Design and evaluation of a patient-specific 3D couplant pad for ultrasound image-guided radiation therapy

Abstract

Purpose: Successful patient treatment with ultrasound image-guided radiation therapy (US IGRT) has been hindered by several issues, including patient deformation due to probe pressure, the presence of an image dead zone, and optical tracking disabilities caused by irregular surfaces. The purpose of this study was to overcome these barriers by using a patient-specific three-dimensional (3D) couplant pad created by a patients skin mold using a 3D printing technique. Materials and Methods: Commercial ultrasound-based localization systems (Clarity® 3.1

Elekta Ltd., Montréal, Québec, Canada) equipped with optical tracking devices were installed in a CT simulation room and in a radiation treatment room. To determine the optimal materials for the couplant pad, various mixing ratios of candidate materials were used and the strengths and elasticities of the resultant couplant pads were tested. A patient skin mold was designed using a skin contour of simulation CT images and fabricated by a 3D printer (CubePro®, Cubify, 3DSYSTEMS, Rock Hill, SC, USA). A couplant pad was then fabricated by pouring gelatin solution into a fixed-shape container accommodating the patients skin mold. To examine the effect of the couplant pad on the baseline positional accuracy of our system, a phantom study was carried out with a breast phantom. From the four patients who underwent US IGRT, a total of 486 ultrasound images (including images with and without the couplant pad) were acquired before treatment. The effectiveness of the couplant pad was evaluated in terms of image contrast, tracking accuracy, and inter-observer variation. Results: The positioning accuracies of our US system in the phantom study were 0.9 ± 0.3 mm and 1.3 ± 0.4 mm with and without the couplant pad, respectively. The patient study revealed that if the US image acquired in the first radiation treatment session is used as a reference, the couplant pad can reduce the mean target shift from 4.7 mm to 3.7 mm in 3D vector amplitude. Moreover, the use of the couplant pad reduced the standard deviation from 2.2 mm to 1.7 mm. Use of the couplant pad also improved image contrast around the treatment target by 10 %. Analysis of the effect of US scanning coverage and target deformation due to excessive probe pressure revealed that the centroid offset of the target volume after target position alignment was decreased from 4.4 mm to 2.9 mm upon use of the couplant pad. Inter-fractional target contour agreement calculations revealed that one patient with a small target showed a substantial increase in the Kappa value (from 0.07 to 0.31) with the use of the couplant pad

however, this effect was not significant in other cases with larger targets. Conclusion: Our patient-specific 3D couplant pad, which was generated using a mold by 3D printing technique, is thus a promising strategy for improving tracking accuracy, image quality, and inter-observer variation for ultrasound-based image guided radiotherapy. In addition to its conventional advantage of noninvasiveness, our couplant pad facilitates the use of ultrasound technology in radiotherapy. Since the position, shape, and volume of the treatment target can be confirmed at every treatment, US is more effective for adaptive radiation therapy compared with conventional x-ray imaging.