The investigation study on non-linear filter based preprocessing for MRI image segmentation and classification

Imaging techniques assists the medical practitioners and researchers to diagnose the activities and disorders in human body before persistent surgeries. Among several medical imaging modalities, magnetic resonance imaging provides additional contrast information about the tissues. Magnetic resonance imaging scans are used as significant method for identifying diseases throughout the human body. MRI scan provides sufficient information for patient diagnosis. Three processes namely preprocessing, segmentation and classification are performed on MRI images for finding the existence of disease. Many researchers introduced segmentation and classification techniques for improving the performance of tumor identification with MRI images. But, PSNR, segmentation time and classification accuracy performance of existing techniques was not improved. In order to address these problems, machine learning and deep learning based non-linear teager filter can be used in our research work.