Abstract
Magnetic resonance imaging (MRI) is one of the common and widely used imaging techniques in medical imaging with excellent capability for soft tissues imaging. It is very suitable for brain imaging, muscles and excellent for early diagnosis and detection of brain tumors, treatment monitoring and other brain abnormalities. However, the incorporated noise during MRI image acquisition process makes it difficult for human interpretation as well as computer-aided analysis of the images. The main objective of this paper is to present an effective method for noise removal from brain MRI images which reduce the effect of noise while retaining the structure of the image. In this paper, a hybrid technique has been proposed which ensembles the median and wiener filters through intelligent Machine Learning technique, the Genetic Programming (GP). The GP intelligently correlates and combines the features of the filters with some trigonometric and other functions like sin, cos, and log, mathematical operators and constants with different combinations in an automated manner with the help of fitness function. At the end, GP evolved best mathematical optimal expression which is used to restore the images corrupted with impulse noise. In order to evaluate and show the effectiveness of proposed method, a set of comprehensive experimentation on standard dataset have been performed and validated the performance. Results show that, the performance of the proposed hybrid technique is much better and removes the noise from images both at low as well as at increased noise levels. The results of the proposed method have also been compared with state-of-the-art methods and the performance of the proposed method is also found to be better than previous techniques.