Abstract: Brain tumors pose a serious threat to global health. For treatment to be effective, accurate and timely detection is required. Early diagnosis and efficient treatment of brain tumors are difficult tasks for the medical community. Because of its remarkable spatial resolution, magnetic resonance imaging (MRI) stands out as a non-invasive method of identifying brain malignancies. This research describes a novel approach to automatically identify brain tumors using MRI pictures. First, pre-processing techniques are used to enhance image quality and use a gaussian filter to reduce noise. Then, utilizing LBP and PSO algorithms, pre-processed pictures are employed for feature extraction and feature optimization. In this case, the brain tumor categorization is done using the K-Nearest Neighbor technique. Accuracy can be increased throughout the entire process, yielding accurate results along with curability, sensitivity, and specificity. The created system has the potential to be used in clinical settings and provides an automated and dependable method for MRI image-based brain tumor detection.

Keywords: Brain tumors, MRI Images, Local Binary Pattern, Particle Swarm Optimization.

| DOI: 10.17148/IARJSET.2024.11467