"SABER": A new software tool for radiotherapy treatment plan evaluation

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Accurate radiotherapy treatment planning requires both spatial dose distribution and biological response information.
• Current commercial systems often lack the ability to fully integrate these crucial parameters for comprehensive plan evaluation.

Purpose of the Study:

• To develop an enhanced treatment plan evaluation tool incorporating biological parameters and retaining spatial dose information.
• To improve the optimization of treatment plans and the prediction of clinical outcomes in radiation therapy.

Main Methods:

• Developed a software system integrating hyper-radiosensitivity (induced-repair model) and dose convolution filter (DCF) for simulating dose effects.
• Introduced spatial DVH (sDVH) for evaluating spatial dose distribution and derived generalized equivalent uniform dose (gEUD and gEUD2).
• Implemented models for calculating tumor control probability (TCP), normal tissue complication probability (NTCP), and probability of uncomplicated tumor control (P+).

Main Results:

• The software distinguishes plan features not discernible by commercial systems by retaining spatial and biological dose information.
• Novel tools like sDVH and DCF can significantly alter the perceived plan quality and predicted metrics (TCP, NTCP).
• The calculation method and DCF application can change the ranking order of treatment plans, highlighting the importance of spatial and biological data integration.

Conclusions:

• The new software effectively integrates spatial and biological information into treatment planning.
• The order of applying biological models impacts plan ranking, emphasizing the model-dependent nature of evaluation.
• This tool aids in selecting more biologically optimal treatment plans and potentially predicting outcomes more accurately.