Stereotactic Radiosurgery · Technique

Gamma Knife Workflow

From indication and stereotactic reference to shot construction, plan review, and treatment delivery

A practical guide to how a Gamma Knife case is built. The emphasis is not on memorizing button sequences, but on preserving the clinical intent through every handoff: immobilization, imaging, contouring, planning, verification, delivery, and follow-up.

Timoteo Almeida, MD, PhD | Departments of Neurological Surgery and Radiation Oncology

Orientation

A Gamma Knife plan begins before the frame or mask is applied. The team must know the target, the clinical endpoint, the structures that can limit dose, and whether one fraction is actually the right treatment. Every subsequent step exists to preserve that intent in stereotactic space. A beautiful dose distribution on the wrong image set, through an unstable reference, or for a poorly chosen indication is still a failed plan.

Part I

Choose the Treatment Before the Hardware

1.Define the clinical endpoint

Start by stating what success means for this patient: durable control of a benign tumor, local control of a metastasis, AVM obliteration, pain relief in trigeminal neuralgia, or creation of a functional lesion. That sentence determines the prescription, target definition, acceptable heterogeneity, follow-up interval, and tolerance for latency or toxicity. Confirm prior radiation, systemic therapy, anticoagulation, implanted devices, anesthesia needs, and the imaging required to see both the target and its limiting anatomy.

2.Frame, mask, or another platform

A rigid stereotactic frame provides a direct mechanical reference and remains especially valuable for single-session vascular and functional indications, very small targets, long treatments, and patients in whom motion cannot be tolerated. Icon and Esprit also support mask-based treatment with stereotactic cone-beam CT and intrafraction motion monitoring, enabling fractionation and avoiding pin fixation. Perfexion, still a large part of the installed base, has neither, so at a Perfexion site the mask pathway described here is not available and the fractionation question is answered differently. Mask treatment is not simply "frame treatment without pins": the reference chain, motion model, margins, treatment pauses, and imaging policy must all be explicitly validated.

Immobilization choice is clinical and geometric, not a test of platform loyalty.
Situation	Often favors	Reason
Functional target or trigeminal neuralgia	Frame	Very small target; rigid reference; high single dose
AVM with angiographic integration	Frame	Stable stereotactic chain across multimodality imaging
Perioptic or large target needing 3–5 fractions	Mask	Fractionation improves normal-tissue tolerance
Claustrophobia, involuntary movement, or poor mask fit	Frame or anesthesia-supported workflow	Motion risk may outweigh mask convenience
Target or patient outside Gamma Knife geometry	Alternative SRS/SRT platform	Choose the system that can deliver the intended plan safely

Part II

Create and Image the Stereotactic Reference

3.Frame application is part of planning

Frame position determines target reach, imaging geometry, patient comfort, airway access, and collision clearance. Review the target location before placement. Position the frame to center the likely treatment coordinates within the usable stereotactic volume, while preserving clearance for the nose, shoulders, and fixation posts. Infiltrate each pin site adequately, avoid thin or pathologic bone and known hardware, tighten opposing pins progressively according to the approved frame protocol, and confirm mechanical stability before imaging.

Frame check before leaving the procedure area Patient identity and target reviewed; frame stable; pin sites dry; posts and pins clear of the anticipated target and imaging field; airway and line access preserved; fiducial indicator available; no obvious collision or reach problem; analgesia and imaging sequence confirmed.

4.Acquire images for geometry and biology

Thin-slice postcontrast 3D T1-weighted MRI is the general tumor workhorse, but it is not sufficient for every indication. Add high-resolution T2/CISS/FIESTA-type imaging for cranial nerves and the internal auditory canal, susceptibility-sensitive sequences when blood products or calcification matter, dedicated sellar or vascular sequences when needed, stereotactic CT for bone detail or MRI-distortion concerns, and catheter angiography for AVM definition when indicated. Define stereotactic space by the method the platform and fixation support: a fiducial indicator box with frame-based MRI or CT, or CBCT-based definition on Icon and Esprit, which is validated for framed as well as mask-fixated patients. Whichever method is used, it must be the one the local program commissioned.

Before contouring, inspect the entire image set for motion, geometric distortion, incomplete coverage, wrap, fat-shift, contrast timing, and registration error. MRI distortion can be clinically important near the skull base and at the edge of the field. Registration is not accepted because the software produced a green check; it is accepted only after an anatomic review in all three planes, with particular attention to the target and nearby organs at risk.

Never let fusion create anatomy If the target appears to move relative to bone, vessels, the brainstem, or the ventricular system across registered studies, stop and resolve the discrepancy. A submillimeter display does not guarantee submillimeter truth.

Part III

Build the Plan

5.Contour the disease, then the constraint

Target definition is indication-specific. A metastasis usually follows the enhancing lesion; a postoperative cavity requires reconstruction of the operative bed and meningeal boundaries; an AVM follows the nidus rather than feeders or draining veins; trigeminal neuralgia targets a defined nerve segment rather than a volumetric tumor. A zero geometric expansion is common in rigid frame-based SRS, but GTV equals CTV is not a universal biologic truth, and mask-based PTV margins depend on the complete immobilization and guidance chain.

Contour the structures that can actually change the plan, including optic nerves and chiasm, brainstem, cochlea, lens/globe when relevant, normal brain for dose-volume assessment, and prior high-dose regions in reirradiation. Review contours on the sequence that best shows each structure rather than trusting one fused series for everything.

6.Shot construction: coverage is only the first move

Gamma Knife planning shapes dose by combining isocenters (shots), collimator sizes, sector settings, weights, and positions. Forward planning remains an anatomical craft: start with the largest collimator that can efficiently fill the target, add smaller shots to repair concavities or protect a critical boundary, and use sector blocking or composite shots when they improve selectivity or organ-at-risk sparing. Filling a moderate target with many 4-mm shots may look meticulous while worsening treatment time and plan efficiency.

Inverse planning can generate a strong starting solution and is particularly useful for irregular or multitarget cases, but it does not know which compromise matters clinically. The final plan still requires direct review of target coverage, prescription-isodose geometry, high-dose distribution, gradient, normal-brain spill, critical-structure dose, collision feasibility, and treatment duration. Two planning inputs are easy to accept by default and worth stating explicitly: the skull or head definition sets the geometry the dose engine uses for depth, and the algorithm matters, since TMR 10 assumes a homogeneous head while the convolution algorithm accounts for heterogeneity and can change dose near air cavities and bone. Record which was used, and treat plans computed with different algorithms as not directly comparable.

Plan review: metrics answer different questions and should be read together.
Metric	Question	Common failure of interpretation
Coverage	How much target receives the prescription dose?	Accepting undercoverage without identifying whether it is deliberate OAR protection
Selectivity	How much prescription volume is actually target?	Sacrificing necessary coverage to chase a cosmetically perfect number
Paddick conformity index	Coverage and selectivity combined into one number, since the index is their product	Comparing targets of very different size or shape as if CI were context-free
Gradient index	How quickly does dose fall outside the target?	Using one universal cutoff despite strong target-size dependence
Normal-brain dose volume	What uninvolved brain receives an injury-relevant dose?	Treating V12 or its multifraction analog as a binary threshold rather than continuous risk
Maximum dose and prescription isodose	Is intratarget heterogeneity appropriate for this indication?	Assuming a homogeneous plan is always safer or a hotspot is always beneficial
Treatment time	Can the plan be delivered reproducibly and tolerated? Cobalt-60 decays with a half-life of about 5.3 years, so beam-on time for the same prescription lengthens between reloads and can roughly double across a source cycle	Ignoring source decay, motion opportunity, sedation, and patient endurance

7.Prescription is a clinical decision

Prescription dose and fractionation should be chosen from disease-specific outcome data, prior radiation, target volume, critical-structure dose, and the patient's goals, using the framework on the Dose Selection & Fractionation page. The prescription isodose line is a planning consequence, not a ritual number. Gamma Knife plans often prescribe to a lower isodose surface than many LINAC plans, intentionally creating central heterogeneity, but the appropriate hotspot depends on target biology, size, geometry, and nearby normal tissue.

Part IV

Verify and Deliver

8.The treatment time-out is a second plan review

Before irradiation, independently confirm patient, indication, target number and laterality, prescription and fraction, approved image set, immobilization mode, critical-structure doses, prior-radiation context, and completion of required physics checks. The layered QA program behind that line, from machine-level radiation-focus and Winston-Lutz testing through independent secondary dose calculation and end-to-end verification, is set out on the Planning, Constraints & QA page. Review every shot for coordinate plausibility and clearance. Verify the stereotactic reference against the method the plan was built on. For an indicator-box definition, confirm the frame and the stereotactic definition. For a CBCT definition, framed or mask-fixated, acquire the localization CBCT, inspect the registration at the target rather than at the skull as a whole, and confirm the coordinate transform before the first shot. For a fractionated mask course, do not stop at the transform: use adaptive dose control to compare the dose about to be delivered at today's position against the approved plan, and carry delivered dose forward across fractions so the course is judged on what was actually given.

Final spoken check Right patient; right target(s); right fraction; right dose; right stereotactic reference; critical constraints reviewed; plan deliverable without collision; motion-management and interruption thresholds understood; treating neurosurgeon, radiation oncologist, and physicist agree.

9.Motion management is a treatment policy

Frame fixation limits motion mechanically but does not remove the need to observe the patient and respond to pain, anxiety, or frame concerns. Intrafraction motion monitoring uses an infrared camera and a reflective marker at the nose, with tracking accuracy well under a millimeter and a user-selectable gating threshold spanning roughly 0.5 to 3 mm, so the threshold a program adopts is a real clinical decision rather than a factory setting. It is most often paired with mask fixation but can also run with a frame to document fixation stability. Because surrogate motion is not identical to target motion, thresholds, image reacquisition, repositioning, and restart rules must be defined by the locally commissioned workflow. Long treatments should be reassessed when repeated interruptions or progressive drift occur.

10.Close the loop after treatment

At completion, document delivered dose and fraction, interruptions, imaging or positional corrections, and any deviation from the approved plan. Remove the frame with attention to pin-site bleeding and vasovagal symptoms, or release the mask and reassess the patient. Steroids, anticonvulsants, analgesics, and medication changes are indication-specific rather than automatic. The discharge plan should state expected latency, urgent warning symptoms, follow-up imaging interval, and which service owns surveillance.

Part V

Failure Modes Worth Anticipating

11.Problems that begin upstream

Wrong fractionation: choosing one fraction before testing whether the target dose and OAR constraints can coexist.
Poor frame geometry: a stable frame that cannot reach the target or collides at the required coordinates.
False confidence in MRI: accepting motion, distortion, incomplete coverage, or a misregistration because the images look sharp.
Contour by habit: applying the same target rule to a metastasis, cavity, AVM, benign tumor, and functional target.
Metric worship: degrading clinically important coverage or OAR sparing to improve a conformity or gradient score.
Shot overcomplexity: excessive small shots that add time without meaningful dosimetric benefit.
Unmodeled motion: using mask fixation without a complete localization, monitoring, interruption, and reimaging policy.
Handoff failure: correct plan, wrong fraction, wrong lesion label, or uncommunicated prior radiation.

Key points

Start with the clinical endpoint and fractionation decision, not the immobilization device.
Frame position is part of treatment planning: target reach, clearance, imaging, comfort, and airway all matter.
Use the imaging sequence that best defines each target and OAR; inspect fusion at the anatomy that limits the plan.
Build with efficient large shots, then refine with smaller collimators, weighting, and sector shaping.
Coverage, selectivity, conformity, gradient, normal-brain dose, OAR dose, and treatment time answer different questions.
Mask treatment requires a commissioned motion policy; intrafraction monitoring is a surrogate, not direct target tracking.
The final time-out is an independent clinical and technical plan review, not an administrative pause.

References

Cirino E, Benedict SH, Dupre PJ, et al. AAPM-RSS Medical Physics Practice Guideline 9.b: SRS-SBRT. J Appl Clin Med Phys. 2025. AAPM
Petti PL, Rivard MJ, Alvarez PE, et al. Recommendations on the practice of calibration, dosimetry, and quality assurance for gamma stereotactic radiosurgery: report of AAPM Task Group 178. Med Phys. 2021. AAPM
Seibert TM, White NS, Kim GY, et al. Distortion inherent to magnetic resonance imaging can lead to geometric miss in radiosurgery planning. Pract Radiat Oncol. 2016;6(6):e319–e328. PubMed
Wright G, Harrold N, Hatfield P, Bownes P. Validity of the use of nose-tip motion as a surrogate for intracranial motion in mask-fixated frameless Gamma Knife Icon therapy. J Radiosurg SBRT. 2017;4(4):289–301. PubMed
Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg. 2000;93(Suppl 3):219–222.
Grishchuk D, Dimitriadis A, Sahgal A, et al. ISRS technical guidelines for stereotactic radiosurgery: treatment of small brain metastases (≤1 cm in diameter). Pract Radiat Oncol. 2023;13(3):183–194. PubMed
Paddick I, Lippitz B. A simple dose gradient measurement tool to complement the conformity index. J Neurosurg. 2006;105(Suppl):194–201.
U.S. Food and Drug Administration. Leksell Gamma Knife Esprit 510(k) record, K222047. FDA

Educational workflow for neurosurgery and radiation-oncology trainees; not a substitute for platform training, local commissioning, approved device instructions, or case-specific multidisciplinary judgment.