Winston Lutz unable to detect BB size Varian sugar cube

Google Groups
1

I’m having an issue analyzing a WL using Varian’s sugar cube. I’m assuming the BB size is too small (~2 mm). I think the issue is in def _find_bb, but I’m not entirely sure what to change for the code. I was wondering if it’s possible to run the analysis on BBs under 2 mm. I attached a zip file of the images.

File “”, line 1, in
File “C:\Users\htran\anaconda3\lib\site-packages\pylinac\winston_lutz.py”, line 205, in analyze
img.analyze(bb_size_mm)
File “C:\Users\htran\anaconda3\lib\site-packages\pylinac\winston_lutz.py”, line 834, in analyze
self.bb = self._find_bb(bb_size_mm)
File “C:\Users\htran\anaconda3\lib\site-packages\pylinac\winston_lutz.py”, line 933, in _find_bb
raise ValueError(
ValueError: Unable to locate the BB. Make sure the field edges do not obscure the BB and that there is no artifacts in the images.

WL images.zip (29.9 KB)

2

Hi,

Since Version 3.0.0 you can set the bb size in the analyze method:
https://pylinac.readthedocs.io/en/latest/changelog.html?highlight=py_gui#id31

regards
Thomas

3

I tried using wl.analyze(bb_size_mm) but it still seems to not recognize the bb.

4

Hi,

What bb sizes did you try? If you have a 2mm BB did you also try a.e. 3mm BB size? .analyze(bb_size_mm=3)

regards
Thomas

5

I’ve not seen 8-bit images in quite some time, much less a truebeam; I didn’t know that was possible! Not only is the bb quite small, the image appears quite noisy. I’m not sure if this is from a low number of MUs, but the bb, even if the size wasn’t an issue, is difficult to find in this image. Here’s a representative image converted to binary (the step to find the bb). I can’t make out the BB and neither can the algorithm. I’d try again w/ more MU and/or bigger field if possible.

noisy_wl.png