Purpose

To evaluate the geometric accuracy of applicator reconstruction in gynecological brachytherapy using low-field magnetic resonance imaging.

Material and methods

A 0.55T magnetic resonance (MR) scanner was used in this study. First, accuracy of needle reconstruction based on MR and computed tomography (CT) images were compared. Seven plastic needles were inserted into three phantoms, which were scanned with MR and CT. Five observers determined tip and entry points of the needles on both types of images. Needle lengths and tip coordinates were compared. Then, inter-observer variation in localization of applicator and needle tips in patients was evaluated on MR images. Twenty fractions of eleven patients treated with Venezia applicator with needles were selected. In tandem and ovoid tubes, MR markers were inserted, with only air in the needles. Differences between observers in determining the coordinates were depicted using standard deviation (SD).

Results

The mean difference between needle depths was 1.16 ±0.8 mm measured on MR and CT images. Uncertainties among observers in determining the position of tips were generally small, with SD of less than 1.0 mm for both imaging modalities. A good correlation (R² > 0.99) was found between the tip coordinates calculated on MR and CT images. In patients, SD of applicator tips with MR markers was smaller than that of needle tips without markers (X: 0.18 vs. 0.27 mm, Y: 0.24 vs. 0.58 mm, Z: 0.21 vs. 0.34 mm). Regarding direction, the greatest variability was constantly observed in the Y direction that corresponded to the greater slice thickness.

Conclusions

The 0.55T MR scanner can provide clinically acceptable accuracy in the reconstruction of brachytherapy (BT) applicators and needles. MR markers allow greater precision in determining the tips of applicators. The accuracy of reconstruction using MR images is comparable to that achieved with CT images.