Study of AI-Assisted Versus Manual Contouring for Breast Cancer Radiotherapy Using LOCATOR Software

What this Trial Is About

This research investigates the use of artificial intelligence AI to assist in contouring during radiation treatment for breast cancer patients. The trial compares AI-assisted contouring using the in-house LOCATOR software with fully manual contouring done by radiation oncologists. The study aims to evaluate whether AI-assisted contouring matches the quality of manual contouring and whether it can save oncologists time. It is a multicenter, phase II randomized controlled trial focusing on contour quality and efficiency. Participants are randomly assigned in a 31 ratio to either AI-assisted contouring or manual contouring groups. The AI model is initially trained on data from previous breast cancer cases and is updated regularly as more patients participate. In the AI-assisted group, contours are generated by the LOCATOR software and then manually reviewed and adjusted by oncologists. The manual group receives the standard care with contours created entirely by oncologists. During the study, researchers assess contour quality using the MD Anderson Contouring Grade Scale and various geometric and dosimetric measures. They also evaluate time savings, clinician-reported toxicity, patient-reported quality of life, and economic costs. Follow-up includes acute and late toxicity and quality of life assessments over 18 months to 5 years. Participant involvement includes contouring procedures, toxicity monitoring, questionnaires, and outcome evaluations over the study period.

Locally Optimised Contouring With AI Technology for Radiotherapy