867 / 1161
What have we learned after >10 years searching
the holy grail for PET based GTV-segmentation?
• Using molecular imaging for GTV delineation
at all
is more important
than finding the right method to include the last voxel
• Maybe drawing one line is not what resembles the information
needed for future RT planning
• If we need one line,
visual
delineation is not a bad idea, institutional
standardisation makes sense
• Automatic delineation (by something else than simple thresholding)
speeds up
the contouring process but should be used as a starting
point for user review
• The use
4D
imaging for TVD will not be possible without automation
14