Long-term stability analysis of beam shape in a robotic radiosurgery system

Abstract

Purpose: This study aimed to investigate the long-term stability of CyberKnife beam profile parameters and assess their compliance with existing quality assurance (QA) guidelines. We evaluated beam profiles in both standard and diagonal planes over 3.5 years post-installation to detect potential issues and ensure consistent beam quality. The findings will contribute to validating and refining current QA practices for CyberKnife systems.

Methods: Beam profile measurements were performed monthly using an Octavius 1000SRS detector array. The profiles were evaluated in terms of the beam shape constancy within 2%, and the penumbra, symmetry, and flatness were analyzed using statistical process control methods. Temporal changes in the dose profiles were visualized using dose difference heat maps. The relationship between the beam parameters and accumulated monitor units was also investigated.

Results: The 2% profile constancy check accurately detected magnetron deterioration 2 months before failure, confirming its high sensitivity for beam stability monitoring. While symmetry and flatness remained within 0.7% throughout over 100 × 10⁶ monitor units of operation, penumbra exhibited greater responsiveness to magnetron-induced changes but did not consistently flag all orientations. Additionally, statistical analyses and heat maps revealed gradual profile shifts independent of acute component failures, highlighting the importance of multifaceted QA strategies.

Conclusions: These findings reinforce the effectiveness of the 2% profile constancy check for early detection of magnetron failure and support its adoption in current CyberKnife guidelines. At the same time, symmetry, flatness, and penumbra parameters remain valuable for characterizing gradual profile variations. Collectively, this study underscores the need for comprehensive beam monitoring and regular maintenance to sustain optimal CyberKnife system performance.

Keywords: CyberKnife; flatness; long‐term analysis; penumbra; symmetry.

FIGURE 1

Schematic diagram of the measurement setup. The figure shows the arrangement of the Octavius 1000SRS detector array and Solid Water HE relative to the CyberKnife beam. The scanning plane is also illustrated to demonstrate the profile measurement orientation.

FIGURE 2

Pass rates for TG‐135 criteria (2% dose difference) compared to baseline over time. The graph shows the results for longitudinal, lateral, first diagonal, and second diagonal orientations. The black dashed line indicates the magnetron replacement.

FIGURE 3

Penumbra measurements for both sides over time for longitudinal and lateral orientations. (a) Longitudinal positive side, (b) longitudinal negative side, (c) lateral positive side, (d) lateral negative side. The black dashed line indicates the magnetron replacement. LCL, lower control limit; UCL, upper control limit.

FIGURE 4

Penumbra measurements for both sides over time for first diagonal orientation. (a) First Diagonal Positive Side, (b) First Diagonal Negative Side, (c) Second Diagonal Positive Side, (d) Second Diagonal Negative Side. The black dashed line indicates the magnetron replacement. UCL: upper control limit, LCL: lower control limit.

FIGURE 5

Symmetry measurements over time for (a) longitudinal, (b) lateral, (c) first diagonal, and (d) second diagonal orientations. The black dashed line indicates the magnetron replacement. UCL: upper control limit, LCL: lower control limit.

FIGURE 6

Flatness measurements over time for (a) longitudinal, (b) lateral, (c) first diagonal, and (d) second diagonal orientations. The black dashed line indicates the magnetron replacement. UCL: upper control limit, LCL: lower control limit.

FIGURE 7

Time series heatmaps of dose profile changes for (a) longitudinal, (b) lateral, (c) first diagonal, and (d) second diagonal orientations. The red triangle indicates the magnetron replacement. The relative dose differences represent the differences between each measurement and the baseline in relative dose.

FIGURE 8

(a) 2D heatmap of the mean of dose differences in beam profiles over time relative to the baseline. The mean of absolute dose differences represents the average of the differences between each monthly measurement and the baseline. (b) 2D heatmap of the baseline dose profiles.

FIGURE A1

Effects of magnetron current on (a) longitudinal and (b) lateral profiles, beam current on (c) longitudinal and (d) lateral profiles, steering X2 current on (e) longitudinal profile, and steering Y2 current on (f) lateral profile.

FIGURE A2

Comparison of beam profiles immediately before magnetron failure (green) and baseline profiles (blue) for (a) longitudinal, (b) lateral, (c) first diagonal, and (d) second diagonal orientations. Black dashed lines mark the 80% and 20% dose levels, red dashed lines indicate the 2% dose difference threshold, and purple dashed lines depict the relative dose differences between the baseline and pre‐failure profiles.

FIGURE A3

(a) Beam profile differences for the CyberKnife longitudinal beam, measured with the Octavius detector's longitudinal and first‐diagonal channels. (b) Beam profile differences for the CyberKnife first‐diagonal beam, measured with the Octavius detector's first‐diagonal and lateral channels.