ENAL++: a new and effective off-line correction protocol for rotational setup errors when using a robotic couch.

Daan Martens,Mark Luesink,Kasper L. Pasma,Henk Huizenga

doi:10.1120/jacmp.v16i6.5583

Abstract

Cone‐beam CTs (CBCTs) installed on a linear accelerator can be used to provide fast and accurate automatic six degrees of freedom (6DoF) vector displacement information of the patient position just prior to radiotherapy. These displacement corrections can be made with 6DoF couches, which are primarily used for patient setup correction during stereotactic treatments. When position corrections are performed daily prior to treatment, the correction is deemed "online". However, the interface between the first generation 6DoF couches and the imaging software is suboptimal. The system requires the user to select manually the patient and type the match result by hand. The introduction of 6DoF setup correction for treatments, other than stereotactic radiotherapy, is hindered by both the high workload associated with the online protocol and the interface issues. For these reasons, we developed software that fully integrates the 6DoF couch with the linear accelerator. To further reduce both the workload and imaging dose, three off‐line 6DoF correction protocols were analyzed. While the protocols require significantly less imaging, the analysis assessed their ability to reduce the systematic rotation setup correction. CBCT scans were acquired for 19 patients with intracranial meningioma. The total number of CBCT scans was 856, acquired before and after radiotherapy treatment fractions. The patient positions were corrected online using a 6DoF robotic couch. The effects on the residual rotational setup error for three off‐line protocols were simulated. The three protocols used were two known off‐line protocols, the no action level (NAL) and the extended no action level (eNAL), and one new off‐line protocol (eNAL++). The residual setup errors were compared using the systematic and random components of the total setup error. The reduction of the rotational setup error of these protocols was optimized with respect to the required workload (i.e., number of CBCTs required). Rotational errors up to 3.2° were found after initial patient setup. The eNAL++ protocol achieved a reduction of the systematic rotational setup error similar to that of the online protocol (pitch from 0.8° to 0.3°), while requiring 70% fewer CBCTs. With a 6DoF robotic couch, translation, and rotation patient position corrections can be performed off‐line to reduce the systematic setup error, workload, and patient scan dose.PACS numbers: 87.56.Fc, 87.56.Da, 87.57.‐s

Highlights

178 Martens et al.: Setup protocols for rotations three translations and three rotations)
A systematic rotational error can lead to a dose distribution that does not coincide with the planning target volume (PTV) as intended
A systematic rotational error of 2° causes a 1.7 mm translation at 50 mm from the isocenter. Correction of these rotational errors allows for the clinical target volume (CTV) to PTV margin of 2 mm that was used for the patient group described here

Summary

Introduction

178 Martens et al.: Setup protocols for rotations three translations and three rotations). As a consequence of this CBCT imaging, it has become apparent that sometimes significant corrections for rotations are required.[1,2,3,4,5,6]. The full rotational information available from CBCT imaging is not used. Neglecting a typical rotational setup error of 2° may cause up to 1.7 mm deviation at a distance of 50 mm from the isocenter. This can occur when the planning target volume (PTV) or an organ at risk (OAR) is located off-axis, or if the PTV is not shaped spherically. In the patient group investigated in this study, rotational errors up to 3.2° were observed, in spite of the mask fixation

Methods

Results

Discussion

Conclusion