A hybrid PET-MR scanner was used to compare proposed MR techniques to the 8F-FDG PET gold standard for brown adipose tissue detection. The subjects were cooled inside of the magnet while PET and MR fat fraction and BOLD techniques were obtained. It was found that MR techniques alone are not sufficient to accurately predict BAT volume but can be used to increase along side PET to increase BAT specificity. Brown Adipose Tissue (BAT) helps maintain core body temperature by burning fat and producing heat in a process called “non-shivering thermogenesis”. The discovery that BAT is far more active in adult humans than previously thought lead to a lot of interest in BAT with regard to metabolic disorders such as obesity, diabetes and cachexia.2-4 Strong links between BAT and both diabetes and obesity have been recently shown in mice but similar links in human have been more difficult in humans to research due in large part to difficulty imaging adult human BAT.5,6 18F-FDG PET-CT imaging, the gold standard, presents issues with low specificity for BAT and high radiation dose for longitudinal studies. MR methods of detection have been proposed and tested but results have been varied and it has been difficult to make a direct comparison when scanning on separate modalities. 7,8 By using a hybrid PET-MR scanner we aimed to evaluate the accuracy of MR BAT detection methods relative to the gold standard of 18F-FDG PET within the same scan.Seven participants were cooled according to an individualized cooling protocol inside of a hybrid PET-MRI scanner (Biograph mMR, Siemens Healthcare, Germany) using a modified ArcticSun 2000 temperature management system (Medivance, Louisville, CO). A two-echo chemical-shift dixons, gradient recalled echos and a static FDG-PET on the upper thorax were obtained. Fat fractional maps were created from the dixon water and fat phase images. Regions of interest (ROIs) were defined posteriorly using the fat fraction maps and static PET images using a range of acceptable fat fraction and a minimum FDG uptake for BAT. The accuracy of the MR and FDG methods were compared and the ROIs were used to fat fractional decreases and changes is GRE signal in BAT during cooling.Five of the seven volunteers were positive for brown fat showing a standard uptake value of greater then 1.5 SUV despite all seven containing tissue between 20 and 80 percent in the supraclavcular region. In the five subjects positive for BAT MR fat fraction techniques were found to overestimated BAT volume by 70±30% when compared to BAT volume confined by PET uptake and fat fraction. Three of the five subjects positive for BAT showed significant decreases in BAT fat fraction (P<0.05) and in one of the two subjects negative for PET a significant decrease (P<0.05) in fat fraction was observed in supraclavicular fatty tissue. In four of the five subjects there was no significant decrease in fat fraction in the area PET negative but within the fat fraction range. Significant changes in the MR signal intensity from BAT were not observed in any subject. The study showed that by themselves MR techniques are not able to accurately define BAT volumes. Unlike in mice, human BAT is highly heterogenous and due to the partial volume effect fat fractions greatly overestimate BAT volume. GRE signal differences are highly susceptible motion and the proximity of BAT to the thoracic cavity as well as the heterogeneity of BAT lead to difficulty tracking changes in signal intensity. Because the primary substrate for thermogenesis is endogenously derived fatty acids looking for changes in fat fraction can serve as a direct marker for thermogenesis. Glucose however is not a direct biomarker which helps explain the surprising lack of uptake in two young, very fit subjects. In one we observed very little BAT according to PET and no BAT according to PET despite showing a significant decreases in fat fraction in their supraclavicular fatty tissue and being in the population in which we expect the highest BAT activity. It is possible that this is caused a change in glucose metabolism in the BAT of extremely fit young subjects. FDG-PET is known to have difficult identifying BAT in obese subjects but our results suggest that by looking for decreases in fat fraction MR techniques may be able to increase sensitiity to BAT in some young fit adults as well. Fat fraction techniques alone are not an accurate prediction of BAT volume. Fat fraction decreases can be used to more directly observe BAT thermogenesis and in conjunction with FDG-PET measurements to increase BAT specificity.