# Comparative evaluation of image‐guided radiation therapy (IGRT)‐based dose calculation accuracy using cone‐beam, megavoltage, and kilovoltage CT modalities

## TL;DR

This study compares the accuracy of three imaging techniques for radiation therapy, finding that all are clinically acceptable when properly calibrated.

## Contribution

The study provides a comparative evaluation of IGRT modalities for dose calculation accuracy using modality-specific HU-D calibrations.

## Key findings

- IG-kVCT showed the highest agreement with CT-sim, with gamma passing rates above 98.5%.
- CBCT and MVCT had slightly lower accuracy, with gamma passing rates around 96.5% to 97.2%.
- All modalities demonstrated clinically acceptable agreement with CT-sim and measured doses.

## Abstract

This study evaluates the dosimetric accuracy of three image‐guided radiotherapy (IGRT) imaging modalities, cone‐beam computed tomography (CBCT), megavoltage computed tomography (MVCT), and image‐guided kilovoltage computed tomography (IG‐kVCT), using modality‐specific HU‐to‐density (HU‐D) calibrations. Dose calculations from IGRT images were compared to computed tomography simulation (CT‐sim) references in phantoms and validated against measurements from head and neck and prostate patient plans to assess the feasibility of each modality for precise dose calculation in adaptive radiotherapy (ART).

Two phantoms, the Tomo phantom HE and CIRS Thorax phantom, were used for HU‐to‐density (HU‐D) calibration. IGRT images were acquired using Elekta Synergy XVI (CBCT) and Radixact X9 (MVCT and IG‐kVCT), and calibration curves were generated for each modality. Dose distributions calculated from IGRT images were then compared with those from CT‐sim in phantom studies. For measurement‐based evaluation, 10 patient plans (head and neck and prostate cases) were delivered to a phantom and measured using the ArcCHECK system, and recalculated doses on CT‐sim and IGRT images were compared to the measured doses. Gamma analysis was performed to assess dosimetric accuracy.

IG‐kVCT showed the closest agreement with CT‐sim, achieving gamma passing rates (GPR) of 99.8% ± 0.3% for 3%/3 mm and 98.5% ± 0.7% for 3%/2 mm criteria, with dose differences below 1%. CBCT and MVCT demonstrated slightly lower accuracy, with GPRs of 97.2% ± 1.1% and 96.5% ± 1.3% for 3%/3 mm, respectively, and dose differences up to 2%. Similar trends were observed when compared to measured doses. All IGRT modalities showed clinically acceptable agreement, and no statistically significant differences were found between CT‐sim and any IGRT modality.

All three IGRT modalities demonstrated clinically acceptable accuracy for adaptive dose calculation with modality‐specific HU‐D calibration curves.