Automated brain tumour segmentation from multi-modality magnetic resonance imaging data based on new particle swarm...

Area of Science:

• Medical Image Processing
• Artificial Intelligence
• Computational Neuroscience

Background:

• Brain tumour segmentation is a critical yet challenging task in medical imaging.
• Manual segmentation is time-consuming and prone to errors, necessitating automated solutions.
• Computer-aided detection systems aim to reduce physician workload and improve segmentation accuracy.

Purpose of the Study:

• To develop an artificial intelligence-based level set method for automated brain tumour segmentation.
• To introduce novel segmentation algorithms based on Particle Swarm Optimisation variants.
• To evaluate the performance of these algorithms in segmenting 3D brain tumour images.

Main Methods:

• A level set method integrated with artificial intelligence was proposed.
• Three algorithms were developed: Modified Particle Swarm Optimisation (MPSO), Darwin Particle Swarm Optimisation (DPSO), and Fractional Order Darwinian Particle Swarm Optimisation (FODPSO).
• These algorithms were applied to segment 3D brain tumour images.

Main Results:

• The proposed segmentation technique was validated using the MICCAI RASTS 2013 database for high-grade glioma patients.
• Performance was assessed using accuracy, sensitivity, specificity, and Dice similarity coefficient.
• The MPSO algorithm demonstrated superior performance in segmentation accuracy and robustness.

Conclusions:

• The Modified Particle Swarm Optimisation (MPSO) algorithm consistently outperformed DPSO and FODPSO in brain tumour segmentation.
• The developed AI-based approach shows promise for accurate and efficient 3D brain tumour segmentation.