The Application of Deep Learning to Quantify SAT/VAT in Human Abdominal Area

摘要

MRI imaging is less risky to humans than CT scans, but is more difficult to extract information from due to the ambiguous gray levels. The objective of the paper is to describe a novel technique that has been developed to automate the measurement of the abdominal subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT) from MRI images for obese patients before and after fasting. Ground truth has been established with clinicians' inspection. Three hundred thirty images have been utilized in this study where deep learning and Convolutional Neural Networks have been employed to quantify SAT and VAT. This work would minimize the time doctors spend analyzing MRI images.