Brain Tumor-Level Detection in 3D MRI Images

The field of neuroradiology has advanced significantly with the implementation of 3D Magnetic Resonance Imaging (MRI) technology. This imaging modality enables early-stage tumor diagnosis by detecting minute tumors and subtle abnormalities with high precision. Brain tumors are classified into three primary categories: benign, premalignant, and malignant. A brain tumor is defined as an abnormal proliferation of cells and tissues, with malignant tumors being characterized by rapid growth, irregular surfaces, lack of encapsulation, and deep attachments to surrounding structures. These tumors can exert substantial pressure on adjacent tissues and spread rapidly. The proposed method leverages 3D MRI scans for precise brain tumor detection. Users can choose between two advanced classification approaches: the Support Vector Machine (SVM) or Adaptive Neuro-Fuzzy Inference System (ANFIS) classifiers. The preprocessing phase, which includes contrast enhancement, noise reduction, and grayscale conversion, optimizes image quality. Segmentation techniques are then applied to isolate the tumor region, followed by feature extraction to identify vector-based characteristics. These features are input into the classifier to accurately determine the tumor type. Tumor-affected MRI images are visualized through various graphs generated by a Self-Organizing Map (SOM), enhancing interpretability and aiding in diagnostic decision-making.Keywords: 3D brain Model, Detection, Tumour Stages, 3D MRI, segmentation.