Grey-Wolf-Based Wang's Demons for Retinal Image Registration

Area of Science:

• Medical Imaging
• Computer Vision
• Optimization Algorithms

Background:

• Image registration is crucial in medical imaging for accurate analysis.
• Retinal image registration presents unique challenges due to non-rigid deformations.

Purpose of the Study:

• To propose and evaluate a novel Grey Wolf Optimizer (GWO)-based non-rigid demons registration framework.
• To compare the GWO-based approach against other metaheuristic algorithms and existing demons methods for retinal image registration.

Main Methods:

• Implementation of a non-rigid demons registration algorithm optimized with the Grey Wolf Optimizer (GWO).
• Comparative analysis of GWO-based demons registration against Cuckoo Search, Firefly Algorithm, and Particle Swarm Optimization.
• Evaluation of GWO-optimized Wang's, Tang's, and Thirion's demons registration methods.

Main Results:

• The GWO-based framework achieved a high correlation of 0.9977 and demonstrated faster processing times.
• GWO-based Wang's demons exhibited superior accuracy and the lowest registration error (8.36 × 10^-5) compared to Tang's and Thirion's demons.
• The proposed GWO optimization significantly enhanced the performance of demons registration methods.

Conclusions:

• The Grey Wolf Optimizer provides an effective and efficient method for optimizing non-rigid demons registration in medical imaging.
• GWO-based Wang's demons registration is a highly accurate and robust technique for retinal image alignment.
• This framework offers a promising advancement for precise medical image analysis and diagnosis.