Medical physicist striving for better radiation treatment planning

When it comes to planning radiation treatment for cancer patients, ‘good enough’ just doesn’t cut it for Ed Bender, a medical physicist at Mercy Regional Cancer Center.

That’s why he created and secured a patent for a new method of automatically generating treatment plans based on patient data. This came after developing an early version of the technique about six years ago for his own use.

“My fundamental goal in developing this automated technique was not to try to reproduce something that I’ve done over and over again — I wanted to push the limits of what is even possible,” said Bender, a former assistant professor in the UW School of Medicine and Public Health.

He says he often comes across treatment plans that technically meet objectives set by national guidelines, but could be better.

Plans for radiation treatment include millions of adjustable parameters like radiation intensity, alignment and beam modulation, so it’s not something a human could personally develop, Bender explains. His method of automation optimizes the computer assistive techniques long used in the field.

“What I was really trying to get at with my techniques is trying to push those no-compromise optimizations where you could get a better plan than one that was ‘good enough,’” he told “That’s the fundamental piece… how could we develop methods to get a plan that could not be better, because that’s really what we’re after.”

The Wisconsin Alumni Research Foundation is currently seeking commercial partners for this technology, including Philips’ Pinnacle brand, according to Bender.

He sees it as commercially viable, because it takes out busy work weighing down health care providers, and builds in some strategy from the beginning of the planning process, rather than starting from scratch.

When he first came up with this idea, he says he was spending “a tremendous amount of time” generating dosing plans for cancer treatment.

“As the saying goes, necessity is the mother of invention. So I just started working on coming up with ways to try to streamline the process and try to automate some of it,” he said. “I developed methods to automate a lot of these things, a lot of best practices, rules of thumb that I knew would get me where I want to go.”

Importantly, he says he never intended to make it completely automated, “where I literally click one button and then the plan is done — although with some of the simpler cases I could have done that — I was really more interested in the more complicated disease sites, where some manual intervention is needed as well.”

He says his ultimate goal is to find a treatment planning system manufacturer or vendor — such as Philips — that’s interested in partnering up to develop the method and integrate it with existing products. He says this is the “fastest and best way forward” to “get this out into the real world, so to speak.”

If companies don’t bite, he says he would likely integrate his method with a newer planning system currently being installed at MercyHealth, the cancer center where he works in Janesville.

His first automated methods were developed for cancers of the head and neck area, as that’s one of the more complicated disease sites to treat. But ultimately, the method can apply to most diseases encountered in radiation oncology, he said.

“That would include things like lung cancer, prostate cancer — there’s another common one — and brain cancer as the main other sites,” said.

See WARF’s summary of the patent details:

–By Alex Moe