Using this method allows researchers to obtain highly quantitative measures of active brown adipose tissue, or BAT, in human subjects. Our methodology is designed to avoid common false positives and provide in-depth instruction on how to identify and classify brown adipose tissue depos. The main advantage of this technique is exclusion of false positives and the classification of metabolically active brown adipose tissue depos.
Begin this procedure with loading and prepping of the PET-CT images as described in the text protocol. After navigating through slices on all views, select edit at the top left of the toolbar and select brown fat, ROIs from the drop-down menu that appears. In the new dialogue box that appears, check use SUV and use CT check boxes.
Then, select one of the three vauxhall inclusion criteria. Select interior to apply the BAT detection algorithm to examine the vauxhalls inside the area of the ROI. Now, input an SUV lower limit normalized to the individual's measured or predicted lean body mass and an upper limit sufficient to accommodate high activity levels.
Input the BAT density range in the second row of free text fields. Check the check box located beneath volume times mean so that all vauxhalls deemed to be BAT will be highlighted in blue while the brown fat ROI window is open. Draw the ROIs by clicking the draw button in the brown fat ROI dialogue box.
Then, click anywhere within one of the three views to begin drawing an ROI. To compile ROIs and obtain the total BAT volume, set the starting and ending slice limit to the same slice so that the ROI will only apply to the current axial slice. Circle one depo of BAT without completing the ROI.
Continue the ROI by extending a connecting line across the body to the distant segment of BAT. Enclose the second BAT depo and double click on the previously identified point at the start of the second region. Adjust ROI points as necessary to further reduce the possibility of false positives.
Use unique anatomical landmarks, such as vertebral shape, other bony structures, and the presence of organs to identify the current anatomical region. To identify BAT in the cervical region, navigate to the axial view at the third cervical vertebra. Begin the ROI on the lateral side of the adipose tissue depo, avoiding neck muscles around the spinous process of the vertebra and creating a border just posterior to the lower edge of the mandible.
Now, identify BAT in the dorsocervical region by carefully including subcutaneous adipose tissue only where metabolic activity occurs. To identify BAT in the supraclavicular region, begin by drawing the ROI on the side most superficial, close to the highly active BAT region. For identification of BAT in the axillary region, select BAT near where the arm begins to separate from the torso, but avoids ribs and the lungs.
To identify bat at the mediastinal region, select BAT where the sternum begins to appear at the beginning of T2, near the anteriormost region of the individual's thoracic cavity and continue ROIs inferiorly until the end of the xyphoid process. Identify BAT in the paraspinal region by drawing ROIs around BATs surrounding the body, not the spinous process of the vertebra. Finally, identify BAT in the abdominal region inferior to T12 by tracing active fat directly surrounding the kidneys until metabolic activity is no longer present.
Generate a BAT mask in the brown fat ROI editor by selecting the mask tab and press, make masked PET. Close the PET-CT viewer, but leave the individual boxes open. Then, reopen a new PET-CT viewer window.
Select the following check boxes in the dialogue box that appears. The CT set and the latest PET set. Change the view of the PET-CT images to sagittal and start drawing all ROIs for region-wide analysis starting at the same sagittal slice.
Then, draw and label the cervical ROI by beginning at the top of C3 and extending the ROI to C7, drawing a line under the body of C7 before closing the ROI. Now, draw and label the supraclavicular ROI. Begin at C7, but do not include the body of the thoracic vertebrae while extending the ROI to T3.Then extend the left border of the ROI to the top of the manubrium of the sternum.
To draw and label the axillary ROI, begin at T3, but do not include the body of the thoracic vertebrae while extending the ROI to T7.Then, extend the left border of the ROI just short of the body of the sternum. Draw and label the mediastinal ROI by encompassing the entire sternum within a single ROI. Next, draw and label the paraspinal ROI, beginning at T1, including all thoracic vertebrae within the ROI.
Draw and label the abdominal ROI by beginning at the top of L1 and include any BAT that was not accounted for in any of the previous regions within the abdominal ROI. Finally, draw and label the dorsocervical ROI. Include the region of dorsal subcutaneous fat near the cervical and top of the paraspinal region.
This is where the subject's body has made contact with the scanning bed. Check show all to display the ROIs of all regions. Now, line up all ROIs in order to prevent overlapping or underestimation.
Position the perimeter of adjacent ROIs flush with each other so that no BAT is included in two regions and that no BAT is missing from all regions. To avoid false positives in BAT quantification, PET, CT, and anatomical information must all be considered when drawing ROIs. Several common regions to include and avoid when quantifying whole-body BAT in cold-stimulated subjects are shown, such as metabolically active cervical BAT versus salivary glands, vocal cords, and thyroid.
Also, supraclavicular BAT should be included, while shivering muscle near borders of air and solid tissue should be avoided. Finally, when including abdominal BAT, avoid the calyces of the kidneys. After the ROIs of each axial slice are compiled, BAT deposits can be segmented in the sagittal plane to examine intra, inter, individual differences in regional BAT activation.
The regions shown here include cervical, supraclavicular, axillary, mediastinal, paraspinal, abdominal, and dorsocervical. The composite image with all regions is also shown. It is important to use identifiable organs and structures within the image frame to understand your location in the body and to determine whether a structure is truly brown fat.
Ultimately, we hope to develop an automated method using deep learning or AI methodology. Creating a detailed atlas of brown fat is essential to ensure that the automated method is identifying BAT. Using this image processing method, we've developed a quantitative and noninvasive map of human brown adipose tissue.
With this method, researchers can make more comprehensive comparisons of funcitonal BAT in humans.
