Classification of White and Brown Adipose Tissue using a Support Vector Machine

Determining the volume and distribution of white adipose tissue (WAT) and brown adipose tissue (BAT) by magnetic resonance imaging (MRI) is clinically important. Previous WAT and BAT classification has relied on using fat fraction and proton relaxation time via fixed multi-peak spectroscopic models. However, the recently proposed Multi-Varying-Peak MR Spectroscopy (MVP-MRS) model allows for the selection of appropriate classification features for differentiation between WAT and BAT. Furthermore, these multi-peak features allow prediction of a ‘browning’ or ‘beigeing’ process of WAT by using a Support Vector Machine (SVM) learning algorithm.