Therapy Physics Boards Field Guide

Physics-only rapid review for radiation oncology trainees who have already learned the material and now need fast retrieval under board conditions. Biology is intentionally out of scope.

Legend

A = λN equation or calculation trigger

0.511 MeV number worth memorizing

TG-51 protocol, report, or standard

linac hardware or clinical delivery detail

bold red common board-style trap

Purpose

This is not a textbook. It is a retrieval map: equations, units, assumptions, and the small distinctions that turn a familiar topic into a board question. The writing is intentionally concise and conversational, with mutable protocol and regulatory details checked against public AAPM, NRC, NIST, eCFR, and ICRU reference pages.

Graduating Resident Target

Use this as a cram sheet after first learning the physics. It is optimized for "recognize the question type, choose the governing equation or protocol, avoid the trap, calculate if needed." The densest sections are calculation templates, machine/component logic, QA/protocol tolerances, and brachy/protection rules.

Equation Dashboard

E = hf = hc / λ
Use hc = 1240 eV nm.

A = λN
Activity is disintegrations per second.

T_1/2 = 0.693 / λ
Mean life = 1 / λ = 1.44 T_1/2.

A(t) = A₀e^-λt = A₀(1/2)ⁿ
n = t / T_1/2.

I = I₀e^-μx
HVL = 0.693 / μ; TVL = 2.303 / μ.

Inverse square = (d_old / d_new)²
Short SSD makes this unforgiving.

Equivalent square = 4A / P
Use for rectangular field scatter lookup.

D_w^Q = M k_Q N_D,w^60Co
Core TG-51 relation.

d_ref,e = 0.6R₅₀ - 0.1 cm
Electron reference depth.

R₅₀ ≈ E₀ / 2.33
Electron mean incident energy.

B = P d² / (WUT)
Primary barrier transmission.

TG-43: Ddot = S_KΛGgF / G₀
Geometry, radial dose, anisotropy.

BED = nd[1 + d/(α/β)]
Fractionation comparison.

EQD2 = BED / [1 + 2/(α/β)]
Convert to 2 Gy/fx equivalent.

Calculation Playbook

Units, Decay, and Fractionation

Question Type	Template	Speed Check
Frequency to wavelength	Convert prefix first, then λ = c/f.	3000 MHz = 3 x 10⁹ Hz gives 0.1 m.
Photon energy	E = hf = hc/λ. For wavelength in nm, use 1240 eV nm.	Energy increases as wavelength decreases.
Activity from atoms	Convert half-life to seconds, compute λ = 0.693/T_1/2, then A = λN.	If T_1/2 = 6.93 x 10³ s and N = 10⁴, activity = 1 Bq.
Elapsed half-lives	fraction remaining = (1/2)^t/T_1/2.	1, 2, 3, 4 half-lives = 50%, 25%, 12.5%, 6.25%.
Mean life	τ = 1/λ = 1.44T_1/2.	At one mean life, activity is 36.8% of initial.
Effective half-life	T_eff = T_physT_bio/(T_phys + T_bio).	Always shorter than both physical and biologic half-life.
Fractionation comparison	BED = nd[1 + d/(α/β)], then EQD2 = BED/[1 + 2/(α/β)].	Late-effect tissues use low α/β values; higher dose per fraction matters more.
Nuclear reaction balancing	Conserve total A and total Z on both sides.	Do A first, then Z. It prevents isotope-choice guessing.

Beam Geometry and Monitor Units

Question Type	Template	Fast Trap Check
SSD/PDD MU	MU = Rx / [O S_cS_pWF TF OAR (PDD/100) ISF].	PDD table means SSD geometry. Use field size at patient for S_p.
SAD/TMR MU	MU = Rx / [O S_cS_pWF TF OAR TMR].	Parallel-opposed equally weighted beams: divide Rx by 2 before MU per beam.
Depth-dose transfer	If MU/geometry is unchanged, dose ratio between depths equals PDD or TMR ratio at those depths.	For same SSD beam, D₃/D₆ = PDD₃/PDD₆.
Equivalent square	eq square = 4A/P = 2ab/(a+b) for rectangle.	Use collimator equivalent square for S_c; blocked/MLC field for S_p.
Wedge angle	θ_wedge = 90 - hinge/2.	Hinge 135 degrees -> 22.5 degree wedge.
Gap for matching fields	gap = d(L₁/SSD₁ + L₂/SSD₂)/2.	Use half-field length from central axis to match line. Half-beam block has no divergence on blocked side.
Radiologic depth	d_rad = Σ ρ_relt.	Low-density lung reduces water-equivalent depth; bone increases it.

Shielding and Source Timing

Question Type	Template	Fast Trap Check
HVL/TVL	HVL = 0.693/μ; TVL = 2.303/μ; n = log₁₀(I₀/I).	Each TVL reduces intensity by 10x, not by 90% of original repeatedly in linear units.
Primary barrier	B = Pd²/(WUT).	Uncontrolled area design goal is much lower than controlled area.
IMRT leakage workload	W_L = C W, commonly C = 2-10.	More MU means more leakage workload.
HDR decay correction	t_now = t_nominalS_nominal/S_now.	If the source has decayed, treatment time gets longer.
Permanent implant total dose	Total dose is proportional to initial dose rate x mean life, D ∝ R₀/λ.	Shorter half-life needs higher initial dose rate for same total dose.
Transit dose	More source travel events -> more transit contribution.	For same total dose, more HDR fractions usually means more transit dose.

Most questions are not asking for a fact in isolation. They tie two pieces of information together: a unit plus its base dimension, a machine component plus what changes in electron mode, a formula plus the condition under which it is valid, or a protocol plus what it does not cover.

Default board behavior: identify the quantity, check units, check geometry, check energy range, then ask whether the word "EXCEPT" flipped the task. If the question contains a table, decide first whether it is an SSD/PDD problem or an SAD/TMR problem.

Question Style	Fast Response	Common Trap
Derived SI unit	Gy, Sv, J, N, Bq are derived; kg is base.	Confusing named SI units with base SI units.
Changes with filtration	HVL and effective energy increase; intensity decreases; max energy unchanged.	Thinking filtration increases kVp.
Electron mode on linac	Target out, flattening filter out, scattering foil or scanning in, applicator in, monitor chamber still used.	Removing the monitor chamber.
Short SSD superficial x-rays	Inverse square sensitivity is huge.	Blaming air attenuation first.
Orthovoltage vs MV	Bone dose rises with kV because photoelectric effect is Z-sensitive.	Applying MV Compton logic to kV beams.

Trap Decoder Library

If The Question Mentions...	Think First	Likely Answer Direction
"EXCEPT"	Underline the negation.	The right answer is the non-member, not the most familiar true statement.
SI units	Base vs derived.	kg, m, s, A, K, mol, cd are base; Gy/Sv/J/N/Bq are derived.
Inelastic collision	Conserved quantity.	Momentum conserved; kinetic energy not necessarily conserved.
Cyclotron energy limit	Relativistic mass/phase mismatch.	At high velocity, particle no longer stays synchronized with fixed RF.
Filtered kV beam	Beam hardening.	HVL/effective energy up, intensity down, max energy unchanged.
Gamma Knife in brain	Co-60 energy in soft tissue.	Compton dominates.
18 MV photons	Photonuclear threshold.	Neutron dose is a concern.
SF₆ in waveguide	Insulation/dielectric strength.	Do not choose cooling or acceleration.
TLD vs OSLD	Readout stimulus.	TLD = heat; OSLD = light.
Rounded MLC leaf end	Consistent penumbra plus dosimetric leaf gap.	Do not confuse DLG with physical leaf gap alone.
Annual SRS/SBRT end-to-end	Measured vs calculated dose tolerance.	Think 5% for end-to-end dosimetric agreement.
Patient release after implant/unsealed source	Public exposure from released patient.	Release if another individual unlikely to exceed 5 mSv TEDE.
Proton range question	HU-to-stopping-power and beam-direction uncertainty.	Range margin is not a uniform isotropic PTV expansion.
DICOM object list	RTPLAN/RTSTRUCT/RTDOSE/RTIMAGE/REG/waveforms.	Machine log file is the non-DICOM-style answer.

Unit Sense and Constants

Quantity	Board Value	Use
Speed of light	3.00 x 10⁸ m/s	c = fλ
Planck relation	E = hf = hc/λ	hc = 1240 eV nm
Electron rest energy	0.511 MeV	Pair production threshold = 1.022 MeV.
Proton/electron mass ratio	1836	Proton is about 1800 times electron mass.
Atomic mass unit	931.5 MeV/c²	Mass defect to binding energy.
Avogadro	6.022 x 10²³ mol^-1	Atoms per mole.
Curie	3.7 x 10¹⁰ Bq	Old activity unit.
Roentgen	2.58 x 10^-4 C/kg air	Exposure for photons in air; 1 R ~ 0.876 cGy in air.
Gray	1 J/kg	Absorbed dose.
Sievert	Gy x weighting factors	Protection quantity, not tumor prescription dose.

Base vs Derived SI Units

Base SI	Derived / Named SI	Physics Board Translation
meter (m), kilogram (kg), second (s)	newton, joule, gray	N = kg m s^-2; J = N m; Gy = J kg^-1.
ampere (A), kelvin (K), mole (mol), candela (cd)	coulomb, volt, becquerel, sievert	Bq = s^-1; Sv = J kg^-1 after weighting factors.

Mechanics and Accelerator One-Liners

Inelastic collision: momentum is conserved; kinetic energy is not. If the bodies stick together, it is perfectly inelastic.
Work-energy: W = Fd cosθ and KE = 1/2 mv² for nonrelativistic particles.
Relativistic energy: E² = (pc)² + (m₀c²)²; for photons, E = pc.
Cyclotron limit: fixed-frequency cyclotrons lose synchrony at relativistic speeds because effective mass increases; synchrocyclotrons vary RF frequency to compensate.
Synchrotron: magnetic field and RF frequency are varied so particles stay in a fixed orbit while energy increases.

Atomic and Nuclear Bookkeeping

Atomic number Z is protons; mass number A is nucleons; neutrons = A - Z.
Isotopes share protons, isobars share A, isotones share neutrons. Memory: P-A-N.
Shell capacity = 2n². K shell has 2, L has 8, M has 18.
Characteristic x-rays occur when an outer-shell electron fills an inner-shell vacancy. Auger electrons are the competing non-radiative route.
Binding energy per nucleon peaks near iron. Fission of heavy nuclei and fusion of light nuclei both move toward higher binding energy per nucleon.

Board check: A photon has frequency 3000 MHz. What is the wavelength?

Convert first: 3000 MHz = 3 x 10⁹ Hz. Then λ = c/f = 3 x 10⁸ / 3 x 10⁹ = 0.1 m. The trap is failing to move MHz to Hz.

Board check: Which particle can have a bremsstrahlung interaction?

Any charged particle can radiate bremsstrahlung when accelerated/decelerated in a Coulomb field. In practice, electrons and positrons matter most because radiative losses scale inversely with particle mass squared. Neutrons, photons, and neutrinos are not the classic charged-particle bremsstrahlung answer.

Decay Mode Signatures

Mode	Signature	Board Memory
Beta-minus	n → p + e^- + anti-ν; A same, Z increases by 1	Neutron-rich nucleus moves up in Z.
Beta-plus	p → n + e⁺ + ν; A same, Z decreases by 1	Needs at least 1.022 MeV for positron creation.
Electron capture	p + e^- → n + ν; A same, Z decreases by 1	Produces characteristic x-rays/Auger electrons from shell vacancy.
Alpha	A decreases by 4, Z decreases by 2	Heavy nuclei; discrete alpha energies.
Gamma	No A or Z change	Nuclear de-excitation.
Internal conversion	No A or Z change; orbital electron ejected	Competes with gamma emission.

Decay Diagrams Without Panic

Vertical transition on a decay scheme = nuclear de-excitation, usually gamma or internal conversion.
Beta spectra are continuous because energy is shared with a neutrino/antineutrino.
Alpha and gamma emissions are discrete-energy processes in simple diagrams.
Electron capture and internal conversion both produce atomic vacancies, so both can lead to characteristic x-rays or Auger electrons. The nuclear change is different.
Positron annihilation produces two 0.511 MeV photons traveling nearly 180 degrees apart after the positron slows.

Activity Math

A = λN. If you know atoms and half-life, convert half-life to seconds, compute λ, then activity in Bq.
A(t)=A₀(1/2)^t/T_1/2. Count half-lives when possible; it is faster and safer than logs.
Rule of 72: approximate percent decay per unit time = 72 / T_1/2 when the half-life is in that unit.
T_eff = (T_physT_bio) / (T_phys + T_bio). Effective half-life is always shorter than either component.
Total decays from a source = A₀ / λ. Permanent implant total dose is proportional to initial dose rate divided by λ.

Parent-Daughter Equilibrium

Parent vs Daughter	Name	What Happens
Parent half-life much longer than daughter	Secular	Daughter activity grows until it approximately equals parent activity.
Parent longer, but not enormously longer	Transient	Daughter rises, peaks, then appears to decay with parent; daughter can exceed parent.
Parent shorter than daughter	No useful equilibrium	Parent disappears before sustaining daughter.

Exam trap: gamma emission and internal conversion compete with each other. Beta-minus and beta-plus do not "compete" in the same nucleus in the same simple way; they reflect opposite N/Z pressures.

Generator and Activation Logic

A useful generator has a parent long enough to ship/use and a daughter short enough to grow in quickly.
For neutron activation, add the incident particle to the target nucleus and conserve A/Z after emitted particles leave.
Specific activity increases as half-life decreases for the same number of radioactive atoms because A = λN.

Charged Particle Logic

Stopping power is energy loss per path length. Collision stopping makes ionization/excitation; radiative stopping makes bremsstrahlung.
Collision stopping roughly increases with z²/v². Doubling particle charge matters more than doubling mass.
Bremsstrahlung yield rises with particle energy and absorber Z, and is far more important for electrons than heavy charged particles.
LET is local energy deposition density. High LET generally means dense ionization, higher RBE, and lower OER, but clinical protons are still assigned RBE 1.1 in standard planning.
Electron multiple scattering is why electron fields spread laterally and why small cutouts lose lateral charged particle equilibrium.

Particle	Dominant Exam Physics	Clinical/Shielding Memory
Electron	Multiple Coulomb scattering, collision + radiative stopping.	Finite practical range, broad penumbra with depth, bremsstrahlung tail.
Proton	Collision stopping, Bragg peak, nuclear interactions.	Range uncertainty and distal edge sensitivity.
Alpha	High charge, high LET, short range.	Dangerous internally, stopped by very thin external shielding.
Neutron	Indirectly ionizing; elastic/inelastic scatter and capture.	Hydrogenous moderation, capture gammas, high radiation weighting factor.

Photon Interaction Logic

Interaction	Where It Matters	High-Yield Consequence
Coherent/Rayleigh	Low energy, high Z	No energy transfer to matter; contributes to scatter blur in imaging.
Photoelectric	kV range, high Z	Strong Z dependence; bone and iodine stand out.
Compton	Therapy MV range in soft tissue	Depends mainly on electron density, not Z. Dominant for Co-60/Gamma Knife in brain.
Pair production	> 1.022 MeV, rises with energy and Z	Produces positron/electron pair; annihilation photons after positron slows.
Photonuclear	High-energy linac photons, especially >10 MV	Neutron production concern for 15/18 MV beams.

I = I₀e^-μx is narrow-beam attenuation. Broad-beam geometry adds scatter back into the detector, so apparent attenuation is less severe. HVL hardens a polyenergetic beam because low-energy photons are preferentially removed.

Interaction	Approximate Dependence	Answer Shortcut
Photoelectric	Roughly Z³/E³	Dominates kV contrast and bone dose.
Compton	Electron density, weak Z dependence	Dominates MV soft-tissue therapy dose.
Pair production	Threshold 1.022 MeV, rises with Z and energy	Relevant above threshold, more with high-energy beams/high-Z materials.
Photonuclear	High-energy photons interacting with nucleus	Neutrons from high-MV treatment heads.

Board check: For a soft-tissue head target, why does Gamma Knife dose deposition mainly follow Compton physics?

Co-60 emits about 1.17 and 1.33 MeV photons, average about 1.25 MeV. In soft tissue at those energies, Compton dominates.

Neutron and Proton Logic

Fast neutrons lose energy most efficiently in hydrogen-rich material through elastic scattering; slow neutrons are more likely captured.
Neutron protection needs hydrogenous moderation plus capture shielding. For high-energy photon vaults, maze design and door composition matter.
Protons have finite range and a Bragg peak. The distal edge is biologically and geometrically sensitive, so avoid placing it in a critical serial OAR when possible.
Range uncertainty comes from CT number to stopping-power conversion, setup/anatomy changes, and motion. Robust optimization is not optional thinking in proton planning.

Kilovoltage Tube Logic

Thermionic emission releases electrons from the cathode. Tube current controls photon quantity; kVp controls spectrum endpoint and beam quality.
λ_min = hc/eV. Maximum photon energy equals electron kinetic energy; filtration does not change that endpoint.
Filtration removes low-energy photons, increasing HVL/effective energy and decreasing intensity.
Anode angle increase makes the effective focal spot larger, improves field coverage, reduces heel effect severity, and worsens spatial resolution.
The oil bath removes heat and provides electrical insulation.
Tungsten K-characteristic x-rays need incident electrons above the K-shell binding energy; K lines are roughly 59 and 67 keV.

Parameter Change	What Increases	What Does Not Increase
Increase mA	Photon fluence/output, heat	Endpoint energy, effective energy
Increase kVp	Endpoint energy, effective energy, output, penetration	Tube current
Add filtration	HVL, effective energy, mean energy	Maximum photon energy
Increase anode angle	Effective focal spot size, field coverage	Spatial resolution, heel effect severity
Decrease focal spot	Spatial resolution	Heat capacity

Linac Head and RF Chain

Component	Board Role	Trap
Magnetron	Generates microwaves directly; common in lower-energy machines.	Not an amplifier.
Klystron	Amplifies RF from a low-power source.	Does not generate microwaves from scratch.
Thyratron	High-speed switch/controlled rectifier for pulsed power.	Not the voltage step-up transformer.
Waveguide	Accelerates electrons with RF fields; standing or traveling wave.	Transmission waveguide carries RF to accelerator structure.
SF₆ gas	Insulates high-voltage/RF waveguide components because of high dielectric strength.	Not a coolant and not the accelerating medium.
Bending magnet	Directs electrons from accelerator to target/foil; energy analyzes/focuses.	Acts on electrons, not produced photons.
Target	Bremsstrahlung photon production.	Removed for electron mode.
Flattening filter	Flattens forward-peaked photon beam at a specified depth.	Removed in FFF and electron mode.
Scattering foil	Broadens electron beam.	Not used in photon mode.
Monitor chamber	Terminates dose by MU; checks symmetry/flatness.	Still present in electron mode.

Treatment Machine Family Tree

Machine	What It Accelerates/Uses	Exam Hook
Linear accelerator	Electrons accelerated in a straight waveguide	Photon mode uses target; electron mode removes target.
Betatron	Electrons induced by changing magnetic flux	Can produce electrons or photons with a target; mostly historical.
Microtron	Electrons recirculated through RF cavity	Energy gain per pass with magnetic recirculation.
Cyclotron	Charged particles in spiral path, fixed magnetic field	Relativistic mass/velocity limits classic fixed-frequency design.
Synchrotron	Charged particles in fixed ring	Magnetic field and RF vary with particle energy.
Gamma Knife	Many Co-60 sources	Co-60 average photon energy about 1.25 MeV.
Tomotherapy	Helical fan-beam 6 MV linac	Binary MLC, no flattening filter profile.
CyberKnife	Robotic compact linac	Non-isocentric beams, image guidance, often cones/MLC.

Collimation, MLC, and FFF

Modern MLC transmission is typically 1-3%. Interleaf leakage is between leaves; intraleaf leakage is through a leaf; tongue-and-groove reduces leakage but can make a narrow underdose band.
Dynamic MLC delivery is more sensitive to systematic leaf-position errors than static shaping. A 1 mm systematic error can matter.
FFF removes the flattening filter: dose rate rises, head scatter falls, out-of-field dose falls, spectrum softens, and the profile is peaked rather than flat.
FFF does not magically mean "higher energy." A nominal 6X FFF beam is often less penetrating than flattened 6X.
Rounded MLC leaf ends make penumbra more consistent as leaves move across the divergent beam. The price is partial transmission through the rounded tip, modeled with the dosimetric leaf gap.

MLC Error Language

Term	Meaning	Why It Is Tested
Dosimetric leaf gap	Modeling parameter for rounded leaf-end transmission and geometry.	Small gap errors can bias IMRT/VMAT dose.
Tongue-and-groove	Interlocking leaf design to reduce interleaf leakage.	Can cause narrow underdose if adjacent leaves expose a region sequentially.
Interleaf leakage	Leakage between adjacent leaves.	Important with high MU treatments.
Intraleaf transmission	Transmission through the leaf body.	Typical modern value around 1-3%.
Leaf positioning accuracy	Actual leaf vs planned leaf location.	Systematic errors matter more than isolated random errors.

Quantity Decoder

Quantity	Definition	Do Not Confuse With
Fluence	Particles crossing a sphere per cross-sectional area.	Planar fluence through a flat plane.
Energy fluence	Radiant energy per area; for monoenergetic beam = fluence x energy.	Fluence rate.
Kerma	Kinetic energy released per unit mass.	Dose, unless charged particle equilibrium holds.
Absorbed dose	Energy deposited per unit mass.	Equivalent/effective dose in Sv.
Exposure	Ionization in air from photons.	Not defined for electrons/protons or in tissue.

Detector Choice

Ion chambers are the reference workhorse: stable, near tissue equivalent after corrections, and NIST-traceable through calibration coefficients.
Diodes are sensitive to energy, temperature, field size, dose rate history, and angle. They are not meaningfully pressure-dependent like air chambers.
TLDs are read by heat. OSLDs are read by light. Radiochromic film is near tissue equivalent and self-developing; use net optical density.
Modern EPIDs are usually indirect detectors: scintillator plus photodiode array, commonly amorphous silicon.
Free-air ion chambers are standards-lab devices, not the usual meter for vault barrier surveys.

Detector	Strength	Board Weakness / Correction
Farmer chamber	Reference calibration in broad fields	Volume averaging in small fields/high gradients.
Parallel-plate chamber	Electron dosimetry, buildup/surface region	Guard ring, polarity, and calibration details matter.
Diode	High sensitivity, small fields	Energy, temperature, angular, dose-rate, and field-size dependence.
Diamond	Near tissue equivalent, small field utility	Detector-specific correction factors still needed.
Film	High spatial resolution planar dose	Scanner, orientation, calibration curve, net optical density.
Ion chamber survey meter	Barrier photon dose-rate surveys	Use appropriate energy range and pulsed-field response.
GM meter	Contamination/presence checks	Poor quantitative therapy barrier meter.

Reference Calibration

TG-51 is an absorbed-dose-to-water protocol for Co-60, MV photons, and electrons. It is not a brachytherapy source-strength protocol for Ir-192.

Item	Photon	Electron
Reference field	10 x 10 cm²	Usually cone/applicator field with adequate size for energy
Reference depth	10 cm water-equivalent depth	d_ref = 0.6R₅₀ - 0.1 cm
Beam quality	%dd(10)_x	R₅₀
Core correction	k_Q	k_Q plus gradient/replacement conventions
Corrected reading	M = M_rawP_TPP_ionP_polP_elec plus local leakage/timing corrections if needed.
ADCL calibration cycle	Ion chamber/electrometer calibration is usually kept on a 2-year cycle; annual machine output calibration is a separate clinic QA item.

Electron depth-ionization conversion: for I₅₀ in cm, R₅₀ = 1.029I₅₀ - 0.06 for I₅₀ <= 10 cm, and R₅₀ = 1.059I₅₀ - 0.37 for I₅₀ > 10 cm. Mean incident electron energy is E₀ = 2.33R₅₀.

TG-51 Correction Factor Triggers

Correction	Corrects For	Memory Hook
P_TP	Temperature/pressure effect on air density	Open-to-air ion chamber reading changes with weather.
P_ion	Ion recombination	More important for pulsed/high-dose-per-pulse beams.
P_pol	Polarity effect	Compare readings at opposite bias polarity.
P_elec	Electrometer calibration	Chamber-electrometer system factor.
k_Q	Beam quality conversion from Co-60 calibration quality	Photon quality uses %dd(10)_x; electron quality uses R₅₀.

Saturation curve logic: in the saturation region, collected charge should be nearly independent of bias voltage. If increasing voltage still meaningfully increases reading, recombination is not negligible. Pulsed beams need two-voltage or equivalent recombination correction discipline.

Depth-Dose and Scatter

Factor	Meaning	Board Directionality
PDD	Dose at depth divided by dose at dmax for same SSD setup.	Increases with energy, SSD, and field size; decreases with depth.
TMR	Dose at depth divided by dose at dmax with detector points at same source distance.	Independent of SSD; used for SAD/isocentric calcs.
S_c	Collimator/head scatter.	Depends on collimator setting.
S_p	Phantom scatter.	Depends on field size at phantom/patient.
S_cp	Total output factor.	Measured in phantom; often split into S_cS_p.
OAR	Off-axis ratio.	Profile change relative to central axis, not organ-at-risk in physics tables.

Percent Depth Dose Directionality

Change	PDD Effect	Why
Increase beam energy	Increases PDD, deeper dmax	More penetrating beam.
Increase field size	Increases PDD	More phantom scatter contributes dose at depth.
Increase SSD	Increases PDD	Inverse square difference between dmax and depth is reduced.
Increase depth	Decreases PDD after dmax	Attenuation and inverse square dominate.
Add low-density lung in path	Can reduce attenuation but perturb electron equilibrium	Simple equivalent path length can fail near interfaces.

PDD to TMR approximation: conceptually remove the SSD inverse-square effect from PDD. A common working form is TMR(d,r_d) approx PDD(d,r,SSD)/100 x [(SSD+d)/(SSD+d_max)]² x [S_p(r_dmax)/S_p(r_d)]. On a board problem, the point is usually directionality: TMR is an SAD quantity and is not the same thing as PDD/100.

Monitor Unit Patterns

First branch: SSD question uses PDD. SAD/isocenter question uses TMR/TPR. If the table hands you PDD and says 100 cm SSD, do not use TMR because it feels more modern.

Setup	Skeleton	Notes
SSD	MU = Rx / [O x S_c x S_p x modifiers x PDD/100 x ISF]	ISF corrects from calibration geometry to treatment point geometry.
SAD	MU = Rx / [O x S_c x S_p x modifiers x TMR]	Use dose per beam if fields are equally weighted.
Wedge pair	wedge angle = 90 - hinge angle/2	Hinge 135 degrees -> wedge 22.5 degrees.
Radiologic path	∑ density x thickness	3 cm lung at density 0.25 behaves like 0.75 cm water-equivalent path.
Field gap	gap ~ d(L₁/SSD₁ + L₂/SSD₂)/2	Half-beam block removes divergence on that side.

Dose Perturbation and Out-of-Field Rules

Situation	Fast Answer	Trap
Carbon-fiber couch in beam	Attenuates the beam; if unmodeled, delivered dose at target can be low for beams through couch.	It is a dosimetric effect, not just a collision issue.
Bone/high-Z interface	Can increase dose near entrance side from backscatter and perturb dose around exit side.	Direction and magnitude depend on energy, material, and geometry.
Metal implant	Causes CT artifact and dose perturbation; overrides/model-based algorithms help but do not remove uncertainty.	High density on CT is not automatically accurate material assignment.
Out-of-field near target	Mostly patient/internal scatter.	Shielding only at the head will not remove near-field patient scatter.
Out-of-field far from target	Head leakage and collimator scatter become relatively more important.	High-MU IMRT/VMAT can increase leakage contribution.

Special Beam Rules

Beam	Rules	Traps
Electrons	R_p ~ E/2, R₉₀ ~ E/3.2, practical treatment depth ~ E/3.	Surface dose increases with energy. Small cutouts, obliquity, and low energy can change PDD.
Electron field size	Need lateral scatter equilibrium; small fields reduce output and alter depth dose.	Cutout factor is not just area.
Bolus	Raises surface dose and shifts depth-dose curve toward surface.	Bolus does not increase penetration.
Protons	Finite range, Bragg peak, SOBP, RBE convention 1.1.	Distal edge has high LET/range uncertainty; motion interplay matters in PBS.
TBI	Spoiler increases surface dose; lung blocks reduce lung dose.	Long SSD lowers dose rate; AP/PA may be easier for weak patients.
TSEI	Large SSD, multiple fields/poses, high surface dose.	Soles, scalp, perineum, and skin folds need attention.

Electron Clinical Rules

Approximate therapeutic range: R₉₀ ~ E/3.2; practical range: R_p ~ E/2.
Minimum field dimension should generally exceed practical range for stable output/depth dose.
Obliquity increases surface dose and shifts dose proximally; it also broadens penumbra.
Air gaps reduce output and alter effective SSD; extended SSD corrections are electron-specific, not simple photon inverse square alone.
High-Z shielding in electron fields can make bremsstrahlung; use appropriate thickness and consider low-Z bolus/covering material near skin.

Proton Board Rules

Concept	High-Yield Rule	Trap
RBE	Clinical proton plans conventionally use RBE 1.1; photons/electrons use 1.0.	Carbon ion RBE is variable and model-dependent.
Range	Range is dominated by proton energy and patient stopping power. A 200 MeV proton has range on the order of 26 cm in water.	HU-to-stopping-power calibration drives range uncertainty.
Range uncertainty	Common mental model: distal/proximal margin includes a percent of range plus setup.	Range margin is along the beam direction, not a uniform 3D shell.
SOBP modulation	Wider modulation adds lower-energy components and generally increases entrance dose.	There is no photon-like buildup region for protons.
Lateral penumbra	Aperture systems can be sharp near surface, but multiple Coulomb scattering broadens penumbra with depth.	Distal edge and lateral edge are different uncertainty problems.
PBS interplay	Spot scanning plus moving anatomy can create hot/cold interplay.	Repainting, gating, compression, and robust optimization are mitigation tools.

TG-142 Linac QA Values To Know

Cadence	Test	Tolerance
Daily	Photon output constancy	3%
Daily/weekly	Electron output constancy	3%; daily for machines with unique electron monitoring needs.
Daily	Lasers	Non-IMRT 2 mm; IMRT 1.5 mm; SRS/SBRT 1 mm.
Daily	ODI	2 mm conventional/IMRT; 1 mm SRS/SBRT.
Monthly	Photon/electron output and backup monitor chamber	2%
Monthly	Light-radiation coincidence, symmetric jaws	2 mm or 1% per side.
Monthly	Dynamic MLC leaf positioning	About 1 mm; static positioning about 2 mm.
Annual	TG-51 output calibration	1% from baseline/expected.
Annual	Flatness/symmetry baseline change	1%
SRS	Overall localization/delivery accuracy	1 mm is the board number.

Protocol Index: What Each Report Is For

Protocol / Report	Use	Classic Exam Association
TG-51	Reference absorbed dose to water calibration	k_Q, %dd(10)_x, R₅₀.
TG-142	Medical accelerator QA	Daily 3%, monthly 2%, SRS tighter geometry.
TG-43	Brachytherapy dose calculation formalism	Air-kerma strength, dose-rate constant, g(r), F(r,θ).
TG-56	Brachytherapy QA	HDR source strength and positional checks.
TG-66	CT simulator QA	Laser/imaging plane/table alignment.
TG-100	Risk-based quality management	FMEA, process maps, fault trees, quality management by risk.
TG-101	SBRT recommendations	Steep gradients, image guidance, normal tissue constraints.
TG-119	IMRT commissioning test cases	Benchmarking planning/delivery accuracy.
TG-135	Robotic radiosurgery QA	CyberKnife-style delivery systems.
TG-203	Cardiac implantable electronic devices	Risk stratification by dose, pacing dependence, neutrons.
TG-218	IMRT measurement-based patient-specific QA	3%/2 mm, 95% tolerance, 90% action.
TG-224	Proton machine QA	Range, spot position, output, imaging/safety checks.
TG-302	Surface-guided radiotherapy	SGRT for setup, DIBH, frameless SRS, motion monitoring, and QA.
TG-307	EPID-based IMRT/VMAT QA	Pre-treatment and transit dosimetry with portal imagers.
MPPG 11.a	Plan and chart review	Minimum prudent physics support for plan/chart review.

Other Safety/QA Numbers

TG-218 commonly tested universal IMRT QA gamma: 3%/2 mm, global normalization, absolute dose, 10% threshold, 95% tolerance, 90% action.
TG-101 SBRT mindset: steep dose fall-off outside PTV is central to reducing toxicity.
MPPG 9.a/9.b SRS/SBRT annual end-to-end dosimetric tolerance for measured vs calculated dose is commonly 5%.
TG-203 cardiac implantable device risk: >5 Gy cumulative device dose and neutron-producing beams are high-risk flags.
MPPG 11.a cadence language: physics chart review should occur at regular intervals, practically within each block of about 5 treatment fractions, and at clinically meaningful transitions.
Small field definition is not simply "less than 10 x 10." Think source occlusion, loss of lateral charged particle equilibrium, and detector size relative to field.

Risk Analysis Language

Term	Meaning	Exam Hook
FMEA	Prospective failure modes and effects analysis	Used before an event to identify high-risk process steps.
RCA	Root-cause analysis	Reactive; performed after an event or near miss.
Process map	Step-by-step workflow diagram	Foundation for FMEA.
Fault tree	Top-down analysis of paths leading to failure	Starts with undesired event.
Statistical process control	Control charts and process drift detection	Monitors stability over time.
Not FMEA's job	Deciding which workflow steps are nonessential	FMEA scores failures in a defined process; it is not just a lean-process trimming tool.

Dose Reporting and Volumes

GTV = visible disease. CTV = microscopic risk. ITV = motion envelope/internal uncertainty. PTV = setup and delivery uncertainty margin. PRV = OAR plus uncertainty.
ICRU 83 shifted IMRT reporting away from a single point: use dose-volume reporting such as D_98% near-min, D_2% near-max, and D_50% median dose.
Conformity asks "does high dose match target?" Homogeneity asks "how even is target dose?" Gradient asks "how fast does dose fall?"

Metric	Meaning	Fast Interpretation
D_98%	Dose received by 98% of target	Near-minimum dose.
D_2%	Dose received by hottest 2% of target	Near-maximum dose.
D_50%	Median dose	Recommended central reporting metric for IMRT-style plans.
V₂₀	Volume receiving at least 20 Gy	Dose-volume threshold, common in lung.
D_0.03cc	Small-volume maximum-like dose	Used because true point max is unstable.

Dose Algorithms

Algorithm	Concept	Where It Breaks
Pencil beam	Fast kernel along ray lines.	Lung, interfaces, small fields.
Convolution/superposition	Models scatter transport better.	Still approximate in extreme heterogeneity.
Collapsed cone / AAA	Clinical workhorse for photons.	Less exact than deterministic/Monte Carlo at interfaces.
Acuros / deterministic	Solves transport equation with material assignment.	Material mapping and reporting mode matter.
Monte Carlo	Particle transport simulation.	Statistical uncertainty and commissioning burden.

Simulation, Motion, and Registration

Changing CT kVp can change the HU-density calibration. mAs mainly changes noise.
Aliasing occurs when sampling frequency is less than twice the highest spatial frequency in the object.
MIP is useful for bright/high-density moving targets such as many lung nodules; MinIP or phase-by-phase contouring may be needed for hypodense lesions.
MRI detects RF signal emitted by relaxing nuclei after excitation, not x-rays.
DIBH for breast reduces heart exposure and improves reproducibility, but it requires active patient participation.
CBCT is excellent for setup, but it is not continuous beam-on motion monitoring by itself.
Beam hardening makes dense structures appear with streak/cupping artifacts; partial-volume averaging blurs small high-contrast structures across voxel boundaries.

4DCT Image Sets

Image Set	Best For	Trap
MIP	Hyperdense/high-contrast moving targets, often lung nodules	Can miss hypodense liver/pancreas lesions.
MinIP	Hypodense structures/lesions when visible	Not a universal ITV tool.
Average CT	Dose calculation for many motion-managed plans	Blurs anatomy; not always adequate for contouring target extent.
Phase CT	Motion review and phase-specific contouring	One phase alone may underrepresent full motion.
Slow CT / free-breathing CT	Rough motion-inclusive anatomy in some settings	Less explicit than 4DCT for respiratory motion.

Registration and Adaptive Therapy

Concept	Meaning	Board Trap
Rigid registration	Translation/rotation only.	Cannot represent organ deformation or weight-loss anatomy change.
Deformable registration	Maps anatomy with spatially varying deformation vectors.	Useful but must be validated; it can create plausible-looking wrong anatomy.
Mutual information	Similarity metric often used for multimodality registration.	Does not require identical image intensity scale.
Online adaptive	Plan is adapted for the anatomy of the same treatment session before delivery.	Requires expedited contouring, dose calculation, QA, and approval workflow.
Offline adaptive	New plan is generated from accumulated interval imaging for future fractions.	Not delivered in the same session that triggered the change.
MR simulation / MR-linac	Excellent soft tissue; dose calculation still needs electron density assignment or synthetic CT.	MRI intensity is not HU.

Modulation and Data Flow

Inverse planning adjusts beamlet weights to minimize a cost function.
Increasing fluence smoothing usually reduces modulation complexity and monitor units.
VMAT can modulate MLC position, gantry speed, and dose rate; beam energy and radiation type are not continuously modulated during the arc.
DICOM objects contain tags with group/element numbers, value representation, value length, and value field. DICOM-RT moves images/plans/structures, but does not directly control the linac.
Machine log files are useful QA data, but they are not themselves standard DICOM radiotherapy objects. RTPLAN, RTSTRUCT, RTDOSE, RTIMAGE, REG, and waveform objects are the safer exam answers.
Pre-treatment IMRT QA in a phantom checks deliverability and dose agreement in that QA geometry; it does not prove patient setup, anatomy, collision clearance, or every transfer step.
FLASH is usually defined at ultra-high dose rates exceeding about 40 Gy/s.

Source-Strength Formalism

TG-43U1 specifies source strength by air-kerma strength rather than apparent activity. Dose-rate calculation separates source strength, dose-rate constant, geometry, radial dose function, and anisotropy.

Term	Meaning	Board Hook
S_K	Air-kerma strength	Unit U = cGy cm² h^-1.
Λ	Dose-rate constant	Converts source strength to dose rate at reference geometry.
G(r,θ)	Geometry factor	Inverse-square/source-length geometry.
g(r)	Radial dose function	Absorption and scatter along transverse plane.
F(r,θ)	Anisotropy	Angular variation, self-attenuation, source construction.

TG-43 vs model-based dose calculation: TG-43 assumes water-based conditions and does not fully model patient heterogeneity, applicator shielding, or finite patient scatter. Model-based brachytherapy algorithms address those effects; the reporting trap is to specify whether dose is reported to water or medium.

Source Timing and Dose Rate

Item	Number	Use
Ir-192	73.8 d, average energy ~0.38 MeV	HDR workhorse; treatment time increases as source decays.
I-125	59.4 d, ~28 keV	Permanent prostate/eye plaque seeds.
Pd-103	17 d, ~21 keV	Shorter half-life, higher initial dose rate for same total dose.
Cs-137	30 y, 662 keV	Historical LDR.
HDR definition	>12 Gy/h	ICRU dose-rate class.
LDR definition	0.4-2 Gy/h	Classic low dose rate range.

HDR treatment time scales inversely with current source strength: t_actual = t_nominal x S_nominal/S_actual.
Transit dose grows with number of fractions for the same total prescription because the source travels in and out each fraction.
Point A is classically 2 cm superior along the tandem and 2 cm lateral. The modern limitation is obvious: point-based prescription ignores patient anatomy.
Well chambers are used for HDR source strength checks; Farmer chambers are not the source-strength tool.
HDR source positioning accuracy is a millimeter-level issue, not a casual 3 mm issue. Full calibration language uses +/-1 mm source positioning accuracy.

IORT and Special Brachy Modalities

Modality	Typical Technology	Board Hook
Electron IORT	Mobile/dedicated electron unit, often up to about 12 MeV.	Sharp practical range; shielding and applicator alignment dominate setup.
Low-kV IORT	Miniature x-ray source around 50 kV.	Rapid dose fall-off; high surface/contact dose.
HDR IORT	Ir-192 afterloader with surface applicator/flap or interstitial catheters.	Source path, transfer tubes, dwell positions, and emergency procedures matter.
Electronic brachytherapy	Electrically generated low-energy x-rays.	No radioactive source decay, but output/source QA still required.

Source QA and Directive Traps

Scenario	Know This	Exam Trap
HDR written directive	Includes patient/site, radionuclide, dose, fractionation, and treatment site/volume before treatment.	Patient exposure rate for release is not an HDR written directive component.
HDR source strength	Measured with a well-type ionization chamber and independently checked.	Not a Farmer chamber measurement.
HDR interlocks/emergency	Check door, radiation monitor, applicator/transfer guide tube integrity, emergency retraction, timer.	Emergency procedures are not "annual-only" in practical programs; know daily/quarterly/annual categories from TG-56-style QA.
Manual permanent seeds	Assay a sample or all sources per institutional/regulatory expectations; compare source strength to vendor certificate.	Air-kerma strength is the specified source strength quantity.
Permanent implant follow-up	Post-implant evaluation of total source strength outside the treatment site is a 60-day regulatory timing concept.	This is separate from the pre-treatment written directive.
Applicator diameter	For same prescription depth, larger cylinder diameter usually lowers mucosal surface dose.	Prescription depth fixed does not mean surface dose fixed.
Point A	Classically 2 cm superior to flange/fornix along tandem and 2 cm lateral.	Point A prescription ignores patient-specific anatomy; modern image-guided brachy uses HR-CTV/OAR DVH metrics.

Barrier Design Logic

Quantity	Meaning	Board Default
W	Workload	Dose delivered per week at 1 m or isocenter reference.
U	Use factor	Fraction of primary beam directed at barrier.
T	Occupancy factor	Fraction of time a person is in the protected area.
P	Design goal	Common weekly values: uncontrolled 0.02 mSv/wk, controlled 0.1 mSv/wk.
TVL	Tenth-value layer	n = log₁₀(1/B) TVLs.

Primary barrier: B = P d²/(WUT).
Secondary barrier includes leakage and scatter. Leakage workload can be much higher for IMRT; a common board range is C = 2-10 times primary workload.
For a 6 MV linac, neutron shielding is not the design driver. Neutrons become a concern for high-energy photon beams above about 10 MV.
Use an ion chamber survey meter for photon dose rate beyond a shielded barrier. GM meters are not the right quantitative choice for pulsed therapy barriers.

Shielding Question Pattern

If This Changes...	Barrier Thickness Direction	Reason
Distance to occupied area increases	Decreases	Inverse square reduces required transmission.
Occupancy factor increases	Increases	More time in area requires more shielding.
Use factor toward a primary barrier increases	Increases	More beam directed at that barrier.
Workload increases	Increases	More dose delivered per week.
Area changes from controlled to uncontrolled	Increases	Lower public design goal.
Photon energy increases	Usually increases	Higher TVL, and neutrons above high-energy thresholds.

Dose Limits and Regulated Events

Limit / Requirement	Board Number
Adult occupational TEDE	50 mSv/y (5 rem/y)
Public dose limit	1 mSv/y (0.1 rem/y)
Declared pregnant worker embryo/fetus	5 mSv for entire pregnancy, with effort toward uniform monthly exposure.
Individual monitoring threshold	Likely to exceed 10% of applicable limit.
Patient release after radioactive material	TEDE to any other individual not likely to exceed 5 mSv.
Survey meter calibration	Before first use, annually, and after repair affecting calibration; cannot use if indicated vs calculated exposure rate differs by more than 20%.
NRC medical event notification	NRC telephone notification no later than the next calendar day; written report within 15 days.
Medical event patient notification	Notify referring physician and patient no later than 24 h after discovery, with the regulatory physician-judgment exceptions.
Regulatory scope	NRC regulates byproduct/source material; state programs generally regulate linac x-ray production.
DOT transport index	Maximum radiation level in mrem/h at 1 m from package surface.

CTDIvol estimates scanner output for a standardized phantom; DLP = CTDIvol x scan length.
Increasing mAs decreases image noise but increases dose roughly linearly.
Increasing kVp increases penetration, changes contrast, and can alter HU-density calibration.
Tube current modulation reduces unnecessary dose by adjusting mA to patient attenuation.
Spatial resolution is about small objects/high spatial frequency; low-contrast resolution is about seeing subtle density differences in noise.
MR-linac advantage: soft-tissue visualization and adaptive anatomy assessment. It is not useful because magnetic fields make oxygen molecules radiosensitizers.

Modality / Concept	High-Yield Point	Trap
CT number	HU = 1000(μ - μ_water)/μ_water	HU-density table is protocol/energy dependent.
Nyquist	Sampling frequency must be at least twice the highest object spatial frequency.	Aliasing occurs when sampling is too low, not too high.
Window/level	Window = displayed HU range; level = center of range.	Display setting does not change underlying image data.
MRI signal	RF signal from precessing/relaxing nuclei after excitation.	Imaging coil detects emitted RF, not the transmitted RF itself.
Geometric distortion MRI	Gradient nonlinearity, B0 inhomogeneity, susceptibility, chemical shift.	Important for MR simulation and fusion.
PET SUV	Activity concentration normalized by injected activity/body habitus.	Uptake is not a direct radiation dose measurement.
PET spatial resolution	Limited by detector size, positron range, non-collinearity, depth-of-interaction, and reconstruction.	511 keV annihilation photons are emitted nearly, but not perfectly, 180 degrees apart.
CT artifacts	Beam hardening, metal streaks, motion, photon starvation, partial volume.	Artifacts can corrupt HU-to-density conversion.
SSDE	Size-specific dose estimate adjusts CTDIvol for patient size.	Still an estimate of scanner output/dose, not patient organ dose.

Cram Tables

Radionuclide Snapshot

Nuclide	Half-Life	Emission / Energy	Use
Co-60	5.27 y	1.17, 1.33 MeV gamma; avg 1.25 MeV	Teletherapy/Gamma Knife historical and current.
Cs-137	30.1 y	662 keV gamma	Historical LDR.
Ir-192	73.8 d	Average ~0.38 MeV gamma	HDR/PDR afterloading.
I-125	59.4 d	Low-energy photons ~28 keV	Permanent seeds.
Pd-103	17 d	Low-energy photons ~21 keV	Permanent seeds, faster dose delivery.
Sr-90/Y-90	28.8 y / 64 h	Beta	Ophthalmic/surface historical; Y-90 therapy.
Tc-99m	6 h	140 keV gamma	Imaging, not therapy.

Question-Type Map

Cluster	What To Master	Typical Trick
Basics and decay	SI units, constants, decay modes, equilibrium, activity math.	Unit conversion before formula.
Machines	X-ray tube, filtration, Co-60, linac modes, RF, MLC, FFF.	Component that does not change when switching modes.
Interactions	Photoelectric/Compton/pair, stopping power, neutrons, protons.	Energy and material decide the answer.
Measurements	Fluence, kerma, dose, TG-51, detector behavior.	Which detector is pressure/temperature/energy dependent?
External beam calcs	PDD/TMR, scatter factors, equivalent square, wedges, gaps, heterogeneity.	Using the wrong table or forgetting dose per beam.
QA and planning	TG-142, TG-218, SRS/SBRT, ICRU 83, motion, DICOM.	Protocol number tied to a tolerance.
Brachy/protection	TG-43, air-kerma strength, HDR timing, shielding, NRC release/events.	Source strength vs activity; public vs release dose.

Number Snapshot

Category	Numbers
Constants	c = 3 x 10⁸ m/s; h c = 1240 eV nm; m_ec² = 0.511 MeV; 1 amu = 931.5 MeV; proton/electron mass ratio 1836.
Photon interactions	Pair threshold 1.022 MeV; Co-60 average 1.25 MeV; Cs-137 662 keV.
Decay	T_1/2 = 0.693/λ; mean life 1.44 T_1/2; 1 Ci = 3.7 x 10¹⁰ Bq.
Electrons	R₅₀ ~ E/2.33; R_p ~ E/2; R₉₀ ~ E/3.2.
Protons	Clinical RBE 1.1; 200 MeV range in water about 26 cm; range uncertainty is beam-direction/stopping-power driven.
Linac QA	Daily output 3%; monthly output 2%; annual calibration 1%; SRS localization mentality 1 mm.
IMRT/SBRT	TG-218 3%/2 mm, 95% tolerance, 90% action; SRS/SBRT annual E2E dose agreement often 5%.
Brachy	HDR >12 Gy/h; Ir-192 73.8 d; I-125 59.4 d; Pd-103 17 d; HDR source position +/-1 mm; Point A 2 cm up, 2 cm lateral.
Protection	Occupational 50 mSv/y; public 1 mSv/y; pregnancy embryo/fetus 5 mSv; patient release 5 mSv; survey meter annual/after repair.
Events	NRC medical event phone report by next calendar day; written report 15 d; patient notification generally 24 h after discovery with physician exceptions.

Last-Pass Checklist

What is the quantity? Dose, kerma, exposure, fluence, activity, equivalent dose?
Are the units already compatible? Seconds vs days, MHz vs Hz, cm vs m, cGy vs Gy?
What geometry is implied? SSD/PDD, SAD/TMR, inverse square, field size at collimator vs phantom?
What energy range is implied? kV photoelectric, MV Compton, high-MV neutron, proton distal edge?
What protocol applies? TG-51, TG-142, TG-43, TG-218, NCRP 151, NRC Part 20/35?
Did the question ask for the exception?

Public Source Trail

Mutable protocol and regulatory details were checked against public standards and reference pages. This page is a review aid, not a substitute for the current full reports, institutional policy, or clinical physicist review.

AAPM TG-51 / NIST citation page: TG-51 protocol; AAPM addendum: TG-51 photon addendum.
AAPM reports: TG-142, TG-218, TG-101, TG-224, TG-43U1, TG-106.
Additional AAPM public pages used for the second pass: TG-100, TG-119, TG-203, TG-302, TG-307, MPPG 11.a, TG-186, and AAPM Report 23.
NRC/eCFR: 10 CFR Part 20; 10 CFR Part 35; NRC medical use FAQ for patient release under Part 35: Part 35 FAQ.
ICRU radiation therapy report overview: ICRU radiation therapy reports.