Clinical Validation of two Surface Imaging Systems for Patient Positioning in Percutaneous Radiotherapy

TL;DR

This study validates two optical surface imaging systems, Catalyst and AlignRT, for patient positioning in radiotherapy, showing they can improve accuracy especially in pelvic and chest regions without additional radiation exposure.

Contribution

The paper provides a clinical validation of two surface imaging systems, demonstrating their potential to enhance patient positioning accuracy in radiotherapy across different body regions.

Findings

AlignRT improved pelvic positioning accuracy by 0.9mm in 75% of cases.

Both systems were most accurate in head & neck regions.

Further improvements are needed to reduce variability in chest and pelvic regions.

Abstract

Precise patient positioning and thus precise positioning of the planning target volume (PTV) is a prerequisite for effective treatment in percutaneous radiation therapy. Conventional imaging modalities used to ensure exact positioning for treatment typically involve additional radiation exposure of the patient. A different approach to patient alignment without additional exposure is provided by optical 3D surface scanning and registration systems. We investigated two systems Catalyst and AlignRT, which allowed us to generate and validate surface images of 50 patients undergoing irradiation on a Tomotherapy system. We compared the positions proposed by these two surface imaging systems (SIS) with the actual adjustments in patient positions made on the basis of MVCT scans. With the two SIS, the surface of every single patient was repeatedly recorded throughout treatment. In addition, we performed a systematic analysis of both systems using a body phantom. At least 10 surface images were available from each patient included in the analysis. The large number of patients allowed us to subdivide the study population into three groups by body region treated: head & neck, chest, and pelvic. The results for each of these three body regions were separately analyzed. Thus, the focus of our investigation is on body regions registered with the two systems rather than tumor entities. For AlignRT, for instance, we found an improvement in positioning for the Pelvic region of 0.9mm in 75% and 6.4mm in 95% of the collected data in comparison to conventional marker alignment. The two systems were most accurate in the head & neck region, where conventional mask alignment is already sufficiently accurate. The other two body regions, pelvic and chest, might benefit from positioning by SIS. But for both systems there is still a need for improvement to reduce statistical variation.