IMRT QA Gamma Settings
Paste reference and measured point clouds, then compute gamma statistics.
Example Data Table
These points match the default textareas above.
| x (mm) | y (mm) | dose |
|---|---|---|
| 0 | 0 | 1.00 |
| 3 | 0 | 0.98 |
| 0 | 3 | 0.97 |
| 3 | 3 | 0.95 |
| x (mm) | y (mm) | dose |
|---|---|---|
| 0 | 0 | 0.99 |
| 3 | 0 | 0.97 |
| 0 | 3 | 0.96 |
| 3 | 3 | 0.94 |
Formula Used
For each measured point m, the algorithm searches nearby reference points r and computes:
- Δd is the spatial distance between points (mm).
- dcrit is the distance criterion you set.
- ΔD is the absolute dose difference.
- Dcrit is the dose criterion (percent or absolute).
- A point passes if γ ≤ 1.0.
How to Use This Calculator
- Paste your reference dose points into the reference box.
- Paste measured dose points into the measured box.
- Set distance and dose criteria, plus a threshold.
- Click Compute Gamma to view results above.
- Use Download CSV for full point results.
- Use Download PDF for a quick summary.
Professional Notes on IMRT Gamma QA
1) Why gamma remains a practical metric
Gamma analysis combines spatial mismatch and dose mismatch into one index. In IMRT QA it helps compare deliveries when leaf motion, model limits, or setup shift dose patterns by millimeters while changing magnitude. The pass rate supports quick trending across machines and dates.
2) Typical criteria and what they imply
Routine checks often use 3%/3 mm, while tighter studies use 2%/2 mm. The dose criterion may be percent of a normalization dose or an absolute tolerance. The distance criterion defines how far the search may move to find agreement. Stricter settings increase sensitivity but can lower pass rate.
3) Global versus local normalization
Global percent uses one reference scale (often the maximum reference dose), which stabilizes scoring in low-gradient regions. Local percent scales the dose criterion to the local reference dose, which increases sensitivity in low-dose areas and near steep gradients. Local mode can highlight small relative errors but may lower the overall passing rate.
4) Dose threshold and low-dose exclusion
Applying a dose threshold (frequently 5% to 20% of global dose) reduces noise-driven failures where absolute dose is small. This calculator excludes measured points whose nearest reference dose is below the threshold. Track the excluded count because a very high exclusion fraction may indicate sparse sampling or an overly aggressive threshold.
5) Discrete sampling and search radius
With discrete point clouds, gamma depends on sampling density. If points are spaced near the distance criterion, agreement may hinge on the search radius. This tool sets radius as a multiplier times the distance criterion (for example, 2×3 mm = 6 mm), helping avoid missed neighbors in coarse grids.
6) Interpreting the summary statistics
Pass rate reports the percentage of evaluated points with γ ≤ 1. Mean γ summarizes overall agreement, while max γ highlights the worst mismatch. A histogram is helpful: a plan with many points in 0.5–1.0 may be close to failing under stricter criteria even if it currently passes.
7) Common failure patterns to investigate
Clustered failures near field edges can suggest setup shifts or penumbra modeling issues. Failures in high-gradient regions may indicate resolution limits, detector volume averaging, or small MLC timing differences.
8) Reporting and audit-ready exports
For documentation, record the exact criteria used: dose type, distance, threshold, and normalization source. The CSV export provides per-point γ and pass/fail status for traceability. The PDF export gives a concise summary for attaching to routine QA records and internal audits.
FAQs
1) What does γ ≤ 1.0 mean?
It means a measured point agrees with at least one nearby reference point within your combined distance and dose criteria. The distance and dose contributions are combined by a root-sum-square calculation.
2) Should I use global or local percent mode?
Global percent is more stable for routine trending and low-dose regions. Local percent is more sensitive to relative differences at each point. Choose the mode that matches your site’s QA protocol.
3) Why are some points marked Excluded?
Excluded points fall below the dose threshold, based on the nearest reference dose. Excluding low-dose points reduces noise-driven failures and focuses scoring on clinically meaningful regions.
4) What does “No match” indicate?
No match means the algorithm found no reference point within the configured search radius for that measured point. This often happens when sampling is sparse or coordinate grids do not overlap well.
5) How should I choose the search radius multiplier?
Start with 2× the distance criterion for discrete point clouds. If you have dense sampling, a smaller multiplier can be adequate. If sampling is coarse, a larger multiplier can prevent false “No match” results.
6) Can I use absolute dose criteria?
Yes. Select “Absolute” and enter a dose tolerance in the same units as your data. Absolute criteria are helpful when doses are already normalized or when you need a fixed tolerance scale.
7) How do I interpret a high max γ but good pass rate?
A small region may have a large discrepancy while most points agree well. Review the CSV to localize failures and determine whether the outlier region is clinically significant or related to sampling artifacts.