What to read on a DVH before changing anything
Three questions, in order, that tell you which knob to turn — and stop you turning all of them.
Built in·updated 2026-08-03
1. Where does the target curve fall off? A shoulder that drops early means cold spots inside the PTV; a long tail to the right means a hotspot. They call for opposite fixes, and the total score cannot tell them apart.
2. Which OAR curve is closest to its limit, in points rather than in Gy? A structure 1 Gy under a limit worth 70 points matters more than one 5 Gy under a limit worth 5. Score weights, not dose, decide what to work on.
3. Is any curve shaped nothing like the others? A structure whose curve sits flat and high is usually a matching problem — the wrong ROI, or one the plan never saw — not a planning problem. Fix that before optimizing anything.
Answer all three before you touch an objective. Most re-optimizations are spent undoing a change made for the first reason when the problem was the third.
How to do it in
- Objective list
- achieved vs requested, read first
- DVH
- for unconstrained structures
- Low-dose end
- check on helical deliveries
Read the objective list before the curve. Precision shows the achieved value against each objective you set, which answers "did I get what I asked for" without any interpretation.
The DVH is for the question the objective list cannot answer: what happened in the structures you did not constrain. That is where a plan is usually quietly wrong.
On a helical delivery, check the low-dose bath as well as the high-dose end. A rotational technique spreads a few Gy over a lot of tissue, and a criterion written for a C-arm plan — a mean dose, a V5 — can be met by a fixed-field plan and missed by a helical one that is better everywhere else.