NCIRF User Manual
The National Cancer Institute Dosimetry System for Radiography and Fluoroscopy (NCIRF) is a reference-grade radiation dose estimation system developed at the National Cancer Institute (NCI) to estimate organ absorbed doses and effective dose for populations undergoing diagnostic radiography, fluoroscopy, and fluoroscopically guided interventional procedures.
NCIRF integrates computational human phantoms with a streamlined GEANT4 Monte Carlo radiation transport engine. Unlike NCICT and NCINM, which rely on pre-calculated dose conversion coefficients, NCIRF performs direct Monte Carlo radiation transport simulations using user-specified imaging and geometric parameters.
NCIRF is intended to support population-based dose evaluation, benchmarking, and research or regulatory-facing analyses, rather than real-time or patient-specific clinical dose estimation.
Intended use
NCIRF is intended for reference dose reconstruction and comparative analyses.
It is not intended for real-time clinical decision support or site-specific clinical optimization.
Two libraries of computational human phantoms are available in NCIRF 3.0:
- Reference size phantoms
- Size-specific phantoms
Reference size phantoms are categorized by body posture:
- Arms Raised
- Arms Lowered
- Arms Rotated
Users select patient age and gender. Available age groups include:
- Newborn
- 1 year
- 5 years
- 10 years
- 15 years
- Adult (assumed ≥20 years)
For ages between ICRP-defined groups, users may:
- Interpolate organ doses between the nearest younger and older age groups, or
- Select the nearest age group as an approximation.
Reference height and weight, as defined by the ICRP, are displayed automatically and are not editable.
Size-specific phantoms include:
- 169 pediatric models (79 male, 90 female)
- 193 adult models (93 male, 100 female)
Users select:
- Age Group (pediatric or adult)
- Gender (male or female)
- Height (cm) and Weight (kg)
A Phantom Height–Weight Map is provided to guide selection. Selecting the correct age group is essential for accurate calculation of active and shallow bone marrow doses, which rely on age-dependent dose response functions.
Users must define x-ray beam characteristics by selecting a spectrum defined by:
- Peak tube potential (kVp)
- Half-value layer (HVL, mm Al)
Available combinations in NCIRF 3.0 are listed below:
| Peak tube potential (kVp) | Half-value layer (Al mm) |
|---|---|
| 50 | 1.890 |
| 50 | 2.800 |
| 50 | 3.300 |
| 50 | 3.750 |
| 60 | 2.250 |
| 60 | 3.420 |
| 70 | 2.610 |
| 70 | 4.050 |
| 70 | 6.830 |
| 80 | 3.010 |
| 80 | 4.610 |
| 80 | 5.570 |
| 80 | 6.380 |
| 80 | 7.700 |
| 90 | 3.380 |
| 90 | 5.180 |
| 100 | 3.750 |
| 100 | 5.710 |
| 110 | 4.110 |
| 110 | 6.180 |
| 110 | 7.330 |
| 110 | 8.230 |
| 110 | 9.680 |
| 120 | 4.530 |
| 120 | 6.520 |
Additional beam parameters include:
- Source-to-isocenter distance (SID, cm)
- Field width and height at isocenter (cm)
Beam field size updates are reflected in the phantom images.
Dose Area Product (DAP, Gy·cm²) must be provided to convert Monte Carlo dose per particle into absolute absorbed doses.
Beam orientation is defined using:
- Positioner Primary Angle (PPA)
- Positioner Secondary Angle (PSA)
Angles may be entered numerically, adjusted using arrows, or selected from predefined beam directions. Definitions follow the NEMA DICOM standard.
Phantom views (top, frontal, and lateral) display:
- Isocenter location
- Beam width and height at isocenter
- X-ray source location and beam direction
Users may interactively reposition the isocenter using the mouse.
Patient bed thickness (Bed Thickness, cm) may also be specified and is explicitly included in Monte Carlo calculations.
Users specify the number of Monte Carlo histories to control statistical uncertainty.
From NCIRF version 2.0 onward, multithreading is supported.
- Number of threads: 1–24
- Optimal selection typically matches the number of available CPU cores
After clicking Calculate Dose, Monte Carlo simulations are executed in the background using GEANT4.
Note: The user interface may appear temporarily unresponsive during simulations.
Performance scaling with thread count has been evaluated and published (Lee et al., BPEX 2023).
Following simulation, the output table displays:
- Organ absorbed doses (mGy)
- Monte Carlo statistical errors (%)
Effective dose (mSv) is calculated using tissue weighting factors from ICRP Publication 103.
From NCIRF version 2.0 onward, peak skin dose (PSD) is also calculated.
Dose smoothing is applied to reduce voxel-level statistical noise. Approximately 10⁵ histories are typically sufficient for stable PSD estimates.
Batch Manager enables automated execution of multiple simulation cases.
Two versions are available:
- Batch Manager – Reference size
- Batch Manager – Size-specific
Each Batch Manager reflects the selected phantom library and includes the appropriate phantom parameters.
Users may:
- Send current settings to Batch Manager
- Save or load batch parameter sets (CSV)
- Execute Monte Carlo runs via Run Batch Set
- Generate MCNP input files for external computing environments
NCIRF is available at no charge for non-commercial research use under an approved Software Transfer Agreement (STA).
- Complete the STA form
https://dceg.cancer.gov/tools/radiation-dosimetry-tools/ncidose-software-transfer-agreement.pdf - Obtain required signatures
- Submit the completed form to:
Dr. Choonsik Lee
choonsik.lee@nih.gov
Commercial use requires a licensing agreement through the NCI Technology Transfer Center.
For evaluation or licensing inquiries, contact:
Dr. Kevin Chang
changke@mail.nih.gov