Are all plan metric objectives achievable?

This article discusses the achievability of plan study metric objectives. For a list of plan study metrics with formulas, please see our Plan Study Metric Library article.

The Question

We are often asked, "Are all of the minimally required and ideal objectives in plan studies actually achievable?" Great question!

Short Answers

To answer the first part: YES, all of the "minimally required" objectives are achievable.

To answer the second part: NO, not all of the "ideal" objectives are achievable, and this is by design.

Explanation

Treatment planning always involves trade offs between OAR sparing and target coverage. Do I give up some coverage to spare an OAR? Do I sacrifice an ideal OAR result to get the required target coverage?

ProKnow plan studies emulate the real world challenges by giving you empirical score functions for each plan metric where you, as the planner, will use the collection of score functions as your guide on where to make sacrifices in order to achieve your optimal score over all the metrics.

The "minimally required" objective for each metric is always achievable; however, in some cases the "ideal" score for a specific plan metric is not achievable. This allows a planner to get as close to the ideal as possible while also balancing the trade-offs in terms of how pushing the limits might adversely affect scores for other metrics.

One example where the "ideal" score for a metric may not be possible is if the top score for a Conformation Number (CN) metric is given for a result of CN = 1.0. A perfect score for a CN is 1.0, but this is also physically impossible. However, for a ProKnow plan study algorithm, the max score for a CN metric might be set to 1.0 to serve as an asymptotic limit so that planners may work towards higher conformation, but back off when there are diminishing returns in terms of overall plan score. Another example would be for OAR metrics such as dose-to-volume, where an ideal might be set very low (or in some cases, to 0.0 Gy). This is based on an "ALARA" mindset, even though 0.0 Gy to a specified % volume or cc may be impossible. It will be up to the planner's skills to figure out how to optimize the total plan score by knowing where to push the limits and where to settle for less-than full points.

NOTE: ProKnow plan studies are designed so that perfect scores (e.g. 150 out of 150) are impossible. This also is by design.