Artificial Intelligence for Improved Dosimetric Calculation in Radiotherapy

Cancer treatment by radiotherapy (RT) is essential yet complex, requiring precise radiation doses to effectively target tumors while sparing healthy tissues. This project addresses the frequent challenge of dose discrepancies-differences between the planned and delivered doses-due to factors such as patient anatomy, organ movement, and equipment variability. These inconsistencies can lead to increased toxicity or reduced treatment efficacy if left uncorrected. To tackle this, the project combines Monte Carlo (MC) simulations, known for their high accuracy, with machine learning to develop an advanced, AI-driven quality assurance (QA) system for radiotherapy. While MC simulations are accurate, they are also computationally demanding, often limiting their clinical application. By integrating machine learning, this project aims to streamline dose calculations, making them faster and more adaptable for clinical use. A predictive model will analyze patterns in dose discrepancies and proactively adjust for potential variations, enhancing accuracy and reducing treatment errors. This approach not only improves patient safety and treatment outcomes but also advances a more adaptive, data-driven approach to cancer treatment. By establishing a robust AI-enhanced QA framework, the project aims to set a new standard for precision in radiotherapy, benefiting both individual patients and broader clinical practice.

Faculty Supervisor:

Moussa Tembely

Student:

Partner:

Université Grenoble Alpes

Discipline:

Engineering

Sector:

Education

University:

Concordia University

Program: