Imaging for radiation therapy and simulation
Simulation CT
CT simulation fixes the treatment geometry: indexed couch, immobilization, isocenter marking, and a protocol with appropriate field of view and slice thickness (≤3 mm for most sites, 1 mm for intracranial radiosurgery). Hounsfield units must be converted to electron density with a calibration phantom, and that curve is scanner- and kVp-specific. Metal artifact, intravenous contrast, and patient motion degrade both contouring and dose calculation.
Other modalities
CBCT and MVCT provide geometric verification, but CBCT Hounsfield values are unreliable for dose calculation without correction. MRI improves soft-tissue delineation but introduces geometric distortion from gradient nonlinearity and susceptibility that must be measured and corrected. PET adds metabolic information; SUV thresholds for target definition require consistent uptake time and reconstruction. 4DCT produces phase- or amplitude-sorted bins; the maximum intensity projection is used to define the internal target volume, while the average CT is used for dose calculation.
Worked example
A scan with CTDIvol = 10 mGy over 40 cm gives DLP = 10 × 40 = 400 mGy·cm. Using k ≈ 0.015 mSv/(mGy·cm) for abdomen, effective dose ≈ 6 mSv - an order-of-magnitude check, not a patient-specific dose.
Traps
Equating Hounsfield units with electron density, using MIP images for dose calculation, ignoring registration uncertainty between modalities, and assuming imaging dose is negligible.
Radiation protection, safety, informatics, and professionalism
Dose limits and ALARA
Occupational effective dose limit is 50 mSv/y; the public limit is 1 mSv/y; a declared pregnant worker is limited to 0.5 mSv in any month and 5 mSv total. The lens is a classic trap: ICRP recommends 20 mSv/y averaged over 5 years with no year above 50 mSv, while the US NRC limit remains 150 mSv/y. ALARA drives shielding design, which uses workload W, use factor U, occupancy T, and distance, with transmission B = P·d²/(W·U·T).
Medical events and directives
10 CFR 35.3045 defines a medical event by category - wrong patient, wrong radionuclide, wrong route, or a dose substantially different from the written directive (≥20% for a single administration; the rule uses a different threshold for fractionated delivery, so read the current text). A written directive is required for teletherapy, HDR, and IMRT, and the authorized user signs it before administration.
Risk, informatics, and professionalism
TG-100 applies failure mode and effects analysis with RPN = severity × occurrence × detectability; the highest-risk steps receive redundant checks. TG-263 standardizes structure names so plans and data transfer are unambiguous. DICOM RT and HL7/IHE-RO govern interoperability; wrong-patient and plan-transfer errors remain leading failure modes, which is why independent second checks, chart rounds, and peer review are expected professional practice.
Worked example
A source reads 4 mSv/h at 1 m. Doubling distance to 2 m gives 1 mSv/h; at 4 m it is 0.25 mSv/h, illustrating the inverse-square rule behind barrier planning.