Title: Optimised brain tumour detection with the Pteropus Unicinctus optimisation reliant deep model

Authors: Sumit Chhabra; Khushboo Bansal

Addresses: Desh Bhagat University, Mandi Gobindgarh, Punjab, India; Department of Computer Science, Khalsa College for Women, Amritsar, India ' Department of Computer Science and Engineering, Desh Bhagat University, Mandi Gobindgarh, Punjab, India

Abstract: This research Pteropus Unicinctus optimisation (PUO) is proposed which is mathematically derived from the echolocation character of Pteropusand and the eye power searching ability of Unicinctus. The integration of convolutional neural networks (CNNs) and long short-term memory (LSTM) in the PUO-reliant deep model presents an optimal solution to enhance the precision and dependability of brain tumour detection. This strategy leverages the unique capabilities of both architectures and fine-tunes their performance through hyperparameter optimisation, which leads to better diagnostic results and improved quality of patient care in brain tumour detection. The proposed approach attained the accuracy of 99.51%, 99.58%, and 99.50%, and considering sensitivity 99.48%, 99.60%, 99.37% attained, and the model attained 99.48%, 99.58%, and 99.26% of specificity for BraTS 2019 dataset, SimBRATS 2019 dataset, and brain tumour dataset respectively.

Keywords: brain tumour detection; AlexNet; feature extraction; optimisation; convolutional neural network; CNN; Pteropus Unicinctus optimisation; PUO; long short-term memory; LSTM.

DOI: 10.1504/IJIIDS.2025.145466

International Journal of Intelligent Information and Database Systems, 2025 Vol.17 No.2, pp.186 - 216

Received: 28 Oct 2023
Accepted: 09 Jul 2024
Published online: 01 Apr 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article