Is Your Voluming Accurate?
Name Answer
Santam Have not assessed this systematically yet. Is there any way the house can help in this? I do know that 0 -1 mm accuracy is unachievable with human hand.
Indranil It varies according to the structure being contoured. Sometimes I am sure I am 4-5mm away.
Andrew I am going to be very confronting and controversial here - how can you be inaccurate by 4-5mm? Can you post a picture of something this uncertain? Actually drawing accuracy can be achieved at a resolution lower than the scan appearances (you can draw a line to include or exclude 1 pixel.
Pranshu Well sure after zooming in, a line could be drawn or include or exclude 1 pixel.. but at that level of zoom, the quality of image is usually not clear enough to be able to draw what needs to be drawn. Also because there always may not be an air-tissue interface to provide that sharp contrast, the graded density fade off will make it even more difficult to contour at that high zoom. So though precise, it is an issue of accuracy here. An example where once could be 4-5mm away is LN enlargement with ECE especially near the blood vessels. There is always a gray zone around the LN mass in these scenarios which seems to merge into the tissue density around the blood vessels and it is absolutely impossible to determine that extent. Hence, we land up over contouring or be generous in our CTVs. Even something as simple as parotid gland can be difficult to be contoured especially around deep lobe as digastric belly just slides past it.. another area where people may over or wrongly undercontour to project better parotid doses. In all this there is one major thing that we may miss.. hand may be accurate when we draw.. but using a mouse is a different ball game.. i know there are these high resolution tablet screens available.. but how many of us do really use it.. so there is really no point to be precise contour to one pixel.. i think an accuracy level of 2-3 mm is what we should be aiming for consistently.
Andrew 1. the issue of "I am not sure where to draw the line", and "this ECE makes the fat all hazy" are not really what I was talking about. In an uninvolved neck, drawing the fat between SCM and neck muscles can be done with an accuracy down to the limits of vision and hand steadiness, in fact I can autocontour the area based on density difference between fat and muscle and the accuracy is <1mm.
Andrew 2. "around deep lobe as digastric belly just slides past it" - when fiddling with the contrast setting on my non-contrast H&N scans, the digastric muscle and parotid are distinctly different! Not a good example. Perhaps we should put more specific questions rather than dwelling in realms of personal opinion. For example, this parotid/digastric thing, is it really easy (as I say) or not (as Pranshu says)? Does the CT settings make that much of a difference? Is it the CT scanner quality? We need to answer these things with something more than ping-pong arguments.
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License