A radiation therapy quality assurance and commissioning device that measures beam parameters such as beam shape and position in real time, using a scintillator, a precision two-mirror periscope, and a calibrated camera. Made in collaboration with Leo Cancer Care.
Most radiation therapy facilities run their daily QA using four sequential setups: a CT phantom, an IGRT target, a surface-guidance phantom, and one or more beam-dosimetry tools. The industry is lacking a comprehensive device that can do all these in one setting. QA Tron consolidates the entire daily QA chain of events into one radio-translucent instrument that connects with the control computer using a single PoE cable.
The CT phantom interface, the surface-guidance fiducials, the energy-degrading wedges, and the scintillator dosimetry chain share one coordinate frame inside one enclosure. No phantom stack to assemble each morning. No jig to maintain. No inter-device transitions.
The upper region is built from low-Z materials (Delrin, carbon fiber, acrylic mirrors), so the same hardware passes through CT as an unremarkable phantom. Conventional ion-chamber dosimetry phantoms can't meet this requirement.
Three integrated polyethylene wedges produce three Bragg-peak measurements in a single beam delivery. Beam energy is verified across the clinical range.
A single CAT-6 PoE cable carries power and data. For upright systems, the backrest QA shelf installs the QA Tron reproducibly into position on the UPPS every day. For supine tables, the QA Tron simply rests on the patient couch.
Every beam parameter QA Tron measures is processed instantly and compared against set tolerances. The operator sees a summary view indicating which parameters passed, are borderline, or failed. The daily go / no-go decision comes from a single review screen.
Daily QA Mode compares live measurements against a stored golden reference and color-codes go / no-go per metric. Precision Mode saves raw images for commissioning, reference data, and annual QA.
Every frame is background-subtracted, denoised with a 5-pixel median filter, lens-distortion-corrected, and scaled to millimeters at the scintillator plane, before any analysis runs.
Every spot in a user-defined region is fit to a rotated 2D Gaussian. The software returns centroid, σₓ, σᵧ, rotation angle, and amplitude, with live vs. reference profile overlays on both axes.
Rectangular ROIs over the three wedge regions extract distal R₅₀% in a single acquisition. Peak position and distal edge shifts reported against the reference for all three energies simultaneously.
Every tolerance (spot centroid, spot width, intensity, R₅₀% shift, magnification) is set from a dedicated configuration tab. Color-coding on the summary view reflects each facility's own thresholds.
All calculated metrics export as CSV or text for downstream analysis, archival, or feeding into the facility's QA reporting system. Raw images can be saved alongside for audit traceability.
Radio-translucent upper region passes through the CT volume as an unremarkable low-Z phantom. Embedded CT contrast features support routine image-quality QA in the same coordinate frame as every other test.
Applied positioner offsets (X = 4 mm, Y = 8 mm, Z = 10 mm, rotation = 1°) recovered via CT registration and compared against tolerance bands. Sub-millimeter agreement validated on PBS proton systems.
External 1 mm fiducials are recovered by the room's surface-guidance system as an independent cross-check on the same applied offsets. Works with OGTS, C-Rad, Vision RT, and equivalents.
2D Gaussian fits return centroid and σₓ, σᵧ for every spot in a user-defined region. Compared against a stored golden reference and color-coded against user-configurable tolerances.
Three integrated polyethylene wedges produce three simultaneous relative Bragg-peak curves at three energies. Distal R₅₀% shift gives proton range with sub-millimeter sensitivity.
Analytical volume under the fitted Gaussian (2π · a · σₓ · σᵧ) is linear in MU and independent of beam energy, robust to small spot-shape variations that distort peak-intensity measurements.
Backrest QA bracket indexes the device to the same interface as the patient backrest, removing room re-survey from daily setup. Residual offset after correction confirmed below 0.5 mm and 0.2°.
Sub-millimeter centroid extraction from a delivered spot grid quantifies the vector from radiation isocenter to the room lasers.
QA Tron's design and sensitivity are documented in two papers and a graduate thesis, with performance measured on a clinical pencil-beam scanning proton system and benchmarked against a commercial TPS.
Sensitivity validation was performed at the Thompson Proton Therapy Center on an IBA ProTeus Plus pencil-beam scanning proton system, against the RayStation treatment planning system. Results below are reported in Shanks et al. (2025) and the companion technical note by Schreuder et al.
QA Tron supports a workflow we call the QA Patient: a daily verification that touches every subsystem the clinical workflow depends on, in the same order, with the same software paths, just with the QA Tron in the chair instead of a patient.
Room interlocks and AV verified. Operator logs into the OIS / Leo Control System and selects the QA Patient.
QA Tron installs on the UPPS via the backrest QA bracket. Positioner moves to a deliberately offset imaging position (X = 4 mm, Y = 8 mm, Z = 10 mm, rotation = 1°).
CT acquisition and registration against the reference CT. Recovered correction vector compared against the applied offsets. CT image quality verified in the same scan.
OGTS (or equivalent surface-guidance system) recovers the same applied offsets via the external 1 mm fiducials, an independent cross-check against the IGRT result.
Correction applied via the hand pendant, second CT acquired, residual-offset check confirms positioner accuracy below 0.5 mm and 0.2°.
Spot positions, beam energies via three-wedge R₅₀%, and cumulative output verified against the golden reference. One screen, color-coded go / no-go.
QA Tron does not require a vendor lock-in, custom protocol, or proprietary control bus. It speaks the same languages your existing hardware and treatment planning system already speak.
Leo Cancer Care MARIE / UPPS. Backrest QA bracket uses the same indexing interface as the patient backrest. Compatible with conventional couch-based PPS through the seat-column QA platform.
Validated with RayStation. CT data imports natively; PLD plans designed against the device CT ensure planned spot grids fall on the scintillator.
OGTS · C-Rad · Vision RT · or any commercial SGRT system that recognizes external 1 mm fiducials on the device exterior.
Vertical or horizontal diagnostic-quality CT · CBCT · planar kV. Radio-translucent upper region means no high-Z artifacts in the IGRT image set.
Pencil-beam scanning proton systems (validated on IBA ProTeus Plus). Platform-neutral. Any PBS system that can deliver a defined spot pattern is supported.
Single CAT-6 connection carries power and data to the PoE camera. Pre-provisioned Triax line for a future in-beam ionization chamber.
The QA-Tron control software runs the full daily workflow on one screen: connect, acquire, compare against your reference, and read the deltas. Three operating modes (QA, Precision, and Offline review) cover the daily check, commissioning-grade analysis, and after-the-fact review of saved images.
Pricing, lead time, and integration with your imaging and reporting systems.