Performance investigations of filtering methods for T1 and T2 weighted infant brain MR images

In recent decades, medical image analysis and diagnostic techniques have undergone significant advancements and have become a relatively important component of clinical practice. The most popular diagnostic resources are diagnostic images acquired from different modalities such as computed tomography and magnetic resonance imaging. Neonatal neuroimaging is an increasingly developing diagnostic imaging discipline with a particular focus on neonatal brain imaging. The neonatal brain growth and numerous neurological defects can be detected by the newborn brain MRI. MRI images consist primarily of objects of low contrast that are hampered in the image capturing by random noise. Noise produces ambiguous representations which influence disease identification and diagnosis, even mortality, leading to severe loses. Medical image de-noising mainly attempts to reconstruct the original image from its noisy observation as accurately as possible while maintaining the necessary graphical features such as textures and edges. It is also necessary that the medical images that assist healthcare practitioners towards precise disease analysis must be de-noised. This paper provides systematic analysis of de-noising methods for neonatal brain MR images in which each technique has its own conclusions, drawbacks and benefits. This work investigates performance as well as thorough study of different image de-noising approaches for T1 and T2-weighted neonatal Brain MR Images. Utilizing different statistical parameters such as PSNR, SSIM, MSE etc. the image de-noising approaches are compared.