Özet:

Abstract The formation of a brain tumour involves a collection of tissue to which abnormal cells are gradually added. The most difficult duty is to categorise brain tumours using magnetic resonance imaging (MRI) so that affected people can receive therapy. Typically, human investigators examine brain MRI scans for tumour identification and categorization. According to the classification, for which various methodologies are created, images are interpreted. Brain tumour MRI segmentation utilising the suggested Reg FConVolNN segmentation method yields information about anatomical patterns and aberrant tissues. Here, a dataset of patients who had previously experienced brain tumour symptoms was combined with their historical medical information. The suggested neural technique can analyse MRI pictures to find cell mutations and pre-process input images to remove components like the vertebral column or skull in preparation. The effectiveness of the strategy suggested utilising a dataset of MRI images is contrasted with that of existing deep learning and machine learning models. The results show that the suggested method outperformed methods that employed the same dataset in terms of accuracy, AUC, precision, recall, and F-1 score for tumour classification.

Anahtar Kelimeler: