Dosimetry and Treatment Verification
Dosimetry and Treatment Verification tests whether you can independently confirm that planned dose is delivered dose, using hand calculations, calibrated chamber measurements, and patient-specific QA. Expect arithmetic plus judgment about action levels.
Monitor unit fundamentals
- SSD setup: MU = prescribed dose / (output x PDD x Scp x inverse-square factor x modifier factors).
- SAD setup: use TMR instead of PDD, with the inverse square of SAD over depth, plus the collimator factor.
- TG-51 reference conditions: 10 cm depth in water, 10 x 10 cm field, 100 cm SSD for photons, with N-sub-D,w from an accredited laboratory.
Chamber corrections
- Temperature-pressure: P-TP = ((273.2 + T)/295.2) x (101.33/P), with T in degrees Celsius and P in kPa.
- Also apply ion-recombination, polarity, electrometer, and gradient corrections as applicable; the gradient correction is unity for photons at 10 cm.
Verification tools
- Independent MU second check agreeing within 2-3 percent (tighter for non-IMRT, looser for IMRT and VMAT).
- Patient-specific IMRT/VMAT QA with gamma analysis: TG-218 recommends 3 percent/2 mm, 10 percent dose threshold, 95 percent pass rate under global normalization.
- In vivo dosimetry (diodes, OSL, EPID) with action levels near 5 percent; ICRU targets about 5 percent overall delivered-dose accuracy.
Worked example
Prescription 200 cGy at 10 cm depth, 6 MV, 100 cm SSD, 10 x 10 cm open field, PDD(10) = 66.5 percent, Scp = 1.000, output 1.000 cGy/MU. MU = 200 / (1.000 x 0.665 x 1.000) = 301 MU. If a 15-degree wedge with factor 0.700 is added, MU = 200 / (0.665 x 0.700) = 430 MU, so the wedge costs roughly 43 percent more monitor units.
Traps
- Omitting the inverse-square term when depth or SSD differs from the calibration geometry.
- Using PDD in an SAD calculation instead of TMR.
- Reporting a pass rate obtained with local normalization as if it were global.
- Treating a 95 percent gamma pass as proof of accuracy when the wrong structure, shift, or plan version was tested.