Reconstruction of Catheter Paths for High Dose Rate Breast Brachytherapy: A Phantom Model

Abstract

Accelerated partial breast irradiation (APBI) is performed in early-stage breast cancer patients after lumpectomy, and treats the surgical bed rather than the entire breast. The treatment time for APBI is one week or less compared to conventional whole-breast irradiation where the treatment time ranges from three to seven weeks. One treatment modality in APBI is multi-catheter interstitial brachytherapy (MIB); however, MIB is a technically challenging procedure because knowledge of both the positioning and the spacing of the inserted catheters is crucial for achieving adequate dosimetric coverage of the lumpectomy cavity. Conventionally, ultrasound, fluoroscopy, or computed tomography (CT) are used to determine the catheter positioning and spacing, but these are either difficult to interpret, or they expose the patient to radiation. We investigated the use of electromagnetic reconstruction (EMR) for this purpose. This study compares the catheter paths recorded with EMR with catheter paths modeled on CT. Two anthropomorphic plastic phantoms were made with simulated tumour beds that could be seen with both ultrasound and CT. In this experiment, thirteen catheters were inserted into the phantoms using ultrasound guidance. EMR was performed by pulling through each catheter an electromagnetic enabled sensor and collecting position information. The position information was then used to reconstruct the catheter path based on a least-squares polynomial fit using the free and open-source 3D Slicer (www.slicer.org) platform. The sensor transmitted position information to a navigation computer using the PLUS toolkit software. CT images were collected and aligned with the reconstructions to conduct an analysis of accuracy. The Vascular Modeling Toolkit software was used to compare the EM-generated paths to the CT generated paths. The paths were then aligned and measured at regular positions along the phantom. The aim of the study was to be accurate within 0.68mm, which is the axial resolution of the CT scanner. The thirteen catheters modeled with EMR were all aligned with the paths imaged on the CT scan. The mean distance between the catheter paths modeled on the CT scan and EMR was 0.38 +/- 0.13 mm, which is lower than the axial resolution of the CT scanner. The use of EMR was accurate and reproducible for determining the paths of the implanted catheters. The results of this phantom experiment suggest that using EMR is a promising tool that is consistent and reproducible. EMR’s ability to accurately determine catheter positioning and spacing in three dimensions may be useful in facilitating planning. Further research is being conducted to use this tool in patients and in clinical trials.