Morphological and Marker-based Watershed Method for Detection and Segmentation of Brain Tumor Regions
DOI:
https://doi.org/10.24949/njes.v16i2.759Abstract
Brain tumor detection is a complex problem in medical image analysis. Brain tumor is an abnormal growth of brain cells
that is usually detected by Magnetic Resonance Images (MRI). In this paper, we propose an efficient algorithm for detecting brain tumors using MRI without skull removal. After applying basic image preprocessing techniques, morphological operations are used to detect the boundaries and sharpen the regions.Segmentation is performed using Otsu thresholding and then a marker watershed technique is used for final brain tumor segmentation. The proposed approach is evaluated on 3000 images from Brain Tumor Image Segmentation Challenge (BRATS) dataset. Experiments showed that with appropriate preprocessing and appropriate thresholding, good segmentation results can be achieved.
Downloads
Published
Issue
Section
License
Copyright (c) 2023 NUST Journal of Engineering Sciences
This work is licensed under a Creative Commons Attribution 4.0 International License.