Brain tumor classification utilizing deep features derived from high-quality regions in MRI images

•Addresses the need for rapid and accurate brain tumor detection to improve patient outcomes.•Introduces a less labor-intensive approach to diagnose brain cancers using MRI imaging.•Trains and evaluates the system using a publicly accessible brain tumor dataset, achieving an overall accuracy of approximately 98.98%.•Enhances resilience, automation, and accuracy of the diagnostic process, offering significant benefits to the healthcare industry.