MPC: Experiences After 12 Months of (Near-) Daily Use
On Nov 3, 2015 we temporarily suspended MPC. Then we replaced a couple of thyratrons (main + deQing), solved some other problems, and resumed daily morning MPC checks.
In December 2015, we installed the interface between the Catalyst HD systems and the TrueBeam (ADI + EXGI):
In the context of MPC, this can produce "false failed" MPC results1. If the user forgets to bypass Catalyst before starting MPC, Catalyst issues an interlock at beam-on. In the meantime, we learned to deal with that.
MPC software version is still the same as last November, so there is still room2 for improvement regarding MPC usability.
The following screenshots show the MPC history between Nov 2015 (with repeated checks) or Aug 2015 (without repeated checks) and March 2016. Without repeated checks, only the last check result of a day is displayed.
TrueBeam #1275
With repeated checks:
Without repeated checks:
TrueBeam #1469
With repeated checks:
Without repeated checks:
The big picture in the meantime looks better than last November. Sometimes, a check had to be repeated. However, it can't be said from the overall result (red or green) whether this was due to a forgotten Catalyst override or some other problem.
Isocenter Size
It was mentioned elsewhere that isocenter size of TrueBeam 1469 was close to warning level (which starts at 0.45 mm), in contrast to the IsoLock results. One year after that, the measured MPC value is close to 0.40 mm, or even lower:
On the other machine (#1275), isocenter size is still slightly better, and the curve is less noisy:
We have neither an explanation for the noisier #1469 curve, nor for the difference between the two machines. We think it is rather unlikely that the improvement in the #1469 values was caused by the DMI upgrade in mid-December: The last measured value in the "orange zone" was on November 18, and the MV detectors were replaced on December 16.
Notes
1A check can be deleted on the TrueBeam console from within MPC. However, the deleted check is still visible on the MPC Offline Client. The mirroring process from TrueBeam to I-drive seems to exclude "delete commands".
2The room has the size of a medium-size concert hall.