Brown adipose tissue (BAT) is the subject of ongoing research on obesity and metabolic diseases. It has the ability to dissipate energy through non-shivering thermogenesis¹. MRI is a non-invasive, ionizing-radiation-free imaging method for BAT detection. BAT amount and activity were quantified in in vitro^2,3 and in vivo^3,4 studies. Fat-fraction (FF) assessments showed that BAT FF decreases over time under cold exposure. Recently, BAT FF decrease was evaluated over time during induced cooling of the skin⁵. In this work, the study was repeated after winter for the same interscapular BAT depots and the results were compared to the previously published results⁵ to evaluate reproducibility.The protocol of the recent BAT study5 (St01) was repeated after winter (St02). Four of five subjects could be included (S01-S03, S05; 22-29 years, body-mass-index (BMI) 19-24). MR examinations were performed with informed consent on a 3T hybrid MR system (Biograph mMR 3.0T, Siemens, Erlangen, Germany), using the spine-array coil, the flexible 3×3-body-matrix-array coil and the 16-channel head/neck coil. A 2-point Dixon⁶ imaging sequence (VIBE;TR=5.85ms;TE_ⁱⁿ/TE_opp=2.46/3.69ms;α=10°;matrix=416×260×64;FOV=500×312×76.8mm³;BW=710 Hz/px;GRAPPA R=2;30sec breath-hold) was applied every 5min over 140min. For activation of the clavicular region, a cooling vest (Polar Products Inc., OH, USA), circulated with temperature-controlled water from a water circulator (Haake F6 Circulator, Artisan Scientific, IL, USA) was used to expose each volunteer to a cold environment. Three temperature phases were employed: baseline (23°C, 20min), cooling (12°C, 90min), and warming (37°C, 30min). Water temperatures at the inlet and outlet of the vest, under the armpit and at the back skin were monitored with fiber-optic temperature probes (LumaSense Technologies Inc., CA, USA). Data evaluation was done with MATLAB (The MathWorks, Natick, AM, USA) and MITK (Medical Imaging Interaction Toolkit, DKFZ, Heidelberg, Germany)⁷. Water images (W) were registered to a target image to compensate different breath-hold positions. Resulting transformations were applied to the corresponding fat images (F). FF maps (FF=F/F+W) were determined for each time point and median filtered (3-by-3-by-1-neighborhood) to reduce noise. FF change rate during activation was estimated applying a linear fit pixel-wise. The following criteria were used for BAT assessment: a)FF before cooling: 0.6 ≤FF_base≤ 0.8, b)FF at end of cooling: FF_cool≥ 0.3, c)negative FF gradient, d)size of BAT region-of-interest (ROI) N≥15px, e)segmentation of manual interscapular target areas (R1,R2) assisted by radiologist. R1 and R2 were compared to subcutaneous adipose (SAT) and muscle tissue (mean ROI size 8ml). For comparison of changes in R1 and R2 between St01 and St02, manual rigid registration was performed to compensate for different patient positions on the table under the assistance of a radiologist. Evaluated regions were restricted to the slab overlap of St01 and St02.Fig.1 shows a FF map of S01 with marked positions of evaluated regions. BAT ROI positions and FF changes were reproducible between studies for each volunteer (Fig.2,Tab.1). During cold activation, a mean FF decrease of (-2.31±1.05)%/h was found (Tab.1). Mean FF_base values (70±2)% decrease to lower mean FF_cool values (66±2)% (Fig.3). S02 and S05 showed a slight FF decrease during the baseline phase in St02 (similar to St01). During the warming phase, no immediate recovery of the initial BAT FF was found. An increase of BAT volume was detected in St02 in S01 and S05 and a decrease in S02 and S03 (Fig.4). In SAT (mean FF_SAT=(79±4)%) and muscle (mean FF_muscle=(7±2)%) no FF change was observed (Fig.3). Measured body temperature stayed constant at (37.6±0.13)°C and back skin temperatures decreased during cold activation. The mean outlet temperature at the vest was about (1.2±0.2)°C higher than the inlet temperature, indicating a mean heat transfer of (61±8)W into the water circuit.The results of St02 reproduce the results found in St01. BAT depots showed a similar FF decrease during cold activation. In subject S02, R1 was not detectable and was not evaluated. However, cold-activated potential BAT depots were found in each subject. We observed an increase of potential BAT depot size in two volunteers (S01,S05). Both volunteers also transferred more heat into the water circuit in St02 (14%,8%). The other volunteers (S02,S03) with decreased potential BAT depot size released less heat (-18%,-16%). Previous studies found an increased BAT depot size during winter^8,9. However, a larger cohort needs to be evaluated to analyze correlations. In the next step, we will optimize our algorithm for detection of supraclavicular BAT depots and increase study cohort size to analyze correlations of BAT mass and other parameters (e.g.,outdoor temperatures, age, sex, BMI). In conclusion, this study (St02) showed reproducible FF decreases during cold activation in potential interscapular BAT depots at reproducible anatomic positions.