Deep Learning and Optimization Approaches in Multi-classification of Brain Tumor MRI Images Using Deep Dynamic Parallel Convolutional Neural Network with Fully Termite Alate Optimization Algorithm: A Review

Abstract

Brain tumours are among the most life-threatening neurological disorders, requiring accurate and early diagnosis for effective treatment planning. Magnetic Resonance Imaging (MRI) plays a crucial role in brain tumour detection due to its high-resolution and non-invasive nature. However, manual interpretation of MRI scans is time-consuming and prone to human error. Recent advancements in deep learning and optimization algorithms have enabled automated and highly accurate multi-class classification of brain tumours. Convolutional Neural Networks (CNNs) have emerged as the dominant approach for MRI-based tumour classification due to their ability to extract hierarchical features. Advanced architectures such as parallel CNNs, multiscale CNNs, and hybrid models have demonstrated superior performance in distinguishing tumour types such as glioma, meningioma, and pituitary tutors. Additionally, optimization algorithms, including Particle Swarm Optimization (PSO), genetic algorithms, and metaheuristic techniques, have been employed to fine-tune model parameters and improve classification accuracy. Recent research focuses on integrating deep dynamic parallel CNN architectures with bio-inspired optimization techniques, such as termite alate optimization, to enhance feature selection, convergence speed, and classification accuracy. This review analyses recent studies (2020–2023), compares different deep learning and optimization approaches, and identifies challenges such as computational complexity and data imbalance. The study highlights future directions for developing efficient and scalable AI-based diagnostic systems.