Machine calibration and quality assurance
Reference dosimetry
TG-51 uses an ion chamber with a cobalt-60 calibration coefficient N_D,w, measured at 10 cm depth, 10 × 10 cm², 100 cm SSD in water, corrected for temperature-pressure (P_tp), ion recombination (P_ion, important for pulsed and scanned beams), polarity (P_pol), and electrometer (P_elec), plus the beam-quality factor k_Q interpolated from the measured quality specifier. TG-21, which referenced dmax in a phantom and used exposure calibration factors, is a classic distractor.
QA program structure
TG-142 organizes tests by daily, monthly, and annual frequency with separate tolerance and action levels. Representative values: output constancy ±3% daily, ±2% monthly, absolute calibration ±1% annually; wall/ceiling lasers ±2 mm daily and ±1 mm monthly; gantry and collimator angle indicators ±1°; MLC leaf position ±1 mm. Confusing tolerance with action level, or applying a monthly tolerance daily, is a frequent question trap. Acceptance testing establishes the baseline before commissioning; commissioning then produces the beam model, and both must be documented.
Worked example
P_tp = [(273.2 + T)/295.2] × (101.33/P). At 22.0 °C and 98.5 kPa: P_tp = (295.2/295.2) × (101.33/98.5) = 1.029. Skipping this correction causes a 2.9% systematic calibration error, which is why thermometer and barometer calibration belong in the QA program.