News of these accidents motivated some of our approach on a more recent radiation therapy machine project. Here's a short paper I wrote about it.

http://staff.washington.edu/jon/cnts/iccr.html

The software we wrote has been used without incident for more than ten years.