Adaptive Mathematical Model of Tumor Response to Radiotherapy Based on CBCT Data.

Area of Science:

• Oncology
• Medical Physics
• Computational Biology

Background:

• Mathematical modeling of tumor response to radiotherapy can enhance treatment planning but requires robust validation.
• Image-guided radiotherapy utilizes CT imaging for tumor delineation and cone beam CT for treatment monitoring.

Purpose of the Study:

• To develop and validate an adaptive macroscopic model for tumor growth and radiation response.
• To assess the model's ability to incorporate volumetric tumor data acquired during treatment.

Main Methods:

• A macroscopic model was developed to adapt to volumetric tumor data from cone beam CT scans.
• Model parameters were learned from 13 uterine cervical cancer patients, using group-specific and general parameter sets.
• Model adaptation and extrapolation performance were tested on independent patient data.

Main Results:

• The model demonstrated fitting errors of 13-21% for group-specific and general parameter sets.
• Online parameter tuning improved the extrapolation performance of the general model.
• Comparable prediction errors were achieved with online tuning, suggesting flexibility beyond strict patient stratification.

Conclusions:

• An adaptive macroscopic model can effectively integrate volumetric imaging data during radiotherapy.
• Online parameter tuning enhances model adaptability and prediction accuracy, mitigating issues with suboptimal patient stratification.
• This approach holds promise for improving personalized radiotherapy treatment planning and outcomes.