An Optimization Numerical Spiking Neural Membrane System with Adaptive Multi-Mutation Operators for Brain Tumor...

Area of Science:

• Medical Imaging
• Computational Intelligence
• Artificial Intelligence

Background:

• Magnetic Resonance Imaging (MRI) is crucial for diagnosing brain tumors, enabling non-invasive imaging.
• Accurate segmentation of brain tumors from MRI is essential for diagnosis and treatment planning.
• Existing segmentation methods face challenges with image artifacts and complexity.

Purpose of the Study:

• To propose a novel threshold segmentation approach for brain tumor images using optimization spiking neural P systems.
• To introduce an optimization numerical spiking neural P system with adaptive multi-mutation operators (ONSNPSamo) for enhanced segmentation.
• To combine ONSNPSamo with connectivity algorithms for improved brain tumor segmentation accuracy.

Main Methods:

• Development of an optimization numerical spiking neural P system with adaptive multi-mutation operators (ONSNPSamo).
• Implementation of a multi-mutation strategy within ONSNPSamo to balance exploration and exploitation.
• Integration of the ONSNPSamo with connectivity algorithms for brain tumor segmentation.

Main Results:

• ONSNPSamo demonstrated superior or comparable performance against 12 other optimization algorithms on CEC 2017 benchmarks.
• The combined ONSNPSamo and connectivity algorithm approach showed enhanced effectiveness in segmenting brain tumor images in BraTS 2019 case studies.
• The proposed method achieved more accurate brain tumor segmentation compared to most involved algorithms.

Conclusions:

• The ONSNPSamo is a promising optimization technique for image segmentation tasks.
• The integration of ONSNPSamo with connectivity algorithms offers a robust solution for brain tumor segmentation from MRI.
• This approach advances the field of medical image analysis for neuro-oncology.