Physicist's Notes · ART Phantom

Seven steps from image-the-phantom to signed QA dossier.

End-to-end treatment-chain QA runs the same workflow as a patient treatment, with dosimeters in place of body tissue. Seven steps from CT-simulation through delivery to dossier filing. This post walks each step, the dosimeter-to-task mapping for the five measurement modalities, and how the dossier fits AAPM TG-119 (commissioning) and TG-218 (patient-specific) expectations.

Seven-step cycle

From CT-simulation to signed QA result

1

Image the phantom

CT-simulate the loaded phantom with the same protocol used for patient imaging — slice thickness, kV, mA, reconstruction algorithm matched.

Use the standard patient-imaging protocol. The CT dataset feeds the treatment planning system as if it were a patient.
2

Plan the treatment

Contour the phantom-internal target volumes (placed via TLD / dosimeter pin positions) and OARs. Plan a clinically realistic IMRT / VMAT treatment.

The plan should match the clinical complexity the QA is verifying — head-and-neck, prostate, breast, lung, etc.
3

Insert dosimeters

Load the chosen dosimeters (TLD, OSL, MOSFET, ion chamber, film) into the slice positions that correspond to the planned target and OAR points.

TLD positions stored in the standard phantom map; clinical-physics group has the position-to-anatomy lookup table.
4

Deliver the plan

Set the phantom up on the linac couch as if it were the patient — couch indexing, lasers aligned to phantom-marked iso-centre. Deliver the full treatment plan.

Couch indexing matters. Misaligned couch position is a common source of measured-dose discrepancy.
5

Read the dosimeters

Remove the dosimeters and read them on the corresponding reader (TLD reader for TLD, OSL reader for OSL, etc.). Record measured dose at each position.

Calibration of the reader against a NIST-traceable source is part of the dosimeter calibration cycle.
6

Compare against TPS

Plot measured dose vs TPS-predicted dose at each position. Accept within ± 3 % (gamma index 3 %/3 mm, 95 % pass rate for IMRT QA per TG-218).

Failure mode is not just absolute dose — relative geometry of error tells you whether the fault is in imaging, planning or delivery.
7

Document the result

File the result in the per-machine, per-modality QA dossier. Annual programmes file the result against the AAPM TG-119 / TG-218 commissioning baseline.

Documentation is what survives accreditation audit. The number alone is not enough; the comparison narrative is.

Source: AAPM TG-119 IMRT Commissioning; AAPM TG-218 Patient-Specific IMRT QA.

Dosimeter-to-task mapping

One phantom across five measurement modalities

Dosimeter	Use case	Phantom insertion
TLD chips	Point dose at target + OAR positions; low-cost per measurement	Pin-hole pluck-and-place; standard 3 cm × 3 cm or 1.5 cm × 1.5 cm grids
OSL dosimeters	Repeated-readout dosimetry across multiple QA cycles; lower fade error than TLD over weeks	Same pin holes as TLD; OSL holders standard accessory
MOSFET	High-resolution point dose for SRS / SBRT QA where mm-scale gradients matter	Small-diameter (1.5 mm) MOSFET fits the 5 mm pin hole
Ion chamber	Reference dosimetry — absolute dose at specific calibration points; gold-standard comparison	Larger pin-hole positions; specific chamber-compatible inserts required
Film	Continuous 2-D dose distribution between slices; gamma-map analysis	External assembly mode — open slice gaps to seat film sheets

AAPM TG-119 — Commissioning

The published commissioning baseline the ART phantom is designed around

AAPM TG-119 IMRT-commissioning planning baselines: H&N (oropharynx + prostate-on-pelvis), C-shape (concave PTV around OAR), multi-target prostate. ART phantom holds the dosimeters at the published TG-119 anatomical positions.
Baseline acceptance: measured dose at each TG-119 dosimeter position within ± 3 % of TPS prediction. Gamma 3%/3mm with 95 % pass rate across the dose distribution.
New-machine commissioning runs the full TG-119 plan-set on the ART phantom. Re-commissioning (post-major-service, TPS upgrade, MLC service) repeats the relevant subset.

AAPM TG-218 — Patient-specific IMRT QA

Per-patient delivery verification

AAPM TG-218 patient-specific IMRT QA: each patient plan re-computed on the phantom geometry and delivered. Measured-vs-predicted dose verifies the delivery before the patient is treated.
Per-patient acceptance: gamma 3%/3mm, 95 % pass rate within the body contour. Lower pass rate triggers plan re-evaluation before treatment proceeds.
ART phantom's anatomically realistic geometry tightens the gamma analysis — flat-slab phantoms relax the gradient comparison; anatomically positioned dosimeters do not.