To answer this, German start-up iRT is creating what it says is a completely new approach to radiotherapy QA. The company's Integral Quality Monitor (IQM), developed in collaboration with the Princess Margaret Cancer Centre in Toronto (Med. Phys. 36 5420), is a real-time beam verification system that monitors the accuracy of radiation delivery throughout each patient treatment without any user interaction.

"IQM continuously reviews every segment of every beam during every treatment fraction," explained iRT's managing director Jürgen Oellig. "It is the first intra-fractional verification system available on the market." If beam delivery deviates from the prescribed plan at any point, the treatment can be paused or halted. "This device replaces error management with error prevention - errors cannot find their way to the patient any more," he added.

As well as improving patient safety, the IQM also increases efficiency, said Oellig. He described an example radiotherapy centre treating 700 patients per year, 300 of which are complex intensity-modulated radiotherapy (IMRT) or volumetric-modulated arc therapy (VMAT) cases. Each linac requires daily QA, plus routine annual and monthly checks. The IMRT and VMAT treatments also require pre-treatment QA, leading to an average of 650 verification sessions per year.

"So if each patient receives 20 fractions, and IQM is used to measure every fraction, this adds 14000 QA sessions," Oellig told medicalphysicsweb. "It's a completely different level – for the first time it is possible to make every patient treatment into a verification session."

Keep it simple

Unlike existing QA systems, which comprise arrays of hundreds of detectors, the IQM uses a single large-area ion chamber. The IQM detector comprises three 1.5 mm-thick aluminium plates: two polarizing electrodes sandwiching a collector electrode. Spatial sensitivity is achieved by linearly varying the separation between electrodes across the chamber.

The IQM is positioned directly underneath the linac collimator, enabling the treatment beam to pass through the ion chamber. Oellig noted that the device does not impact the beam quality. It does attenuate the beam slightly – by 6.5% and 5% for 6 and 18 MV photons, respectively – but this attenuation is homogenous over the entire field and can thus be accounted for during treatment planning.

Clinical workflow

The IQM system is fully automated at all stages, only requiring export of a DicomRT Plan data from the treatment planning system. The IQM's calculation algorithm uses these imported data to calculate an expected signal for each individual beam segment, based on the prescribed monitor units (MU), field size, field shape and location. The IQM chamber sensitivity map used in this calculation comprises a matrix of 801 x 801 data points, equivalent to a grid of 641601 data points.

The system then expands each calculated signal into a "corridor", to allow tolerance for expected daily linac fluctuations, and displays this inner corridor in green. In addition, a user-defined outer corridor, which separates acceptable from potentially harmful deviations, is displayed in red.

Once the patient is on the linac couch and treatment is about to start, the individual plan-related data are automatically loaded from the IQM database. During beam delivery, the IQM compares the measured signal with the calculated reference signal.

"The system displays the measured signal as a dot, and as long as this dot stays within the inner corridor everything is fine," explained Oellig. "If a measurement falls outside the green corridor but is inside the red boundary, we flag the error and inform the physicist." In such cases, it may be possible for the physicist to correct the detected error and continue treatment.

However, if a measured signal deviates from the expected value by more than a user defined tolerance (falls outside the red line), the system creates a clinical alarm and the beam can be switched off. "Other measurement systems can only tell if something went wrong after the fraction has been delivered to the patient," said Oellig. "IQM will tell you whilst the patient is being treated."

After each fraction, the system sends an email to all relevant users with a general warnings report and full details of treatment.

Excelling in accuracy

IQM can detect errors related to dose (incorrect photon energies or MU, for example), beam direction (wrong gantry or collimator angles) and beam segment shape. A 2% change in MU creates a 2% signal change and a 1 mm shift in beam placement creates a 0.5% signal change. For a 3 x 3 cm field size, a 2 mm single leaf deviation creates a 2% signal change, while a 2 mm leaf bank change creates a 7% signal change.

Oellig noted that the IQM is a lot like film in terms of resolution. "Array-based detectors are spaced 7 mm to 1 cm apart, so you need 1 cm of movement to detect an error as there is no active area in between," he explained. "Ours offers almost ten times higher resolution with a single detector."

In a panel discussion at the recent AAPM Annual Meeting, Robert Heaton, from the Princess Margaret Cancer Center and part of the IQM development team, described how the project began about 10 years ago, with the research team brainstorming how to improve IMRT validation and better prepare for the introduction of adaptive radiotherapy. "When considering this problem, we had seen many multi-array detectors, but these had issues with complexity and interpretation of the signals," Heaton explained. "We wanted to develop a system that was simple and reliable, so we went to single-channel detector that could give a quality factor for beam delivery without requiring complex analysis."

IQM developer David Jaffray described IQM as a constant sentinel. "I like the idea that this gives real-time feedback to the therapist," he told the gathered crowd. "The users know what signal to expect and can see whether the measured signal is inside the corridor. It's interesting that we have data streaming all the time; we need to consider how that is going to impact our QA processes."

A pre-clinical prototype of the IQM system is currently being tested at 17 radiotherapy centres worldwide. iRT expects to achieve the European CE Mark in September, with US FDA approval targeted for the end of this year.

Related stories

• The 'Best-in-Physics' at AAPM
• Radiotherapy quality assurance: progress and needs
• QA process focuses on MLC analysis
• Radiotherapy safety: the key role of QA