Nestle

The brown fat ROI dialog has 2 ways of making ROIs, either manually or with autosegmentation. The output of these ROIs can be filtered via fixed SUV values or percent of maximum SUV. Another method has been added from the paper by Nestle. This method uses ROIs created on the radiation planning station which are isocontours which contain the tumors to be radiated. They discuss how well the contours fit the geometry of the tumor. In our case we don't generate isocontours, but rather use the ROIs generated by our software. Beyond this difference we try to follow their method of processing the data.

See Demo to get you started.

You see 0.3n. The n means "use Nestle", and the 0.3 is the usual beta value, which you can change if you like. There is also the possibility to use a fixed SUV like 2.5 or a percentage like 41%. (The upper limit of 20 is used in the fixed SUV case but is ignored in both percentage and nestle.)

The beta in this example is 0.3. The mean intensity is average of points inside the ROI which are >= 70% of the SUVmax. The background is manually chosen in the liver or mediastinum, and is the same for all ROIs. The Nestle is similar to the percentage, except the percentage uses a single point for the SUVmax, whereas Nestle uses a mean of SUV values and adds a background.

The background is reported as specified in Options. To see on the screen to same number as appears in the Background textbox, choose SUV mean. There is no need for this, as the SUV mean will appear in the text box even if SUV max appears on the screen. On the other hand the diameter, and whether a simple circle or a sphere is chosen, will have an effect. (Use show circle for visual feedback on what exactly you are measuring.)