Stereotactic Radiosurgery · Technical Foundations

Small-Field Physics & Dosimetry

Why measuring a tiny, sharp beam is hard — and how getting it wrong propagates

Radiosurgery uses photon fields with at least one dimension below a few centimeters, where the comfortable assumptions of conventional dosimetry fail: charged-particle equilibrium is lost, the radiation source is partially occluded, and detectors perturb the very field they measure. A surgeon does not commission a beam, but understanding why small-field dosimetry is error-prone explains the QA culture around radiosurgery and why a commissioning mistake can silently bias every treatment.

Timoteo Almeida, MD, PhD | Department of Neurological Surgery

Orientation

Three conditions define a "small field," per AAPM Task Group 155: loss of lateral charged-particle equilibrium, partial occlusion of the photon source by the collimator, and a detector that is large relative to the field. Each distorts measurement, and because radiosurgery delivers ablative dose with steep gradients and little margin, a systematic dosimetric error is not averaged away — it scales the dose for every patient treated on that beam model. This is why commissioning, detector choice, and independent checks matter disproportionately in radiosurgery.

Part I

Why Small Fields Misbehave

1.The three small-field problems

As field size shrinks below the lateral range of secondary electrons, charged-particle equilibrium is lost on the central axis and output drops in a way that simple models mispredict. The finite focal spot of the source becomes partially occluded by the collimator, further reducing and broadening the beam. And any detector with a sensitive volume comparable to the field averages dose across a steep gradient and perturbs fluence through density mismatch. The combined result is that output factors and small-field profiles are easy to mis-measure, and errors here bias the absolute dose and the penumbra used in every plan.

2.Detectors and the TG-155 framework

No single detector is ideal across all small fields, so a detector-specific correction framework is required. In practice, small-volume detectors are preferred: microdiamond detectors, small-volume ion chambers, unshielded diodes, radiochromic film, and gel/array systems, each with characteristic over- or under-response that must be corrected. TG-155 (with the IAEA/AAPM small-field code of practice) defines small-field conditions, recommends suitable detectors, and standardizes relative-dosimetry measurements (output factors, PDD/TMR, profiles); it is the reference standard for commissioning radiosurgical beams.

Part II

Beams, Output, and Calculation

3.Flattening-filter-free beams and dose rate

Modern linac radiosurgery commonly uses flattening-filter-free (FFF) beams, which remove the flattening filter to deliver a much higher dose rate (shortening treatment and reducing intrafraction motion exposure) at the cost of a forward-peaked profile that the planning system must model accurately. Dedicated cobalt systems instead rely on tabulated relative dose data for their fixed collimators. Either way, the planning system's beam model must reproduce the measured small-field output, or the delivered dose will systematically differ from the intended dose.

4.Dose calculation in small fields and heterogeneity

Dose-calculation algorithm matters more in radiosurgery than in bulk targets. Simple pencil-beam algorithms can mis-model lateral electron transport in small fields and at tissue/air or tissue/bone interfaces — relevant near the skull base, air cavities, and lung for body SBRT. Monte Carlo and modern model-based algorithms (e.g., linear Boltzmann transport solvers) handle these conditions far better and are preferred for small-field and heterogeneous geometries. A plan that looks acceptable under an inadequate algorithm can deliver a meaningfully different dose at an interface.

A commissioning error is invisible and universal Unlike a setup error that affects one fraction, a small-field commissioning error — a mis-measured output factor, an inadequate detector correction, or a poorly modeled FFF profile — biases every patient treated on that beam and is not visible on any individual plan. This is the deep reason radiosurgery depends on meticulous physics commissioning, independent dose checks, and external audit, covered in the QA section of the planning and constraints page.

5.Plan-quality metrics the clinician actually reads

Beyond raw dose, every radiosurgical plan is judged by a small set of geometric metrics. Coverage is the fraction of the target receiving the prescription (typically aimed at ≥ 95–99%). Conformity — usually the Paddick conformity index — measures how tightly the prescription isodose wraps the target, with an ideal value approaching 1. The gradient index (Paddick: the ratio of the half-prescription-isodose volume to the prescription-isodose volume) quantifies how fast dose falls off outside the target; a steeper gradient (lower index, often < ~3) is what spares adjacent brain and keeps V12Gy small. A central practical difference between platforms is the prescription isodose: dedicated cobalt systems prescribe to a low isodose (~50%), producing a hot, steep-gradient interior, whereas linac plans typically prescribe to a higher, more homogeneous isodose (~70–90%). Neither is inherently superior — they reflect different design philosophies, and both can produce highly conformal plans.

Common radiosurgical plan-quality metrics and their targets (representative; institution- and platform-dependent).
Metric	Definition (in brief)	Typical goal
Coverage	Fraction of target volume receiving the prescription dose	≥ ~95–99%
Paddick conformity index	How tightly the prescription isodose conforms to the target	Approaching 1.0
Gradient index	Half-prescription isodose volume / prescription isodose volume	Low (often < ~3); steeper is better
Prescription isodose	Isodose line to which the dose is prescribed	~50% (cobalt) vs ~70–90% (linac)
Maximum dose / hotspot	Point-maximum dose within the target	Accepted (cobalt) or limited (linac), by design

Part III

What the Clinician Should Take Away

5.The clinically relevant residue

The treating physician does not measure beams, but should understand that: small-field dosimetry is a recognized source of systematic error; the program's commissioning, detector selection (per TG-155), and independent checks are what make the prescribed dose trustworthy; algorithm choice affects dose near interfaces; and questions about an unexpected plan (steep dose near bone or air, a suspiciously high output) are legitimately directed to medical physics. The relationship between clinician and physicist is not a formality in radiosurgery — it is part of the safety system.

Key points

Small fields (TG-155) violate three assumptions: loss of lateral charged-particle equilibrium, source occlusion, and detector-size averaging — all biasing output and penumbra.
No universal detector; use small-volume detectors (microdiamond, small ion chamber, diode, film) with detector-specific corrections per TG-155 and the IAEA/AAPM small-field code of practice.
FFF beams give high dose rate (shorter treatment, less intrafraction exposure) but a peaked profile the model must capture; cobalt systems use tabulated collimator data.
Monte Carlo / model-based algorithms outperform pencil-beam in small fields and at bone/air interfaces (skull base, lung SBRT); algorithm choice changes interface dose.
A commissioning error is invisible on individual plans and affects every patient — the reason radiosurgery depends on rigorous physics, independent checks, and audit.

References

Das IJ, Francescon P, Moran JM, et al. Report of AAPM Task Group 155: megavoltage photon beam dosimetry in small fields and non-equilibrium conditions. Med Phys. 2021;48(10):e886–e921. PubMed
Benedict SH, Yenice KM, Followill D, et al. Stereotactic body radiation therapy: the report of AAPM Task Group 101. Med Phys. 2010;37(8):4078–4101. PubMed
International Atomic Energy Agency / AAPM. Dosimetry of small static fields used in external beam radiotherapy (TRS-483 code of practice). 2017. IAEA

Educational synthesis for neurosurgery and radiation-oncology trainees; physics overview for clinical context, not a commissioning reference. Dosimetry references verified against PubMed or official IAEA/AAPM records during review.