Myelin Quantification From Magnetic Resonance Images

Abstract

The goal of the investigation is to segment and quantify myelin from Magnetic resonance images (MRI). Myelin is the most important component for the appropriate brain function newlineof children as well as adults. Myelin is very important for human brains because it can affect overall intelligence, coordination, ability to solve problems, reading capacity, memory, eaction time, cognitive skills, motor control, and development of the brain, attention span, learning capacity, and overall mood. So quantification of myelin from MRI is essential in medical analysis, which is a very tough task nowadays. Medical experts segment myelin manually which may lead to many practical errors because of the subjective nature of the brain. Segmentation of myelin is very important to quantify myelin. If we identify myelin growth is not sufficient in the early stage then we can proceed with further treatments to improve myelin. Even though much researches are going on in myelin, myelin quantification directly from conventional MRI is not successful till now. It will be a very useful achievement in the medical field if we can quantify myelin directly from conventional MRI, because when compared to other methods conventional MRI is time-effective and cost-effective. T1-weighted image is a very regular type of MRI in medical use. So if we compute myelin volume from T1-weighted MRI it will be extremely beneficial for clinical purposes. Skull stripping is the most essential stage in myelin segmentation. For this, We employ the morphological operations closing and opening. Pre-processing is a very important process before segmentation in medical image processing. In this work, T1 MRI images of the brain are pre-processed using SPM-12 neuro-imaging software. SPM-12 follows spatial pre-processing. Realign, coregister, normalization and smoothing are the four important steps of pre-processing. Segmentation and quantification of myelin from clinical data is not easy as processed data set. Curvelet transform based adaptive Gaussian notch filtering wi

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced