Prior to developing the new system, the team at the Centro di Protonterapia of the Azienda Provinciale per i Servizi Sanitari (APSS) used three commercial devices with a phantom to perform daily QA for a single gantry room. The process took between 40–45 minutes. When the proton therapy centre added a second gantry room, lead author Nicola Bizzocchi and colleagues decided to develop a much cheaper, simpler and faster system.

The system utilizes a commercial ionization chamber detector array (MatriXX-PT from IBA Dosimetry) and the dedicated phantom. The team developed and tested two prototypes prior to the creation of the SPREAD (spot positioning, range energy and absolute dose) phantom. The final design consists of slabs of polymethylmethacrylate (PMMA) with a base volume of 24 x 24 x 29 cm and weighing about 5 kg. The idea behind this design was to adapt the beam characteristics, paired with a proper analysis, to the low resolution of the detector, the authors explain.

The phantom contains two pairs of wedges to sample the Bragg peak at different depths, with the image being a transposition on the transverse plane of the depth dose. Three blocks, with 8 x 8 cm bases and heights of 4, 8 and 12 cm are used to check spot positioning and delivered dose. The thickness of these boxes helps spread the single spot and to fit a Gaussian profile on a low-resolution detector.

The researchers deliver two fields on the SPREAD phantom. The first has 12 mono-energetic spot patterns for range verification, beam positioning and verification. The second field is for absolute dose measurement.

The team emphasizes that to obtain a correct setup, the ionization chamber needs to be placed with sub-millimetre precision in the same position every day, using a metal bar as a guide. The crosshair of the measurement device is positioned at the isocentre using lasers, an X-ray image is taken, and the position of the device is corrected to keep the residual positioning error within 0.3 mm. The SPREAD phantom is then positioned on the ionization chamber using two plastic bars integrated in the phantom itself as guides.

Dedicated dosimeters measure range, spot positioning, spot shape and absolute dose, and immediately afterwards the first field is delivered on the SPREAD phantom in order to obtain a reference (or pristine) measurement.

System tests

The researchers conducted an initial 30-day analysis to test the accuracy and repeatability of the system. They specifically tested whether the system could detect errors larger than their action levels of 1 mm in spot positioning, 2 mm in range and 10% in spot size.

Data acquired from a year of use showed high stability and high reliability in terms of range consistency, spot positioning precision and spot-sigma reproducibility. The team reported that over a 35 day period, the standard deviation of the range was always lower than 0.1% for every energy, spot positioning error distributions were always within tolerance value (±1.0 mm), and the average of the standard deviation, in terms of percentage of the mean, was 0.08% with a maximum deviation of 0.42% for a single energy.

Medical physicist Francesco Fracchiolla, co-designer of the SPREAD phantom, told medicalphysicsweb that it took approximately six months to develop the hardware and the software for the phantom. He noted that medical physicists who have visited the Centro de Protonterapia have expressed interest in building and using the phantom at their own proton therapy centres. He welcomes enquiries (by email to francesco.fracchiolla@apss.tn.it) from readers who may also be interested in constructing the SPREAD phantom.

Related stories

• Adaptive treatments are a must for protons
• Optimized plastics improve proton dosimetry
• Dose calculator minimizes QA bottleneck
• Log files offer QA for spot-scanned protons