Lexicographic optimization-based planning for stereotactic radiosurgery of brain metastases

Area of Science:

• Medical Physics
• Radiation Oncology
• Medical Imaging

Background:

• Stereotactic radiosurgery (SRS) requires precise treatment planning.
• Manual planning for intracranial lesions can be time-consuming.
• Automated planning systems aim to improve efficiency and consistency.

Purpose of the Study:

• To validate the mCycle automated lexicographic optimization-planning system for intracranial SRS.
• To assess the system's performance for single (SL) and multiple (ML) brain metastases (up to 4 lesions).
• To compare automated plans with manually created plans.

Main Methods:

• Thirty-five treatment plans (20 SL, 15 ML) were automatically re-planned using mCycle.
• Automated plans were compared to manual plans based on dose constraints, conformity, modulation complexity score (MCS), and delivery time.
• Radiation oncologists performed a blind comparison of automated and manual plans.

Main Results:

• Automated planning significantly improved target coverage and conformity for both SL and ML cases.
• mCycle demonstrated a clinically acceptable increase in brain V12Gy for ML cases.
• Automated plans were preferred in the majority of blind choices, with reduced planning time.

Conclusions:

• The mCycle system provides acceptable SRS plans with coplanar arcs, significantly reducing planning time for brain metastases.
• Automated planning shows promise for improving efficiency in SRS.
• Further research is recommended for non-coplanar arc plans.