Monitoring anatomical changes of individual patients using statistical process control during head-and-neck...

Area of Science:

• Radiotherapy physics and technology
• Medical imaging and image analysis
• Statistical quality control in healthcare

Background:

• Reducing planning target volume (PTV) margins in head-and-neck radiotherapy (HNRT) aims to decrease toxicity while maintaining loco-regional control.
• Quantifying anatomical changes during HNRT is crucial for monitoring treatment effectiveness and patient safety.

Purpose of the Study:

• To investigate the feasibility of using deformable image registration (DIR) and Exponentially Weighted Moving Average (EWMA) Statistical Process Control (SPC) charts to monitor anatomical changes during HNRT.
• To assess the ability of SPC to detect systematic anatomical deviations and inform treatment adaptation.

Main Methods:

• DIR was applied to daily cone beam computed tomography (CBCT) images against planning CT (pCT) images from twelve HNRT patients.
• EWMA charts were used to analyze trends in anatomical changes, with uncertainty analysis providing 90% confidence limits.
• Comparison was made between SPC process limits and an a priori clinical limit of 2 mm for patient pose changes.

Main Results:

• Trends in patient positioning reproducibility were observed before week four in 54% of cases, with 24% confirmed by SPC at a 90% confidence level.
• Absolute changes in patient pose exceeding 2 mm were detected in 39% of cases, with 82% confirmed.
• Soft tissue trends outside SPC process limits before week four were confirmed in 90% of cases.

Conclusions:

• Structure-specific action thresholds effectively detected systematic anatomical changes within the first four weeks of HNRT.
• Further investigation into the dosimetric impact of observed deviations is necessary to validate SPC's efficacy in timely treatment adaptation.
• SPC charts show potential as a safety net for reduced PTV margins in HNRT.