USING OPTIMIZED FEATURE SELECTION FOR CLASSIFICATION OF BRAIN MRI

Several studies into the detection of brain anomalies have been carried out because of its considerably significant role in the identification of anatomical areas of interest for diagnosing diseases, treating illnesses or even the planning of surgeries. Alzheimer's disease (AD) is the most typical kind of dementia amongst elderly people across the world. Magnetic resonance imaging (MRI) is a method which yields images of excellent quality of the anatomical features of the human body, particularly in the brain and offers clinical data supporting diagnoses as well as for biomedical research. The current study is a features selection as well as classification study for normal brain subjects. During the features selection phase, features are chosen through Bacterial Foraging Optimization (BFO). Experimental evaluation was carried out for several other features selection techniques such as Information Gain (IG), Minimum Redundancy Maximum Relevance (mRMR) as well as classifiers such as Instance-based Learning (IBL), C4.5 as well as Fuzzy Classifiers.