical Imaging
- Computer-Aided Diagnosis
- Biomedical Engineering
Background:
- Accurate breast tumor segmentation in magnetic resonance imaging (MRI) is crucial for diagnosis and treatment planning.
- Existing segmentation methods often require manual intervention or lack sufficient accuracy.
- Developing automated and precise segmentation tools is a key challenge in medical image analysis.
Purpose of the Study:
- To present an automated computer-aided detection (CAD) system for breast MRI tumor segmentation.
- To enhance tumor segmentation accuracy using a modified automatic seeded region growing algorithm.
- To validate the system's performance on a recognized breast MRI dataset.
Main Methods:
- Pre-processing steps including breast skin detection and deletion using level set active contour and morphological thinning.
- Modified automatic seeded region growing algorithm with automated initial seed and threshold selection.
- System validation on 40 test images from the RIDER breast MRI dataset.
Main Results:
- The developed system demonstrated statistically significant improvements in segmentation performance.
- Key evaluation metrics showed significant results: relative overlap (p=0.0002), misclassification rate (p=0.045), true negative fraction (p=0.0001), and sum of true volume fraction (p=0.0001).
- Performance was superior compared to previous segmentation approaches tested on the same dataset.
Conclusions:
- The proposed automated system offers a robust and accurate solution for breast MRI tumor segmentation.
- The modified seeded region growing algorithm with automated parameter selection is effective.
- This advancement holds potential for improving the efficiency and reliability of breast cancer diagnosis.
