Skip to content

About

Open-source framework for automated radiotherapy plan review with PQM evaluation, DVH visualization, and failure-mode logging.

Topics

Resources

Contributing

Stars

7 stars

Watchers

0 watching

Forks

Latest commit

 

History

10 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ClearPlan

Configurable, read-only radiotherapy plan review — shared across the workspace and report.

ClearPlan brings clinical goals, plan checks, DVHs, target and delivery metrics, CT overlays, and optional BEV/geometry views into one review workflow. Its purpose is to make a clinic's own checks and documentation easier to maintain, inspect, and reuse—not to approve a treatment plan automatically.

The current source is 3.2.0-dev.20260916.1, a development snapshot requiring local commissioning. The evaluated code snapshot is distinct from the older v3.2.0 release. Software tests are not evidence of clinical safety or effectiveness.

Start here

I want to… Read
Try ClearPlan without a TPS or patient Patient-free quick start
Set paths, constraints, aliases, and defaults Configuration guide
Introduce ClearPlan at another clinic Clinic onboarding
Understand calculations and limitations Methods and review behavior
Extend checks or connect another scripting API Architecture and developer handoff
Reproduce native ESAPI or Citrix tests Runner and native-test workflow

What it adds

  • Editable local policy: RefDB JSON or Excel constraints, explicit aliases, target rules, field-name conventions, check inclusion, and versioned history. Public starter rules are not a clinic's complete commissioned PlanCheck.
  • One review state: shared structure visibility, wide DVHs, three-plane CT overlays, compact findings, active-plan PDF/HTML reports, and separate plan comparison. German/English selection changes presentation, not identifiers or measurements.
  • Additional measurements: target-specific PAM with physical single- or dual-layer effective apertures, Paddick CI and its reciprocal, GI/HI, total MU, aperture summaries, and explicitly estimated—not measured—dose-rate traces.
  • A reusable boundary: an ESAPI-independent snapshot and shared WPF/reporting components. The Eclipse host is read-only. The RayStation adapter is illustrative and unvalidated, not feature-equivalent clinical support.

See the actual Settings gallery for editable defaults, constraint history, aliases, and check selection. These patient-free captures use the real WPF views with public configuration fixtures and labeled demo history.

Actual English-language Settings view showing configurable check selection

Quick check on Windows

Requires Windows x64, MSBuild, and .NET Framework 4.8 developer tools. No licensed TPS assemblies or patient data are needed:

.\tools\test-portable-review.ps1
.\artifacts\simulator\Debug\ClearPlan.Simulator.exe --scenario mixed-review

The simulator uses explicitly synthetic records. A licensed Eclipse build needs additional dependencies and local checks; follow the build guide. Use your approved execution policy.

Portable tests and the simulator need no Runner. Native integration tests need an authorized ESAPI host; our Citrix workflow uses the separately maintained ESAPI Runner Hub to reopen explicit planning contexts and record launch outcomes. Application assertions and rendered reports are verified separately. ClearPlan can also run in Eclipse or another suitable authorized host; see why this distinction matters.

Safety, privacy, and research scope

Inspection does not modify TPS plans, structures, beams, or dose. Optional ARIA report upload is a separate, explicitly configured patient-record write, disabled by default. Missing evidence is not a successful check. Geometry illustrations are not collision clearance; a rate estimate is not a delivery log.

Public contributions must not contain patient reports, clinical screenshots, private rules, credentials, licensed TPS assemblies, or local device catalogs. Masking a name and ID does not anonymize clinical images or snapshots.

The research aim is a Technical Note about an adaptable review workflow and its software boundaries—not a claim of accepted publication, detection performance, or clinical benefit. Existing paper materials are the historical v3.2.0 synthetic draft, not the current local manuscript or validation of this development snapshot.

Contribute and cite

Read CONTRIBUTING.md, use patient-free examples in Issues, and record the exact evaluated commit and configuration revisions. CITATION.cff provides attribution. The documentation index links methods, configuration, commissioning, and evidence scopes.

About

Open-source framework for automated radiotherapy plan review with PQM evaluation, DVH visualization, and failure-mode logging.

Topics

Resources

Contributing

Stars

7 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages