Journal of Cancer Education (03/10/26) Beddok, Arnaud; Cohen, Emmanuel; Raymond, Gaëtan; et al.
Researchers sought to evaluate the educational efficacy and impact of a national webinar focused on artificial intelligence (AI)-based auto-contouring. AI-based automatic contouring is becoming increasingly common in radiation oncology, yet structured training for clinicians remains limited, prompting this study of the effects of a webinar designed to build foundational knowledge in supervised learning, convolutional neural networks, segmentation objectives, performance metrics, and the human-validation requirements of clinical implementation. In a prospective design, 33 professionals — a multidisciplinary group that included radiation oncologists, trainees, medical physicists, dosimetrists, and radiation therapists — attended the live online session, and 28 completed both the pre- and post-webinar assessment. Participants’ median score on a five-item multiple-choice test rose from 3.5 (IQR 2–4) before training to 5.0 (IQR 4–5) afterward, a statistically significant improvement demonstrating short-term gains in conceptual understanding. These findings underscore the value of scalable, clinically focused education to support the safe and informed integration of AI-based contouring tools into routine radiation therapy practice.
Read More