The first thing that the Medical Physicists at our clinic tested was the IQM’s error detection capability: Systematic 1mm leaf errors were introduced, which made the field smaller and resulted in a systematic signal drop in every beam segment. Checking the same error with the 2D array only shows an underdose in blue, but it doesn’t show that this was a systematic error in all segments.
Our team further checked if the magnitude of the error correlated with the signal deviation. It was clearly detected that the signal deviation increases and decreases according to the magnitude of the error.
The next test evaluated the clinical impact of the errors shown in IQM. For this purpose, the IQM signal deviation was plotted against the dose difference in the PTV. 250 data points were evaluated and a very robust correlation of 0.97 was found. We defined the tolerance corridor of the IQM based on this correlation. A maximum overdose of 5% in the PTV was used to define the upper signal corridor and a maximum allowed underdose of -2% defined the lower tolerance limit. In the entire year of using IQM the clinical staff has never seen a dose difference of more than 5%.
The team also introduced random errors in a single control point. IQM allows us to exactly locate the occurrence of the error. This type of error would not be caught with a 2D array, while IQM clearly shows the location of the error.
Finally, our team introduced various data transfer errors. Complete integration into the clinical treatment delivery process allows IQM to detect these types of errors and alarm the user before the patient treatment even starts, which makes patient treatment a lot safer.
Once we were satisfied with IQM’s ability to detect errors, we tested the detector’s robustness with extended linearity tests, dose rate dependence tests and signal stability tests. We compared these against measurements with our Farmer chamber. The detector showed very robust results, which is very important for the clinical routine.
The long-term stability of the IQM System was checked over a 3 month course of daily QA checks. We analyzed the data acquired on 3 Elekta Versa HD machines with 3 different IQM Detectors during a period of 3 months. The maximum deviation during the entire time period was less than 2%, while the linac output data and the IQM measurements weren’t even normalized with each other. This test shows that both the IQM and the Linac work perfectly and are very stable together.
Another large area of interest was the impact of the IQM system on patient treatment safety. The conventional QA process only checks if a treatment technique can be delivered as planned prior to the actual patient treatment. However, nobody knows if the delivery was accurate in every single fraction. As an example, my colleagues checked the treatment monitoring of a very complex H&N VMAT plan. All signal were always within 1% of the expected signal, indicating a highly accurate delivery of this complex technique during the entire course of treatment.
Another important advantage of the IQM system is the complete integration into the clinical workflow. Treatment plan verification happens without any user interaction. The user simply has to export a plan, which happens with a single mouse click. All other steps of the process are done fully automatically. The only thing the user has to do in addition is to evaluate the QA report which takes about 30 seconds. Setting up the detector takes about 10 seconds. The measurement happens fully automatically. This is a major efficiency increase compared to the workflow of a 2D array, where everything has to be done manually. IQM requires about 2:30 minutes time per day, while our previous conventional VMAT QA using a 2D array took about 21:30 minutes per plan. IQM does not require any user interaction at all.
Our clinic designed its own report, which fully complies with the department regulatory requirements. All statistic and graphs are designed exactly as needed in the clinical workflow. The IQM data is very easily accessible and can be processed very simply. In the future we will waive all reports with green lights and only focus on the reports that showed a problem.
To summarize our experience with IQM after the first year in the clinic:
- IQM is a very simple and robust verification system
- IQM significantly increases patient safety
- IQM provides a fully automated workflow and requires absolutely no user interaction