Detection of dose delivery variations on TomoTherapy using on-board detector based verification

Area of Science:

• Medical Physics
• Radiation Oncology
• Quality Assurance

Background:

• Patient-specific treatment quality assurance (QA) is crucial for accurate radiation delivery.
• TomoTherapy uses a multileaf collimator (MLC) with leaf opening time (LOT) to shape radiation beams.
• Deviations in planned versus delivered doses can compromise treatment efficacy and patient safety.

Purpose of the Study:

• To report a clinical case of significant dose delivery deviations identified during TomoTherapy QA.
• To evaluate the performance of an in-house developed QA system, MCLogQA, in detecting these deviations.
• To investigate the root cause of dose inaccuracies related to MLC LOT.

Main Methods:

• Implemented MCLogQA, a QA system using log files and detector-based MLC LOT, for patient-specific treatment QA.
• Verified detected dose deviations (>10%) using a created delivery QA (DQA) plan.
• Transferred treatment plans to a second TomoTherapy unit and investigated phantom plans to assess unit dependency and LOT uncertainty.

Main Results:

• MCLogQA identified >10% dose deviation in a low dose/fraction plan.
• MLC LOT was found to be higher than planned (2.4% and 6.6% increases on two units), leading to increased mean doses (3.5% and 9.3%) and PTV dose variations (14% and 16%).
• Inaccurate MLC LOT was attributed to a poor latency model at very small LOTs, confirmed by phantom testing.

Conclusions:

• Detector-based MCLogQA effectively detects plan-delivery deviations in TomoTherapy, particularly those related to MLC LOT uncertainty.
• Inaccurate MLC LOT, especially at low values, can cause significant dosimetric variations, impacting treatment accuracy.
• MCLogQA provides valuable information for decision-making during QA verification and helps ensure accurate radiation dose delivery.