Enhanced brain tumor classification via efficient predefined time adaptive neural network optimized with high-level target navigation pigeon-inspired optimization

C. Pabitha

doi:10.1016/j.knosys.2025.114658

Enhanced brain tumor classification via efficient predefined time adaptive neural network optimized with high-level target navigation pigeon-inspired optimization

2025-10-25

0

PRE

AI

DOI：10.1016/j.knosys.2025.114658

Original

Original request for help

Save

Abstract

En 中文

Brain tumors are abnormal cell growths in or near the brain, affecting various brain tissues and structures. This paper proposes a Brain Tumor Classification utilizing Efficient Predefined Time Adaptive Neural Network optimized with High-level Target Navigation Pigeon Inspired Optimization (BTC-EPTANNHTNPIO). This novel framework utilizes publicly available brain tumor MRI datasets for training and analyzing. It incorporates Dual Image Adaptive Learnable Filters (DIALF) for preprocessing to enhance the MRI scan quality, Fast Continual Multi-view Clustering (FCMC) for accurate tumor segmentation, and Revised Tunable Q-Factor Wavelet Transform (RTQWT) for feature extraction to capture critical tumor characteristics. Classification is performed using EPTANN, optimized by HTNPIO to ensure precise and efficient brain tumor identification. The BTC-EPTANNHTNPIO model is evaluated using Accuracy, Precision, Sensitivity, F1-score, and Computational time metrics, showing significant improvements over existing methods with accuracy gains of up to 23.37 %, 23.35 %, and 21.45 %, along with similar improvements in precision and recall. The proposed method achieves a remarkable classification accuracy of 98.62 %, outperforming state-of-the-art techniques.

Enhanced brain tumor classification via efficient predefined time adaptive neural network optimized with high-level target navigation pigeon-inspired optimization

Abstract

Journal

Organization