Presentation by Abdul-Amir Yassine at SPIE.PHOTONICS WESTAdapting to the patient: online monitoring for interstitial photodynamic therapy using machine learning Presenter/Author Abdul-Amir Yassine
Univ. of Toronto (Canada)
Author Lothar D. Lilge
Univ. Health Network (Canada)
Author Vaughn Betz
Univ. of Toronto (Canada)
Abstract
Treatment planning is of utmost importance in interstitial photodynamic therapy, as it predicts the required light delivery to the target volume in the upcoming treatment. However, it remains a major challenge due to uncertainties such as the tissue optical properties and the concentration of the photosensitizer and oxygen. Any small difference in these parameters from the assumed values during planning can significantly affect the outcome of the actual treatment. This work introduces a machine learning model to quickly recover the true optical properties during the treatment and re-optimize the source powers to attain the desired tumor coverage.