Module 05 showed how a small air cavity reports dose to the water around it: charge over air mass, times W/e, times a stopping power ratio. TG-51, the AAPM protocol for reference dosimetry of megavoltage photon and electron beams, keeps that physics but hides most of it. The lab measures the chamber in cobalt directly in dose to water, so W/e and the stopping power ratio never appear explicitly. What remains for the clinic is to correct the reading and to apply k_Q, the ratio of the chamber's dose-to-water response in cobalt to its response in the clinical beam.

Three answers carry the module:

1. Every P factor corrects the reading toward what an ideal chamber at reference air conditions would have collected. Most sit within a percent of 1.000; P_TP can move by several percent.
2. Beam quality picks k_Q. Photons use %dd(10)_x; electrons use R₅₀.
3. Different jobs need different detectors. Ion chambers for reference dose, diodes and diamonds for small fields and in vivo checks, film for planar distributions, survey meters for protection.

Every gas-filled detector is the same object: two electrodes with gas between them and a voltage across it. Radiation frees ion pairs in the gas, and the field drags them to the electrodes. What changes from one instrument to the next is only the voltage, and that changes what happens to the ions on the way.

At very low voltage, ions drift slowly and many recombine before they arrive, so the signal is incomplete. Raise the voltage and collection becomes complete: every ion pair formed is collected and no more. That plateau is the ionization chamber region, and it is where dosimetry lives, because the charge is proportional to the energy deposited. Raise it further and the freed electrons gain enough energy between collisions to ionize the gas themselves, multiplying the charge by a fixed gain: the proportional region. Higher still, each event triggers an avalanche along the whole anode and every pulse is the same size, regardless of what started it: the Geiger-Müller region. Beyond that, the gas discharges continuously and the detector is useless.

A clinical ion chamber is a small, open-to-air cavity with a thin conducting wall and a central collecting electrode, connected by a triaxial cable to an electrometer, a very sensitive charge meter. A Farmer chamber holds about 0.6 cm³ of air and collects roughly 20 nC per gray. The bias voltage, typically 300 V or less, is chosen to sit on the plateau of region II.

Chamber shape follows the job. A cylindrical (thimble) chamber is robust and nearly isotropic, so it is the reference chamber for photons. A parallel-plate chamber has a thin entrance window and a narrow gap, so it can measure close to a surface and in steep gradients: buildup region, low-energy electron beams. A well chamber surrounds a brachytherapy source and measures its air kerma strength (Module 15). Small-volume chambers (0.01 to 0.1 cm³) reduce volume averaging in small fields at the cost of signal.

Where the chamber "measures"

A cylindrical chamber replaces a cylinder of water with air. Electrons reaching the cavity come mostly through its upstream wall, so the chamber reads as if it sat slightly upstream of its centre. For depth-dose curves, physics shifts the curve by 0.6 r_cav upstream for photons and 0.5 r_cav for electrons, where r_cav is the cavity radius. For a parallel-plate chamber, the point of measurement is the inner face of the entrance window. In TG-51 photon reference dosimetry, the chamber centre goes at 10 cm and no shift is applied, because k_Q already contains the gradient effect.

Temperature and pressure, P_TP

The chamber is open to the room, so the mass of air in it changes with the weather. Warmer air is less dense, so fewer ion pairs form for the same dose and the raw reading is low. Lower pressure does the same. P_TP scales the reading to the air density at TG-51 reference conditions, 22 °C and 101.33 kPa (760 mmHg). Its shape is the ideal gas law: density goes as P/T, so the correction goes as T/P.

Ion recombination, P_ion

Even on the plateau a small fraction of ions recombine. Initial recombination happens within one particle's track. General recombination happens between ions from different tracks, and it grows with the density of ions, which in a pulsed linac means dose per pulse. Raising the voltage collects faster and loses less, so the loss can be measured by reading at two voltages. For a pulsed beam, with V_H the normal voltage and V_L at most half of it:

Polarity, P_pol

Reversing the bias should give the same magnitude of charge. It does not quite, because the beam itself deposits or knocks out charge in the collector, stem and cable, and that stray charge adds with one polarity and subtracts with the other. Averaging the two readings removes it.

Electrometer and leakage

P_elec is the electrometer's own calibration factor, set to 1.000 when the lab calibrates chamber and electrometer as one unit. P_leak corrects for charge collected with no beam; the photon addendum asks that leakage stay under 0.1% of the reading.

TG-51 (Almond et al., 1999), updated for photons by the 2014 addendum (McEwen et al.), defines one measurement: dose to water at 10 cm depth on the central axis, 10 × 10 cm field, at 100 cm SSD or 100 cm SAD. Ten centimetres is chosen because contaminant electrons from the head have died out there, while the chamber is still deep in a region with a gentle gradient.

Why k_Q exists

The chamber's calibration coefficient is measured in cobalt-60. In a 6 or 18 MV beam the water-to-air stopping power ratio is slightly smaller and the chamber's wall and cavity perturb the fluence slightly differently. k_Q is the ratio of the chamber's calibration coefficient in your beam to its coefficient in cobalt. It is below 1 and falls as energy rises, by a few percent at 18 MV. It is chamber specific, which is why the protocol tabulates it per chamber model.

Beam quality: %dd(10)_x

To look up k_Q you need a number that describes the beam's penetration. TG-51 uses %dd(10)_x: the percent depth dose at 10 cm for a 10 × 10 cm field at 100 cm SSD, for photons only. The subscript x matters. At 10 MV and above, electrons from the head add dose near d_max, which raises the denominator and lowers the measured %dd(10). To remove their effect, the depth dose is measured with a 1 mm lead foil in the beam, which replaces the head's contaminant electrons with a smaller, known contribution, and then converted.

With %dd(10)_x in hand, k_Q for a reference-class chamber comes from the addendum's fit, valid from 63% to 86%:

Electron depth doses have no long exponential tail to sit in. The dose rises, peaks, and falls steeply to the practical range within a few centimetres, and the electron spectrum changes with depth, so the stopping power ratio changes too. TG-51 handles this by measuring at a depth tied to the beam's own range.

Beam quality: R₅₀

R₅₀ is the depth in water where dose falls to 50% of its maximum, measured at 100 cm SSD with a field of at least 10 × 10 cm (20 × 20 cm when R₅₀ exceeds 8.5 cm). An ion chamber measures ionization, not dose, so physics first finds I₅₀, the depth of 50% ionization, and converts:

Original protocol (1999)

The 1999 protocol split k_Q into three pieces. k_ecal is a fixed, chamber-specific photon-to-electron conversion at one reference electron quality. k′_R50 carries it to the actual R₅₀. P_gr^Q is a gradient correction for cylindrical chambers, measured as the ratio of readings with the chamber centre at d_ref + 0.5 r_cav and at d_ref. Parallel-plate chambers were required for low energies (R₅₀ ≤ 4.3 cm, about 10 MeV and below) and need no gradient correction.

2024 electron addendum

The AAPM working group on TG-51 published an electron addendum in 2024 (Medical Physics, WGTG51 Report 385). It allows calibrated reference-class cylindrical chambers in all electron beams, 4 to 22 MeV, with no separate gradient correction: the gradient effect is folded into a single Monte Carlo-based k_Q that depends on R₅₀. It reports that dose per MU may come out up to about 2% higher than with the 1999 protocol. Learn the 1999 structure for the exam, because older questions use it, and know that your clinic may now calibrate electrons with a Farmer chamber at every energy.

Your chamber's N_D,w means something only because it ties back to a primary standard. In the United States, NIST realizes absorbed dose to water in a cobalt beam with a water calorimeter: a gray of dose raises the temperature of water by only about 0.24 mK, and that temperature rise, with the specific heat of water, is dose by definition. NIST calibrates the reference chambers of the Accredited Dosimetry Calibration Laboratories (ADCLs). The ADCL calibrates your chamber and electrometer in its cobalt beam. TG-51 then carries the coefficient to your linac.

TG-51 requires the chamber and electrometer to have been calibrated within the previous two years. With reference-class equipment and full care, the photon addendum puts the combined standard uncertainty in dose at the reference point at about 0.9%; with questionable equipment and typical shortcuts it is about 2%.

Between annual calibrations, TG-142 sets output tolerances: 3% on daily constancy checks, 2% on monthly checks against the baseline, and 1% for the annual reference calibration. A daily device that reads 3.5% high is a reason to hold treatment and call physics, not to treat and note it.

Ion chambers are accurate because air is predictable, but air is thin and gives little signal per volume. Solid detectors trade some of that predictability for size and sensitivity. Each one fails in a characteristic way, and the failure follows from its material.

Diodes

A silicon diode is a solid-state ion chamber. Radiation frees electron-hole pairs in the depletion region and the built-in field collects them with no bias voltage. Silicon is nearly 2000 times denser than air and needs only about 3.6 eV per pair instead of 34 eV, so per unit volume a diode gives roughly 18,000 times the signal of an air chamber. That makes it tiny and fast, ideal for small fields, scanning and in vivo checks. Its weaknesses come from silicon itself: Z = 14 over-responds to low-energy scatter; sensitivity changes with temperature (a few tenths of a percent per °C), with dose per pulse (so with SSD), with beam angle, and falls with accumulated dose as the crystal is damaged. In vivo diodes carry a buildup cap, so they report dose near d_max, not at the skin.

MOSFETs

A MOSFET stores charge in its oxide layer, which shifts its threshold voltage in proportion to dose. It is very small, reads out immediately, and is used in vivo, but it has a finite lifetime and modest precision.

TLD and OSLD

Both are crystals that trap electrons in defect states and release them, with light, when stimulated. TLD (thermoluminescent dosimeter, usually LiF:Mg,Ti, effective Z about 8.2) is heated; the light versus temperature is the glow curve. LiF is close to tissue, becomes supralinear above roughly 1 Gy, and reading it erases most of the signal. OSLD (optically stimulated luminescence, Al₂O₃:C, effective Z about 11.3) is stimulated with light instead of heat, can be re-read because each readout takes only a small fraction of the signal, and over-responds several-fold to kilovoltage photons because of its higher Z. OSLDs are the badge dosimeters for staff and the mailed audit dosimeters for IROC.

Film

Radiographic film (silver halide) needs chemical processing and over-responds strongly at low energy. Radiochromic film (for example EBT3 or EBT4) darkens by polymerization with no processing, is close to tissue equivalent, has fine spatial resolution, and is read on a flatbed scanner after the colour has stabilized, usually about a day. It is the standard for planar dose checks and small-field profiles; its weaknesses are scanner non-uniformity and orientation dependence.

Diamond and scintillators

Synthetic diamond detectors (carbon, Z = 6) are near water equivalent, very small and radiation hard, and are favoured for small-field output and profiles. Plastic scintillators are the most water-equivalent detectors in use: the plastic matches water in electron density and stopping power, so they barely perturb the field and show almost no energy dependence. Their one artifact is Cherenkov light produced in the optical fibre, which must be subtracted.

A survey meter answers a different question from a dosimeter: is there radiation here, and how much per hour? The choice depends on whether you need sensitivity or accuracy.

A Geiger-Müller counter works in region V. One ion pair triggers a full avalanche, so it detects almost anything, which makes it the tool for finding contamination, a lost seed, or a leak. Every pulse is the same, so it measures counts, not energy, and its exposure-rate reading is reliable only for the energy it was calibrated at. After each pulse the tube is dead for roughly 100 μs while the avalanche clears. In a very intense field the tube spends all its time dead and can read low or even zero; this is paralysis, and it is the reason a GM counter is never the only meter used near a high-activity source.

An ion chamber survey meter works in region II with a large, often pressurized air volume. It is far less sensitive, but its response is nearly flat with energy and it reads exposure rate accurately in high fields. It is the instrument for shielding surveys of a linac vault and for measuring dose rate at 1 m from a patient before release.

A scintillation detector (for example NaI(Tl)) converts gamma energy to light. It is the most sensitive photon survey instrument and can identify isotopes by energy.

Neutrons carry no charge and ionize nothing directly. A neutron detector therefore surrounds a proportional counter filled with BF₃ or He-3 with polyethylene. The hydrogen in the moderator slows fast neutrons to thermal energies, where boron-10 or helium-3 captures them and releases charged particles that the counter sees. Shaped correctly, the moderator makes the response follow dose equivalent, which is why this is called a rem meter. Bubble detectors are passive alternatives. Neutron surveys are needed around vault doors and mazes for beams above about 10 MV (Module 21).

This is the photon calibration your physicist runs once a year, with every factor visible. The defaults are a plausible 6 MV calibration at 100 cm SSD. Change the room conditions, the chamber or the readings and watch which factors move and how much the final cGy per MU moves with them.

Work each one on paper first. The constants are given in the problem, as they are on the exam.

You will not run TG-51. You will sign plans whose every MU depends on it, read in vivo results, order film QA, and stand next to survey meters. These are the numbers and the questions that make those moments productive.

Ballpark numbers to carry in your head

Choosing a detector for a clinical question

Is the TSET or chest wall skin dose right: OSLD or TLD on the skin, with no buildup cap, read against a calibration done under similar conditions. Is the pacemaker dose acceptable: OSLD or diode near the device, per the institutional policy for cardiac devices. Does an SRS output factor look odd: ask which detector measured it and whether the small-field correction was applied. Did the IMRT QA pass: the answer depends on the array or film, its resolution, and the gamma criteria (Module 11).

Sources

1. American Board of Radiology. Study Guide for Medical Physics for Radiation Oncology (2023 domain ranges; domain II.B, radiation measurement and calibration).
2. Almond PR et al. AAPM's TG-51 protocol for clinical reference dosimetry of high-energy photon and electron beams. Med Phys 1999;26:1847-1870.
3. McEwen M et al. Addendum to the AAPM's TG-51 protocol for clinical reference dosimetry of high-energy photon beams. Med Phys 2014;41:041501.
4. Muir B et al. AAPM WGTG51 Report 385: Addendum to the AAPM's TG-51 protocol for clinical reference dosimetry of high-energy electron beams. Med Phys 2024;51(9):5840-5857.
5. Klein EE et al. Task Group 142 report: Quality assurance of medical accelerators. Med Phys 2009;36:4197-4212.
6. IAEA Technical Reports Series 398. Absorbed Dose Determination in External Beam Radiotherapy. Vienna, 2000, and its 2024 revision.
7. IAEA TRS-483. Dosimetry of Small Static Fields Used in External Beam Radiotherapy. Vienna, 2017.
8. Khan FM, Gibbons JP. The Physics of Radiation Therapy, 5th ed. Chapters on measurement of ionizing radiation and dosimetry instruments.
9. Attix FH. Introduction to Radiological Physics and Radiation Dosimetry. Wiley, 1986. Chapters on ionization chambers and integrating dosimeters.
10. US Nuclear Regulatory Commission. 10 CFR 35.61, Calibration of survey instruments; 10 CFR 35.630, Dosimetry equipment.

The worksheet applies the published photon formalism to invented readings. Clinical numbers come from your physicist's calibration, the chamber's ADCL report and your machine's commissioning data.