World Journal of Engineering Research and Technology

Arjan Singh* and Dr. Anil Pal

ABSTRACT

Brain tumors are truthfully one of the scarier things out there and are still a major reason behind illness and death worldwide. That's why catching them early and knowing precisely where they are is super imperative on the off chance that specialists need to treat them right and give patients a much better shot. The traditional method of interpreting MRI scans is laborious and time-consuming. It is for the foremost portion manual and takes a though, and let’s confront it, presently and after that it reasonable isn’t that reliable or fast. To create the things a bit more astute (and less dull), we’ve put it together this framework that the inclines on AI and machine learning to do the overwhelming lifting when it comes to the detecting and the classifying brain tumors using the MRI scans. In addition, it is not just basic automation we are talking about – solid image pre-processing, feature picking magic, and models that know what to look for. The system is laid out to spot and sort out differing tumor sorts like Gliomas, Meningiomas, and those dubious Pituitary tumors. What really makes it tick is the way it grabs a bunch of features—stuff like texture, shape, and even deep-level data— using CNNs trained on MRI images. Then, for classifying the tumors, we tested out different models—like SVMs and Random Forest—and compared them to deep learning ones. Spoiler alert: deep learning rather crushed it, with better accuracy and more reliable results overall. The big takeaway? MRI scan data plays a huge role here, especially since it captures all those tiny spatial patterns and details that traditional methods might miss. And inside the conclusion, this whole setup may be a honest to goodness game-changer for doctors— speeding things up, diminishing the riddle, and giving way more exact comes almost without having to go full manual every time.

[Full Text Article] [Download Certificate]