Evaluating Treatment Planning System Commissioning Data
Treatment planning system commissioning turns measured machine behaviour into a clinical calculation model. The quality of that model depends on far more than entering beam profiles and depth-dose curves. Physicists must decide whether the data are complete, internally consistent, suitable for the intended treatment techniques and supported by defensible evidence.
For Australian departments, the task often involves balancing local protocols, vendor recommendations, staffing capacity and the needs of varied clinical services. A system used for conventional three-dimensional conformal therapy may require a different commissioning depth from one supporting VMAT, stereotactic radiotherapy or adaptive workflows. The evaluation must reflect those differences.
The 2014 COMP meeting in Banff provided a professional setting for discussing scientific methods, clinical governance and practical implementation. Delegates preparing a departmental record of such technical discussions could use this conference report guide to preserve decisions, limitations and follow-up actions rather than recording only presentation titles.
Establish The Clinical Scope First
Commissioning begins with a written statement of intended use. Identify treatment units, beam energies, applicators, wedges, imaging systems, dose calculation algorithms and treatment techniques that will be supported. Include whether the model will be used for simple fields, breast tangents, total body irradiation, stereotactic cases, electrons or highly modulated plans.
The scope should also identify patient and site-specific challenges. A service in Sydney with multiple linear accelerators may need a formal process for matching machines and managing changes between sites. A smaller regional centre may prioritise robust three-dimensional planning, while a major Melbourne or Brisbane service may need detailed validation of flattening-filter-free beams, small fields and image-guided workflows.
Commissioning criteria should be agreed before reviewing results. This prevents the team from changing tolerances to accommodate an unexpected discrepancy. Use published guidance, vendor specifications, professional recommendations and the department’s risk assessment, while documenting where a local tolerance is stricter or more permissive.
Assemble Reliable Measurement Data
The data set should cover the measurements required by the calculation algorithm, not simply the data that are easiest to collect. Typical inputs include percentage depth dose or tissue-phantom ratios, central-axis profiles, output factors, wedge factors, tray factors and absolute calibration information. For electrons, include depth-ionisation or depth-dose curves, practical range, cone factors and profiles for each clinically relevant applicator.
Measurement conditions must be recorded with enough detail to make the data reproducible. Detector type, chamber orientation, scanning resolution, field definition, phantom material, reference depth, source-to-surface distance, temperature, pressure and electrometer settings can all influence interpretation. Small-field measurements require particular care because detector volume, perturbation and positioning errors may be comparable with the physical signal being measured.
Essential Data Integrity Checks
- Confirm that field sizes, depths and setup distances match the treatment planning system conventions.
- Compare repeated measurements to quantify repeatability before importing values.
- Check that profile symmetry, penumbra and output trends are physically plausible.
- Verify absolute dose against the department’s calibrated reference procedure.
- Preserve raw files, processed files and import settings under version control.
Data should be reviewed visually as well as numerically. A smooth curve can conceal a misplaced scan point, an incorrect sign convention or an interpolation artefact. Plotting measured and imported values together often reveals errors that a pass-or-fail table misses.
Validate The Beam Model
A commissioned model must reproduce independent measurements across the clinical range. Start with open fields and standard geometries, then move towards wedges, asymmetric fields, irregular apertures, MLC-defined fields and small segments. Compare dose at central-axis points, high-gradient regions and clinically relevant depths rather than relying on a single global percentage difference.
For photon beams, assess build-up, dmax, dose fall-off, out-of-field behaviour and profile shape. Examine how the algorithm handles heterogeneities such as lung, bone and air cavities. Electron models need careful assessment of surface dose, practical range, bremsstrahlung tail and lateral scatter. Stereotactic applications demand additional attention to detector suitability, jaw and MLC modelling, output factors and dose gradients.
The Australian context makes independent verification especially valuable. Local services may use equipment from several suppliers, with Varian, Elekta and other platforms represented across public and private networks. The principles in TRS-398, AAPM task group reports and ACPSEM practice guidance can support the evaluation, but they should be applied alongside machine-specific evidence and the requirements of the hospital’s quality system.
Test Clinical Planning Scenarios
Point-dose checks are necessary but insufficient. A treatment planning system can achieve good central-axis agreement while producing clinically important errors in regions affected by modulation, heterogeneity or steep gradients. Build a validation suite that mirrors actual patient treatments and includes challenging geometries.
Use representative plans for head and neck, prostate, lung, breast, pelvis and brain treatments, then extend the suite for stereotactic and paediatric work where applicable. Compare dose-volume metrics, target coverage, organ-at-risk doses, conformity, homogeneity and monitor units. Independent dose calculations, portal dosimetry, phantom measurements or another validated planning system can provide complementary evidence.
When reviewing results, distinguish between a calculation error and a measurement limitation. A disagreement near a sharp gradient may reflect setup uncertainty rather than a defective algorithm. Record the uncertainty budget, repeat questionable measurements and escalate unresolved differences through a structured technical review. A short accommodation advice page may help conference delegates plan attendance, but the same principle applies clinically: preparation reduces avoidable variability before the technical work begins.
Govern Changes And Ongoing Checks
Commissioning does not finish when the system is released for patient planning. Software upgrades, beam steering changes, target replacement, MLC servicing, imaging updates and new calculation options can affect the model. Establish triggers for partial or full recommissioning, with responsibilities assigned to named physicists and approval documented by the appropriate clinical authority.
A practical change-control process should link each modification to affected data, tests and clinical restrictions. If a new dose algorithm is enabled only for selected sites, the treatment planning system should make that limitation visible. If a machine is temporarily operated with a changed energy or altered accessory, planners and therapists need a clear communication pathway.
Ongoing Assurance Activities
- Review a defined sample of clinical plans against independent checks.
- Repeat selected output, profile and depth-dose measurements after major service work.
- Audit imported data and calculation settings after software upgrades.
- Track near misses, rejected plans and recurring discrepancies.
- Reassess tolerances when new techniques or detectors are introduced.
Australian departments should also account for regulatory and accreditation expectations, including local radiation licensing requirements and the broader safety framework associated with ARPANSA guidance. Records need to be accessible during internal audits, external reviews and incident investigations, even when the original commissioning physicist has moved to another service.
Document Evidence For Clinical Release
A commissioning report should explain what was tested, why it was tested and how the results support clinical use. Include system identifiers, software versions, beam data sources, measurement equipment, calibration status, calculation settings, acceptance criteria, deviations and approval signatures. Attach representative plots and tables, while retaining complete raw data in a controlled repository.
The report should state limitations plainly. For example, a model may be validated for conventional photon planning but not for very small fields, proton therapy, total skin electron therapy or a particular adaptive workflow. Clear boundaries are safer than broad statements that imply every available feature has been tested.
Conference participation can strengthen departmental practice when lessons are converted into local actions. A technical presentation from Banff, a discussion over coffee after a scientific session or an informal Banff icebreaker can prompt useful questions about detector choice, acceptance criteria and audit design. The value comes from translating those discussions into measurable changes in the clinical environment.
Use the completed evidence package to authorise specific clinical applications, train staff and schedule future review. Before releasing a new model, verify that treatment planners understand the validated scope and that independent checking procedures reflect the same assumptions. In cities such as Perth, Adelaide and Canberra, where specialist resources may be shared across services, a clear report also supports consistent practice when expertise is distributed.
Apply a structured review to every commissioned model: define the clinical scope, verify the measurement chain, challenge the algorithm with independent tests, assess realistic treatment plans and control future changes. A defensible record gives the department confidence that calculated dose is fit for purpose and provides a clear basis for safe clinical release.