A novel method for segmenting brain tumor using modified watershed algorithm in MRI image with FPGA.

Area of Science:

• Medical Imaging
• Neuroscience
• Computer Vision

Background:

• Accurate brain image segmentation is crucial for diagnosing neurological disorders like Alzheimer's and Parkinson's.
• Current Magnetic Resonance Imaging (MRI) segmentation methods face challenges due to image noise and artifacts, lacking a universal approach.
• Existing techniques often have limitations, necessitating improved segmentation strategies for reliable analysis.

Purpose of the Study:

• To develop an improved segmentation approach for Magnetic Resonance Imaging (MRI) to accurately delineate Regions of Interest (ROI).
• To address existing limitations in automated brain image segmentation techniques.
• To enhance the precision of brain structure identification for diagnostic purposes.

Main Methods:

• A three-phase approach was implemented: preprocessing, edge detection, and segmentation.
• Input MRI images underwent preprocessing using a high-pass filter for enhancement.
• The Enhanced Canny Edge Detection (ECED) and Modified Watershed Segmentation (MWS) algorithms were employed for segmentation.

Main Results:

• The proposed method successfully segmented the ROI from MRI images.
• Testing demonstrated that the developed system achieved superior results compared to existing strategies.
• The implementation was successfully realized on a Xilinx Virtex-5 FPGA.

Conclusions:

• The proposed segmentation approach offers a significant improvement for brain MRI analysis.
• The combination of ECED and MWS provides a robust method for ROI extraction.
• This technique holds promise for enhancing the diagnosis of brain-related conditions.